Compare commits

..

8 Commits

Author SHA1 Message Date
Alex Chi Z
cd32bb43a0 better wording
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-12 12:33:14 -04:00
Alex Chi Z
9a2068af56 feat(storcon): passthrough detach v2 requests
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-12 11:57:02 -04:00
Alex Chi Z
08babb4f50 resolve comments
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-12 11:40:50 -04:00
Alex Chi Z
ac8b4048fd fix err msg
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-11 15:15:56 -04:00
Alex Chi Z
5ea3662494 fix ancestor behavior
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-11 14:20:44 -04:00
Alex Chi Z
d29d506b63 no reparenting for behavior v2
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-11 13:38:25 -04:00
Alex Chi Z
b225b336e8 handle ts correctly
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-10 16:01:43 -04:00
Alex Chi Z
7d6c9d0557 feat(pageserver): support detaching from multiple ancestors
Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-03-10 15:45:14 -04:00
170 changed files with 1746 additions and 5372 deletions

View File

@@ -0,0 +1,21 @@
## Release 202Y-MM-DD
**NB: this PR must be merged only by 'Create a merge commit'!**
### Checklist when preparing for release
- [ ] Read or refresh [the release flow guide](https://www.notion.so/neondatabase/Release-general-flow-61f2e39fd45d4d14a70c7749604bd70b)
- [ ] Ask in the [cloud Slack channel](https://neondb.slack.com/archives/C033A2WE6BZ) that you are going to rollout the release. Any blockers?
- [ ] Does this release contain any db migrations? Destructive ones? What is the rollback plan?
<!-- List everything that should be done **before** release, any issues / setting changes / etc -->
### Checklist after release
- [ ] Make sure instructions from PRs included in this release and labeled `manual_release_instructions` are executed (either by you or by people who wrote them).
- [ ] Based on the merged commits write release notes and open a PR into `website` repo ([example](https://github.com/neondatabase/website/pull/219/files))
- [ ] Check [#dev-production-stream](https://neondb.slack.com/archives/C03F5SM1N02) Slack channel
- [ ] Check [stuck projects page](https://console.neon.tech/admin/projects?sort=last_active&order=desc&stuck=true)
- [ ] Check [recent operation failures](https://console.neon.tech/admin/operations?action=create_timeline%2Cstart_compute%2Cstop_compute%2Csuspend_compute%2Capply_config%2Cdelete_timeline%2Cdelete_tenant%2Ccreate_branch%2Ccheck_availability&sort=updated_at&order=desc&had_retries=some)
- [ ] Check [cloud SLO dashboard](https://neonprod.grafana.net/d/_oWcBMJ7k/cloud-slos?orgId=1)
- [ ] Check [compute startup metrics dashboard](https://neonprod.grafana.net/d/5OkYJEmVz/compute-startup-time)
<!-- List everything that should be done **after** release, any admin UI configuration / Grafana dashboard / alert changes / setting changes / etc -->

View File

@@ -1,16 +1,14 @@
import itertools
import json
import os
import sys
source_tag = os.getenv("SOURCE_TAG")
target_tag = os.getenv("TARGET_TAG")
branch = os.getenv("BRANCH")
dev_acr = os.getenv("DEV_ACR")
prod_acr = os.getenv("PROD_ACR")
dev_aws = os.getenv("DEV_AWS")
prod_aws = os.getenv("PROD_AWS")
aws_region = os.getenv("AWS_REGION")
build_tag = os.environ["BUILD_TAG"]
branch = os.environ["BRANCH"]
dev_acr = os.environ["DEV_ACR"]
prod_acr = os.environ["PROD_ACR"]
dev_aws = os.environ["DEV_AWS"]
prod_aws = os.environ["PROD_AWS"]
aws_region = os.environ["AWS_REGION"]
components = {
"neon": ["neon"],
@@ -41,23 +39,24 @@ registries = {
outputs: dict[str, dict[str, list[str]]] = {}
target_tags = [target_tag, "latest"] if branch == "main" else [target_tag]
target_stages = (
["dev", "prod"] if branch in ["release", "release-proxy", "release-compute"] else ["dev"]
)
target_tags = [build_tag, "latest"] if branch == "main" else [build_tag]
target_stages = ["dev", "prod"] if branch.startswith("release") else ["dev"]
for component_name, component_images in components.items():
for stage in target_stages:
outputs[f"{component_name}-{stage}"] = {
f"docker.io/neondatabase/{component_image}:{source_tag}": [
f"{registry}/{component_image}:{tag}"
for registry, tag in itertools.product(registries[stage], target_tags)
if not (registry == "docker.io/neondatabase" and tag == source_tag)
outputs[f"{component_name}-{stage}"] = dict(
[
(
f"docker.io/neondatabase/{component_image}:{build_tag}",
[
f"{combo[0]}/{component_image}:{combo[1]}"
for combo in itertools.product(registries[stage], target_tags)
],
)
for component_image in component_images
]
for component_image in component_images
}
)
with open(os.getenv("GITHUB_OUTPUT", "/dev/null"), "a") as f:
with open(os.environ["GITHUB_OUTPUT"], "a") as f:
for key, value in outputs.items():
f.write(f"{key}={json.dumps(value)}\n")
print(f"Image map for {key}:\n{json.dumps(value, indent=2)}\n\n", file=sys.stderr)

View File

@@ -1,110 +0,0 @@
#!/usr/bin/env bash
set -euo pipefail
DOCS_URL="https://docs.neon.build/overview/repositories/neon.html"
message() {
if [[ -n "${GITHUB_PR_NUMBER:-}" ]]; then
gh pr comment --repo "${GITHUB_REPOSITORY}" "${GITHUB_PR_NUMBER}" --edit-last --body "$1" \
|| gh pr comment --repo "${GITHUB_REPOSITORY}" "${GITHUB_PR_NUMBER}" --body "$1"
fi
echo "$1"
}
report_error() {
message "$1
For more details, see the documentation: ${DOCS_URL}"
exit 1
}
case "$RELEASE_BRANCH" in
"release") COMPONENT="Storage" ;;
"release-proxy") COMPONENT="Proxy" ;;
"release-compute") COMPONENT="Compute" ;;
*)
report_error "Unknown release branch: ${RELEASE_BRANCH}"
;;
esac
# Identify main and release branches
MAIN_BRANCH="origin/main"
REMOTE_RELEASE_BRANCH="origin/${RELEASE_BRANCH}"
# Find merge base
MERGE_BASE=$(git merge-base "${MAIN_BRANCH}" "${REMOTE_RELEASE_BRANCH}")
echo "Merge base of ${MAIN_BRANCH} and ${RELEASE_BRANCH}: ${MERGE_BASE}"
# Get the HEAD commit (last commit in PR, expected to be the merge commit)
LAST_COMMIT=$(git rev-parse HEAD)
MERGE_COMMIT_MESSAGE=$(git log -1 --format=%s "${LAST_COMMIT}")
EXPECTED_MESSAGE_REGEX="^$COMPONENT release [0-9]{4}-[0-9]{2}-[0-9]{2}$"
if ! [[ "${MERGE_COMMIT_MESSAGE}" =~ ${EXPECTED_MESSAGE_REGEX} ]]; then
report_error "Merge commit message does not match expected pattern: '<component> release YYYY-MM-DD'
Expected component: ${COMPONENT}
Found: '${MERGE_COMMIT_MESSAGE}'"
fi
echo "✅ Merge commit message is correctly formatted: '${MERGE_COMMIT_MESSAGE}'"
LAST_COMMIT_PARENTS=$(git cat-file -p "${LAST_COMMIT}" | jq -sR '[capture("parent (?<parent>[0-9a-f]{40})"; "g") | .parent]')
if [[ "$(echo "${LAST_COMMIT_PARENTS}" | jq 'length')" -ne 2 ]]; then
report_error "Last commit must be a merge commit with exactly two parents"
fi
EXPECTED_RELEASE_HEAD=$(git rev-parse "${REMOTE_RELEASE_BRANCH}")
if echo "${LAST_COMMIT_PARENTS}" | jq -e --arg rel "${EXPECTED_RELEASE_HEAD}" 'index($rel) != null' > /dev/null; then
LINEAR_HEAD=$(echo "${LAST_COMMIT_PARENTS}" | jq -r '[.[] | select(. != $rel)][0]' --arg rel "${EXPECTED_RELEASE_HEAD}")
else
report_error "Last commit must merge the release branch (${RELEASE_BRANCH})"
fi
echo "✅ Last commit correctly merges the previous commit and the release branch"
echo "Top commit of linear history: ${LINEAR_HEAD}"
MERGE_COMMIT_TREE=$(git rev-parse "${LAST_COMMIT}^{tree}")
LINEAR_HEAD_TREE=$(git rev-parse "${LINEAR_HEAD}^{tree}")
if [[ "${MERGE_COMMIT_TREE}" != "${LINEAR_HEAD_TREE}" ]]; then
report_error "Tree of merge commit (${MERGE_COMMIT_TREE}) does not match tree of linear history head (${LINEAR_HEAD_TREE})
This indicates that the merge of ${RELEASE_BRANCH} into this branch was not performed using the merge strategy 'ours'"
fi
echo "✅ Merge commit tree matches the linear history head"
EXPECTED_PREVIOUS_COMMIT="${LINEAR_HEAD}"
# Now traverse down the history, ensuring each commit has exactly one parent
CURRENT_COMMIT="${EXPECTED_PREVIOUS_COMMIT}"
while [[ "${CURRENT_COMMIT}" != "${MERGE_BASE}" && "${CURRENT_COMMIT}" != "${EXPECTED_RELEASE_HEAD}" ]]; do
CURRENT_COMMIT_PARENTS=$(git cat-file -p "${CURRENT_COMMIT}" | jq -sR '[capture("parent (?<parent>[0-9a-f]{40})"; "g") | .parent]')
if [[ "$(echo "${CURRENT_COMMIT_PARENTS}" | jq 'length')" -ne 1 ]]; then
report_error "Commit ${CURRENT_COMMIT} must have exactly one parent"
fi
NEXT_COMMIT=$(echo "${CURRENT_COMMIT_PARENTS}" | jq -r '.[0]')
if [[ "${NEXT_COMMIT}" == "${MERGE_BASE}" ]]; then
echo "✅ Reached merge base (${MERGE_BASE})"
PR_BASE="${MERGE_BASE}"
elif [[ "${NEXT_COMMIT}" == "${EXPECTED_RELEASE_HEAD}" ]]; then
echo "✅ Reached release branch (${EXPECTED_RELEASE_HEAD})"
PR_BASE="${EXPECTED_RELEASE_HEAD}"
elif [[ -z "${NEXT_COMMIT}" ]]; then
report_error "Unexpected end of commit history before reaching merge base"
fi
# Move to the next commit in the chain
CURRENT_COMMIT="${NEXT_COMMIT}"
done
echo "✅ All commits are properly ordered and linear"
echo "✅ Release PR structure is valid"
echo
message "Commits that are part of this release:
$(git log --oneline "${PR_BASE}..${LINEAR_HEAD}")"

View File

@@ -17,12 +17,6 @@
({};
.[$entry.component] |= (if . == null or $entry.version > .version then $entry else . end))
# Ensure that each component exists, or fail
| (["storage", "compute", "proxy"] - (keys)) as $missing
| if ($missing | length) > 0 then
"Error: Found no release for \($missing | join(", "))!\n" | halt_error(1)
else . end
# Convert the resulting object into an array of formatted strings
| to_entries
| map("\(.key)=\(.value.full)")

View File

@@ -7,8 +7,8 @@ on:
description: 'Component name'
required: true
type: string
source-branch:
description: 'Source branch'
release-branch:
description: 'Release branch'
required: true
type: string
secrets:
@@ -30,25 +30,17 @@ jobs:
steps:
- uses: actions/checkout@v4
with:
ref: ${{ inputs.source-branch }}
fetch-depth: 0
ref: main
- name: Set variables
id: vars
env:
COMPONENT_NAME: ${{ inputs.component-name }}
RELEASE_BRANCH: >-
${{
false
|| inputs.component-name == 'Storage' && 'release'
|| inputs.component-name == 'Proxy' && 'release-proxy'
|| inputs.component-name == 'Compute' && 'release-compute'
}}
RELEASE_BRANCH: ${{ inputs.release-branch }}
run: |
today=$(date +'%Y-%m-%d')
echo "title=${COMPONENT_NAME} release ${today}" | tee -a ${GITHUB_OUTPUT}
echo "rc-branch=rc/${RELEASE_BRANCH}/${today}" | tee -a ${GITHUB_OUTPUT}
echo "release-branch=${RELEASE_BRANCH}" | tee -a ${GITHUB_OUTPUT}
- name: Configure git
run: |
@@ -57,36 +49,31 @@ jobs:
- name: Create RC branch
env:
RELEASE_BRANCH: ${{ steps.vars.outputs.release-branch }}
RC_BRANCH: ${{ steps.vars.outputs.rc-branch }}
TITLE: ${{ steps.vars.outputs.title }}
run: |
git switch -c "${RC_BRANCH}"
git checkout -b "${RC_BRANCH}"
# Manually create a merge commit on the current branch, keeping the
# tree and setting the parents to the current HEAD and the HEAD of the
# release branch. This commit is what we'll fast-forward the release
# branch to when merging the release branch.
# For details on why, look at
# https://docs.neon.build/overview/repositories/neon.html#background-on-commit-history-of-release-prs
current_tree=$(git rev-parse 'HEAD^{tree}')
release_head=$(git rev-parse "origin/${RELEASE_BRANCH}")
current_head=$(git rev-parse HEAD)
merge_commit=$(git commit-tree -p "${current_head}" -p "${release_head}" -m "${TITLE}" "${current_tree}")
# Fast-forward the current branch to the newly created merge_commit
git merge --ff-only ${merge_commit}
# create an empty commit to distinguish workflow runs
# from other possible releases from the same commit
git commit --allow-empty -m "${TITLE}"
git push origin "${RC_BRANCH}"
- name: Create a PR into ${{ steps.vars.outputs.release-branch }}
- name: Create a PR into ${{ inputs.release-branch }}
env:
GH_TOKEN: ${{ secrets.ci-access-token }}
RC_BRANCH: ${{ steps.vars.outputs.rc-branch }}
RELEASE_BRANCH: ${{ steps.vars.outputs.release-branch }}
RELEASE_BRANCH: ${{ inputs.release-branch }}
TITLE: ${{ steps.vars.outputs.title }}
run: |
cat << EOF > body.md
## ${TITLE}
**Please merge this Pull Request using 'Create a merge commit' button**
EOF
gh pr create --title "${TITLE}" \
--body "" \
--body-file "body.md" \
--head "${RC_BRANCH}" \
--base "${RELEASE_BRANCH}"

View File

@@ -21,16 +21,9 @@ on:
run-kind:
description: "The kind of run we're currently in. Will be one of `push-main`, `storage-release`, `compute-release`, `proxy-release`, `storage-rc-pr`, `compute-rc-pr`, `proxy-rc-pr`, `pr`, or `workflow-dispatch`"
value: ${{ jobs.tags.outputs.run-kind }}
release-pr-run-id:
description: "Only available if `run-kind in [storage-release, proxy-release, compute-release]`. Contains the run ID of the `Build and Test` workflow, assuming one with the current commit can be found."
value: ${{ jobs.tags.outputs.release-pr-run-id }}
permissions: {}
defaults:
run:
shell: bash -euo pipefail {0}
jobs:
tags:
runs-on: ubuntu-22.04
@@ -40,7 +33,6 @@ jobs:
proxy: ${{ steps.previous-releases.outputs.proxy }}
storage: ${{ steps.previous-releases.outputs.storage }}
run-kind: ${{ steps.run-kind.outputs.run-kind }}
release-pr-run-id: ${{ steps.release-pr-run-id.outputs.release-pr-run-id }}
permissions:
contents: read
steps:
@@ -91,11 +83,7 @@ jobs:
echo "tag=release-compute-$(git rev-list --count HEAD)" | tee -a $GITHUB_OUTPUT
;;
pr|storage-rc-pr|compute-rc-pr|proxy-rc-pr)
BUILD_AND_TEST_RUN_ID=$(gh api --paginate \
-H "Accept: application/vnd.github+json" \
-H "X-GitHub-Api-Version: 2022-11-28" \
"/repos/${GITHUB_REPOSITORY}/actions/runs?head_sha=${CURRENT_SHA}&branch=${CURRENT_BRANCH}" \
| jq '[.workflow_runs[] | select(.name == "Build and Test")][0].id // ("Error: No matching workflow run found." | halt_error(1))')
BUILD_AND_TEST_RUN_ID=$(gh run list -b $CURRENT_BRANCH -c $CURRENT_SHA -w 'Build and Test' -L 1 --json databaseId --jq '.[].databaseId')
echo "tag=$BUILD_AND_TEST_RUN_ID" | tee -a $GITHUB_OUTPUT
;;
workflow-dispatch)
@@ -117,13 +105,3 @@ jobs:
"/repos/${GITHUB_REPOSITORY}/releases" \
| jq -f .github/scripts/previous-releases.jq -r \
| tee -a "${GITHUB_OUTPUT}"
- name: Get the release PR run ID
id: release-pr-run-id
if: ${{ contains(fromJson('["storage-release", "compute-release", "proxy-release"]'), steps.run-kind.outputs.run-kind) }}
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
CURRENT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
run: |
RELEASE_PR_RUN_ID=$(gh api "/repos/${GITHUB_REPOSITORY}/actions/runs?head_sha=$CURRENT_SHA" | jq '[.workflow_runs[] | select(.name == "Build and Test") | select(.head_branch | test("^rc/release(-(proxy|compute))?/[0-9]{4}-[0-9]{2}-[0-9]{2}$"; "s"))] | first | .id // ("Failed to find Build and Test run from RC PR!" | halt_error(1))')
echo "release-pr-run-id=$RELEASE_PR_RUN_ID" | tee -a $GITHUB_OUTPUT

View File

@@ -476,7 +476,7 @@ jobs:
(
!github.event.pull_request.draft
|| contains( github.event.pull_request.labels.*.name, 'run-e2e-tests-in-draft')
|| needs.meta.outputs.run-kind == 'push-main'
|| contains(fromJSON('["push-main", "storage-release", "proxy-release", "compute-release"]'), needs.meta.outputs.run-kind)
) && !failure() && !cancelled()
}}
needs: [ check-permissions, push-neon-image-dev, push-compute-image-dev, meta ]
@@ -487,7 +487,7 @@ jobs:
neon-image-arch:
needs: [ check-permissions, build-build-tools-image, meta ]
if: ${{ contains(fromJSON('["push-main", "pr", "storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
strategy:
matrix:
arch: [ x64, arm64 ]
@@ -537,7 +537,7 @@ jobs:
neon-image:
needs: [ neon-image-arch, meta ]
if: ${{ contains(fromJSON('["push-main", "pr", "storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
runs-on: ubuntu-22.04
permissions:
id-token: write # aws-actions/configure-aws-credentials
@@ -559,7 +559,7 @@ jobs:
compute-node-image-arch:
needs: [ check-permissions, build-build-tools-image, meta ]
if: ${{ contains(fromJSON('["push-main", "pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
permissions:
id-token: write # aws-actions/configure-aws-credentials
statuses: write
@@ -651,7 +651,7 @@ jobs:
compute-node-image:
needs: [ compute-node-image-arch, meta ]
if: ${{ contains(fromJSON('["push-main", "pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
permissions:
id-token: write # aws-actions/configure-aws-credentials
statuses: write
@@ -694,7 +694,7 @@ jobs:
vm-compute-node-image-arch:
needs: [ check-permissions, meta, compute-node-image ]
if: ${{ contains(fromJSON('["push-main", "pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
strategy:
fail-fast: false
@@ -747,7 +747,7 @@ jobs:
vm-compute-node-image:
needs: [ vm-compute-node-image-arch, meta ]
if: ${{ contains(fromJSON('["push-main", "pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
runs-on: ubuntu-22.04
strategy:
matrix:
@@ -773,12 +773,7 @@ jobs:
test-images:
needs: [ check-permissions, meta, neon-image, compute-node-image ]
# Depends on jobs that can get skipped
if: >-
${{
!failure()
&& !cancelled()
&& contains(fromJSON('["push-main", "pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind)
}}
if: "!failure() && !cancelled()"
strategy:
fail-fast: false
matrix:
@@ -805,7 +800,7 @@ jobs:
# Ensure that we don't have bad versions.
- name: Verify image versions
shell: bash # ensure no set -e for better error messages
if: ${{ contains(fromJSON('["push-main", "pr", "storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
run: |
pageserver_version=$(docker run --rm neondatabase/neon:${{ needs.meta.outputs.build-tag }} "/bin/sh" "-c" "/usr/local/bin/pageserver --version")
@@ -826,19 +821,19 @@ jobs:
env:
TAG: >-
${{
needs.meta.outputs.run-kind == 'compute-rc-pr'
contains(fromJSON('["compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind)
&& needs.meta.outputs.previous-storage-release
|| needs.meta.outputs.build-tag
}}
COMPUTE_TAG: >-
${{
contains(fromJSON('["storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind)
contains(fromJSON('["storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind)
&& needs.meta.outputs.previous-compute-release
|| needs.meta.outputs.build-tag
}}
TEST_EXTENSIONS_TAG: >-
${{
contains(fromJSON('["storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind)
contains(fromJSON('["storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind)
&& 'latest'
|| needs.meta.outputs.build-tag
}}
@@ -890,13 +885,7 @@ jobs:
id: generate
run: python3 .github/scripts/generate_image_maps.py
env:
SOURCE_TAG: >-
${{
contains(fromJson('["storage-release", "compute-release", "proxy-release"]'), needs.meta.outputs.run-kind)
&& needs.meta.outputs.release-pr-run-id
|| needs.meta.outputs.build-tag
}}
TARGET_TAG: ${{ needs.meta.outputs.build-tag }}
BUILD_TAG: "${{ needs.meta.outputs.build-tag }}"
BRANCH: "${{ github.ref_name }}"
DEV_ACR: "${{ vars.AZURE_DEV_REGISTRY_NAME }}"
PROD_ACR: "${{ vars.AZURE_PROD_REGISTRY_NAME }}"
@@ -906,7 +895,7 @@ jobs:
push-neon-image-dev:
needs: [ meta, generate-image-maps, neon-image ]
if: ${{ !failure() && !cancelled() && contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind) }}
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
@@ -924,7 +913,7 @@ jobs:
push-compute-image-dev:
needs: [ meta, generate-image-maps, vm-compute-node-image ]
if: ${{ !failure() && !cancelled() && contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
@@ -978,55 +967,16 @@ jobs:
acr-registry-name: ${{ vars.AZURE_PROD_REGISTRY_NAME }}
secrets: inherit
push-neon-test-extensions-image-ghcr:
if: ${{ contains(fromJSON('["push-main", "pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
# This is a bit of a special case so we're not using a generated image map.
add-latest-tag-to-neon-extensions-test-image:
if: github.ref_name == 'main'
needs: [ meta, compute-node-image ]
uses: ./.github/workflows/_push-to-container-registry.yml
with:
image-map: |
{
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}": [
"ghcr.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}"
],
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}": [
"ghcr.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}"
]
}
secrets: inherit
add-latest-tag-to-neon-test-extensions-image:
if: ${{ needs.meta.outputs.run-kind == 'push-main' }}
needs: [ meta, compute-node-image ]
uses: ./.github/workflows/_push-to-container-registry.yml
with:
image-map: |
{
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}": [
"docker.io/neondatabase/neon-test-extensions-v16:latest",
"ghcr.io/neondatabase/neon-test-extensions-v16:latest"
],
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}": [
"docker.io/neondatabase/neon-test-extensions-v17:latest",
"ghcr.io/neondatabase/neon-test-extensions-v17:latest"
]
}
secrets: inherit
add-release-tag-to-neon-test-extensions-image:
if: ${{ needs.meta.outputs.run-kind == 'compute-release' }}
needs: [ meta, compute-node-image ]
uses: ./.github/workflows/_push-to-container-registry.yml
with:
image-map: |
{
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.release-pr-run-id }}": [
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}",
"ghcr.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}"
],
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.release-pr-run-id }}": [
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}",
"ghcr.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}"
]
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.meta.outputs.build-tag }}": ["docker.io/neondatabase/neon-test-extensions-v16:latest"],
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.meta.outputs.build-tag }}": ["docker.io/neondatabase/neon-test-extensions-v17:latest"]
}
secrets: inherit
@@ -1225,7 +1175,7 @@ jobs:
-f deployPgSniRouter=false \
-f deployProxy=false \
-f deployStorage=true \
-f deployStorageBroker=false \
-f deployStorageBroker=true \
-f deployStorageController=true \
-f branch=main \
-f dockerTag=${{needs.meta.outputs.build-tag}} \
@@ -1233,7 +1183,7 @@ jobs:
gh workflow --repo neondatabase/infra run deploy-prod.yml --ref main \
-f deployStorage=true \
-f deployStorageBroker=false \
-f deployStorageBroker=true \
-f deployStorageController=true \
-f branch=main \
-f dockerTag=${{needs.meta.outputs.build-tag}}
@@ -1285,7 +1235,7 @@ jobs:
# The job runs on `release` branch and copies compatibility data and Neon artifact from the last *release PR* to the latest directory
promote-compatibility-data:
needs: [ meta, deploy ]
needs: [ deploy ]
permissions:
id-token: write # aws-actions/configure-aws-credentials
statuses: write
@@ -1295,6 +1245,37 @@ jobs:
runs-on: ubuntu-22.04
steps:
- name: Fetch GITHUB_RUN_ID and COMMIT_SHA for the last merged release PR
id: fetch-last-release-pr-info
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
branch_name_and_pr_number=$(gh pr list \
--repo "${GITHUB_REPOSITORY}" \
--base release \
--state merged \
--limit 10 \
--json mergeCommit,headRefName,number \
--jq ".[] | select(.mergeCommit.oid==\"${GITHUB_SHA}\") | { branch_name: .headRefName, pr_number: .number }")
branch_name=$(echo "${branch_name_and_pr_number}" | jq -r '.branch_name')
pr_number=$(echo "${branch_name_and_pr_number}" | jq -r '.pr_number')
run_id=$(gh run list \
--repo "${GITHUB_REPOSITORY}" \
--workflow build_and_test.yml \
--branch "${branch_name}" \
--json databaseId \
--limit 1 \
--jq '.[].databaseId')
last_commit_sha=$(gh pr view "${pr_number}" \
--repo "${GITHUB_REPOSITORY}" \
--json commits \
--jq '.commits[-1].oid')
echo "run-id=${run_id}" | tee -a ${GITHUB_OUTPUT}
echo "commit-sha=${last_commit_sha}" | tee -a ${GITHUB_OUTPUT}
- uses: aws-actions/configure-aws-credentials@v4
with:
aws-region: eu-central-1
@@ -1305,8 +1286,8 @@ jobs:
env:
BUCKET: neon-github-public-dev
AWS_REGION: eu-central-1
COMMIT_SHA: ${{ github.sha }}
RUN_ID: ${{ needs.meta.outputs.release-pr-run-id }}
COMMIT_SHA: ${{ steps.fetch-last-release-pr-info.outputs.commit-sha }}
RUN_ID: ${{ steps.fetch-last-release-pr-info.outputs.run-id }}
run: |
old_prefix="artifacts/${COMMIT_SHA}/${RUN_ID}"
new_prefix="artifacts/latest"
@@ -1395,5 +1376,5 @@ jobs:
|| needs.files-changed.result == 'skipped'
|| (needs.push-compute-image-dev.result == 'skipped' && contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
|| (needs.push-neon-image-dev.result == 'skipped' && contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind))
|| (needs.test-images.result == 'skipped' && contains(fromJSON('["push-main", "pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
|| needs.test-images.result == 'skipped'
|| (needs.trigger-custom-extensions-build-and-wait.result == 'skipped' && contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind))

View File

@@ -1,36 +0,0 @@
name: Fast forward merge
on:
pull_request:
types: [labeled]
branches:
- release
- release-proxy
- release-compute
jobs:
fast-forward:
if: ${{ github.event.label.name == 'fast-forward' }}
runs-on: ubuntu-22.04
steps:
- name: Remove fast-forward label to PR
env:
GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
run: |
gh pr edit ${{ github.event.pull_request.number }} --repo "${GITHUB_REPOSITORY}" --remove-label "fast-forward"
- name: Fast forwarding
uses: sequoia-pgp/fast-forward@ea7628bedcb0b0b96e94383ada458d812fca4979
# See https://docs.github.com/en/graphql/reference/enums#mergestatestatus
if: ${{ github.event.pull_request.mergeable_state == 'clean' }}
with:
merge: true
comment: on-error
github_token: ${{ secrets.CI_ACCESS_TOKEN }}
- name: Comment if mergeable_state is not clean
if: ${{ github.event.pull_request.mergeable_state != 'clean' }}
run: |
gh pr comment ${{ github.event.pull_request.number }} \
--repo "${GITHUB_REPOSITORY}" \
--body "Not trying to forward pull-request, because \`mergeable_state\` is \`${{ github.event.pull_request.mergeable_state }}\`, not \`clean\`."

View File

@@ -2,8 +2,8 @@ name: large oltp benchmark
on:
# uncomment to run on push for debugging your PR
#push:
# branches: [ bodobolero/synthetic_oltp_workload ]
push:
branches: [ bodobolero/synthetic_oltp_workload ]
schedule:
# * is a special character in YAML so you have to quote this string
@@ -12,7 +12,7 @@ on:
# │ │ ┌───────────── day of the month (1 - 31)
# │ │ │ ┌───────────── month (1 - 12 or JAN-DEC)
# │ │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
- cron: '0 15 * * 0,2,4' # run on Sunday, Tuesday, Thursday at 3 PM UTC
- cron: '0 15 * * *' # run once a day, timezone is utc, avoid conflict with other benchmarks
workflow_dispatch: # adds ability to run this manually
defaults:
@@ -22,7 +22,7 @@ defaults:
concurrency:
# Allow only one workflow globally because we need dedicated resources which only exist once
group: large-oltp-bench-workflow
cancel-in-progress: false
cancel-in-progress: true
jobs:
oltp:
@@ -31,9 +31,9 @@ jobs:
matrix:
include:
- target: new_branch
custom_scripts: insert_webhooks.sql@200 select_any_webhook_with_skew.sql@300 select_recent_webhook.sql@397 select_prefetch_webhook.sql@3 IUD_one_transaction.sql@100
custom_scripts: insert_webhooks.sql@2 select_any_webhook_with_skew.sql@4 select_recent_webhook.sql@4
- target: reuse_branch
custom_scripts: insert_webhooks.sql@200 select_any_webhook_with_skew.sql@300 select_recent_webhook.sql@397 select_prefetch_webhook.sql@3 IUD_one_transaction.sql@100
custom_scripts: insert_webhooks.sql@2 select_any_webhook_with_skew.sql@4 select_recent_webhook.sql@4
max-parallel: 1 # we want to run each stripe size sequentially to be able to compare the results
permissions:
contents: write
@@ -46,6 +46,7 @@ jobs:
PG_VERSION: 16 # pre-determined by pre-determined project
TEST_OUTPUT: /tmp/test_output
BUILD_TYPE: remote
SAVE_PERF_REPORT: ${{ github.ref_name == 'main' }}
PLATFORM: ${{ matrix.target }}
runs-on: [ self-hosted, us-east-2, x64 ]
@@ -56,10 +57,8 @@ jobs:
password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
options: --init
# Increase timeout to 2 days, default timeout is 6h - database maintenance can take a long time
# (normally 1h pgbench, 3h vacuum analyze 3.5h re-index) x 2 = 15h, leave some buffer for regressions
# in one run vacuum didn't finish within 12 hours
timeout-minutes: 2880
# Increase timeout to 8h, default timeout is 6h
timeout-minutes: 480
steps:
- uses: actions/checkout@v4
@@ -90,45 +89,29 @@ jobs:
- name: Set up Connection String
id: set-up-connstr
run: |
case "${{ matrix.target }}" in
new_branch)
CONNSTR=${{ steps.create-neon-branch-oltp-target.outputs.dsn }}
;;
reuse_branch)
CONNSTR=${{ secrets.BENCHMARK_LARGE_OLTP_REUSE_CONNSTR }}
;;
*)
echo >&2 "Unknown target=${{ matrix.target }}"
exit 1
;;
esac
case "${{ matrix.target }}" in
new_branch)
CONNSTR=${{ steps.create-neon-branch-oltp-target.outputs.dsn }}
;;
reuse_branch)
CONNSTR=${{ secrets.BENCHMARK_LARGE_OLTP_REUSE_CONNSTR }}
;;
*)
echo >&2 "Unknown target=${{ matrix.target }}"
exit 1
;;
esac
CONNSTR_WITHOUT_POOLER="${CONNSTR//-pooler/}"
echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
echo "connstr_without_pooler=${CONNSTR_WITHOUT_POOLER}" >> $GITHUB_OUTPUT
- name: Delete rows from prior runs in reuse branch
if: ${{ matrix.target == 'reuse_branch' }}
env:
BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr_without_pooler }}
PG_CONFIG: /tmp/neon/pg_install/v16/bin/pg_config
PSQL: /tmp/neon/pg_install/v16/bin/psql
PG_16_LIB_PATH: /tmp/neon/pg_install/v16/lib
run: |
echo "$(date '+%Y-%m-%d %H:%M:%S') - Deleting rows in table webhook.incoming_webhooks from prior runs"
export LD_LIBRARY_PATH=${PG_16_LIB_PATH}
${PSQL} "${BENCHMARK_CONNSTR}" -c "SET statement_timeout = 0; DELETE FROM webhook.incoming_webhooks WHERE created_at > '2025-02-27 23:59:59+00';"
echo "$(date '+%Y-%m-%d %H:%M:%S') - Finished deleting rows in table webhook.incoming_webhooks from prior runs"
- name: Benchmark pgbench with custom-scripts
- name: Benchmark pgbench with custom-scripts
uses: ./.github/actions/run-python-test-set
with:
build_type: ${{ env.BUILD_TYPE }}
test_selection: performance
run_in_parallel: false
save_perf_report: true
extra_params: -m remote_cluster --timeout 7200 -k test_perf_oltp_large_tenant_pgbench
save_perf_report: ${{ env.SAVE_PERF_REPORT }}
extra_params: -m remote_cluster --timeout 21600 -k test_perf_oltp_large_tenant
pg_version: ${{ env.PG_VERSION }}
aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
env:
@@ -136,21 +119,6 @@ jobs:
VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
- name: Benchmark database maintenance
uses: ./.github/actions/run-python-test-set
with:
build_type: ${{ env.BUILD_TYPE }}
test_selection: performance
run_in_parallel: false
save_perf_report: true
extra_params: -m remote_cluster --timeout 172800 -k test_perf_oltp_large_tenant_maintenance
pg_version: ${{ env.PG_VERSION }}
aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
env:
BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr_without_pooler }}
VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
- name: Delete Neon Branch for large tenant
if: ${{ always() && matrix.target == 'new_branch' }}
uses: ./.github/actions/neon-branch-delete
@@ -159,13 +127,6 @@ jobs:
branch_id: ${{ steps.create-neon-branch-oltp-target.outputs.branch_id }}
api_key: ${{ secrets.NEON_STAGING_API_KEY }}
- name: Configure AWS credentials # again because prior steps could have exceeded 5 hours
uses: aws-actions/configure-aws-credentials@v4
with:
aws-region: eu-central-1
role-to-assume: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
role-duration-seconds: 18000 # 5 hours
- name: Create Allure report
id: create-allure-report
if: ${{ !cancelled() }}

View File

@@ -1,24 +0,0 @@
name: Lint Release PR
on:
pull_request:
branches:
- release
- release-proxy
- release-compute
jobs:
lint-release-pr:
runs-on: ubuntu-22.04
steps:
- name: Checkout PR branch
uses: actions/checkout@v4
with:
fetch-depth: 0 # Fetch full history for git operations
ref: ${{ github.event.pull_request.head.ref }}
- name: Run lint script
env:
RELEASE_BRANCH: ${{ github.base_ref }}
run: |
./.github/scripts/lint-release-pr.sh

View File

@@ -8,6 +8,8 @@ on:
- .github/workflows/build-build-tools-image.yml
- .github/workflows/pre-merge-checks.yml
merge_group:
branches:
- main
defaults:
run:
@@ -17,17 +19,15 @@ defaults:
permissions: {}
jobs:
meta:
get-changed-files:
runs-on: ubuntu-22.04
outputs:
python-changed: ${{ steps.python-src.outputs.any_changed }}
rust-changed: ${{ steps.rust-src.outputs.any_changed }}
branch: ${{ steps.group-metadata.outputs.branch }}
pr-number: ${{ steps.group-metadata.outputs.pr-number }}
steps:
- uses: actions/checkout@v4
- uses: step-security/changed-files@3dbe17c78367e7d60f00d78ae6781a35be47b4a1 # v45.0.1
- uses: tj-actions/changed-files@4edd678ac3f81e2dc578756871e4d00c19191daf # v45.0.4
id: python-src
with:
files: |
@@ -38,7 +38,7 @@ jobs:
poetry.lock
pyproject.toml
- uses: step-security/changed-files@3dbe17c78367e7d60f00d78ae6781a35be47b4a1 # v45.0.1
- uses: tj-actions/changed-files@4edd678ac3f81e2dc578756871e4d00c19191daf # v45.0.4
id: rust-src
with:
files: |
@@ -58,20 +58,12 @@ jobs:
echo "${PYTHON_CHANGED_FILES}"
echo "${RUST_CHANGED_FILES}"
- name: Merge group metadata
if: ${{ github.event_name == 'merge_group' }}
id: group-metadata
env:
MERGE_QUEUE_REF: ${{ github.event.merge_group.head_ref }}
run: |
echo $MERGE_QUEUE_REF | jq -Rr 'capture("refs/heads/gh-readonly-queue/(?<branch>.*)/pr-(?<pr_number>[0-9]+)-[0-9a-f]{40}") | ["branch=" + .branch, "pr-number=" + .pr_number] | .[]' | tee -a "${GITHUB_OUTPUT}"
build-build-tools-image:
if: |
false
|| needs.meta.outputs.python-changed == 'true'
|| needs.meta.outputs.rust-changed == 'true'
needs: [ meta ]
|| needs.get-changed-files.outputs.python-changed == 'true'
|| needs.get-changed-files.outputs.rust-changed == 'true'
needs: [ get-changed-files ]
uses: ./.github/workflows/build-build-tools-image.yml
with:
# Build only one combination to save time
@@ -80,8 +72,8 @@ jobs:
secrets: inherit
check-codestyle-python:
if: needs.meta.outputs.python-changed == 'true'
needs: [ meta, build-build-tools-image ]
if: needs.get-changed-files.outputs.python-changed == 'true'
needs: [ get-changed-files, build-build-tools-image ]
uses: ./.github/workflows/_check-codestyle-python.yml
with:
# `-bookworm-x64` suffix should match the combination in `build-build-tools-image`
@@ -89,8 +81,8 @@ jobs:
secrets: inherit
check-codestyle-rust:
if: needs.meta.outputs.rust-changed == 'true'
needs: [ meta, build-build-tools-image ]
if: needs.get-changed-files.outputs.rust-changed == 'true'
needs: [ get-changed-files, build-build-tools-image ]
uses: ./.github/workflows/_check-codestyle-rust.yml
with:
# `-bookworm-x64` suffix should match the combination in `build-build-tools-image`
@@ -109,7 +101,7 @@ jobs:
statuses: write # for `github.repos.createCommitStatus(...)`
contents: write
needs:
- meta
- get-changed-files
- check-codestyle-python
- check-codestyle-rust
runs-on: ubuntu-22.04
@@ -137,20 +129,7 @@ jobs:
run: exit 1
if: |
false
|| (github.event_name == 'merge_group' && needs.meta.outputs.branch != 'main')
|| (needs.check-codestyle-python.result == 'skipped' && needs.meta.outputs.python-changed == 'true')
|| (needs.check-codestyle-rust.result == 'skipped' && needs.meta.outputs.rust-changed == 'true')
|| (needs.check-codestyle-python.result == 'skipped' && needs.get-changed-files.outputs.python-changed == 'true')
|| (needs.check-codestyle-rust.result == 'skipped' && needs.get-changed-files.outputs.rust-changed == 'true')
|| contains(needs.*.result, 'failure')
|| contains(needs.*.result, 'cancelled')
- name: Add fast-forward label to PR to trigger fast-forward merge
if: >-
${{
always()
&& github.event_name == 'merge_group'
&& contains(fromJson('["release", "release-proxy", "release-compute"]'), needs.meta.outputs.branch)
}}
env:
GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
run: >-
gh pr edit ${{ needs.meta.outputs.pr-number }} --repo "${GITHUB_REPOSITORY}" --add-label "fast-forward"

View File

@@ -38,7 +38,7 @@ jobs:
uses: ./.github/workflows/_create-release-pr.yml
with:
component-name: 'Storage'
source-branch: ${{ github.ref_name }}
release-branch: 'release'
secrets:
ci-access-token: ${{ secrets.CI_ACCESS_TOKEN }}
@@ -51,7 +51,7 @@ jobs:
uses: ./.github/workflows/_create-release-pr.yml
with:
component-name: 'Proxy'
source-branch: ${{ github.ref_name }}
release-branch: 'release-proxy'
secrets:
ci-access-token: ${{ secrets.CI_ACCESS_TOKEN }}
@@ -64,6 +64,6 @@ jobs:
uses: ./.github/workflows/_create-release-pr.yml
with:
component-name: 'Compute'
source-branch: ${{ github.ref_name }}
release-branch: 'release-compute'
secrets:
ci-access-token: ${{ secrets.CI_ACCESS_TOKEN }}

308
Cargo.lock generated
View File

@@ -191,7 +191,7 @@ checksum = "965c2d33e53cb6b267e148a4cb0760bc01f4904c1cd4bb4002a085bb016d1490"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
"synstructure",
]
@@ -203,7 +203,7 @@ checksum = "7b18050c2cd6fe86c3a76584ef5e0baf286d038cda203eb6223df2cc413565f7"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -272,7 +272,7 @@ checksum = "16e62a023e7c117e27523144c5d2459f4397fcc3cab0085af8e2224f643a0193"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -283,7 +283,7 @@ checksum = "b9ccdd8f2a161be9bd5c023df56f1b2a0bd1d83872ae53b71a84a12c9bf6e842"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1021,7 +1021,7 @@ dependencies = [
"regex",
"rustc-hash 2.1.1",
"shlex",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1248,7 +1248,7 @@ dependencies = [
"heck",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1309,7 +1309,6 @@ version = "0.1.0"
dependencies = [
"anyhow",
"chrono",
"indexmap 2.0.1",
"jsonwebtoken",
"regex",
"remote_storage",
@@ -1340,7 +1339,6 @@ dependencies = [
"flate2",
"futures",
"http 1.1.0",
"indexmap 2.0.1",
"jsonwebtoken",
"metrics",
"nix 0.27.1",
@@ -1349,20 +1347,17 @@ dependencies = [
"once_cell",
"opentelemetry",
"opentelemetry_sdk",
"p256 0.13.2",
"postgres",
"postgres_initdb",
"regex",
"remote_storage",
"reqwest",
"ring",
"rlimit",
"rust-ini",
"serde",
"serde_json",
"serde_with",
"signal-hook",
"spki 0.7.3",
"tar",
"thiserror 1.0.69",
"tokio",
@@ -1382,7 +1377,6 @@ dependencies = [
"vm_monitor",
"walkdir",
"workspace_hack",
"x509-cert",
"zstd",
]
@@ -1709,7 +1703,7 @@ checksum = "f46882e17999c6cc590af592290432be3bce0428cb0d5f8b6715e4dc7b383eb3"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1733,7 +1727,7 @@ dependencies = [
"proc-macro2",
"quote",
"strsim 0.10.0",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1744,7 +1738,7 @@ checksum = "29a358ff9f12ec09c3e61fef9b5a9902623a695a46a917b07f269bff1445611a"
dependencies = [
"darling_core",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1807,8 +1801,6 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "fffa369a668c8af7dbf8b5e56c9f744fbd399949ed171606040001947de40b1c"
dependencies = [
"const-oid",
"der_derive",
"flagset",
"pem-rfc7468",
"zeroize",
]
@@ -1827,17 +1819,6 @@ dependencies = [
"rusticata-macros",
]
[[package]]
name = "der_derive"
version = "0.7.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8034092389675178f570469e6c3b0465d3d30b4505c294a6550db47f3c17ad18"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "deranged"
version = "0.3.11"
@@ -1907,7 +1888,7 @@ dependencies = [
"dsl_auto_type",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1927,7 +1908,7 @@ version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "209c735641a413bc68c4923a9d6ad4bcb3ca306b794edaa7eb0b3228a99ffb25"
dependencies = [
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1956,7 +1937,7 @@ checksum = "487585f4d0c6655fe74905e2504d8ad6908e4db67f744eb140876906c2f3175d"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -1979,7 +1960,7 @@ dependencies = [
"heck",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -2124,7 +2105,7 @@ dependencies = [
"darling",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -2134,19 +2115,28 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "186e05a59d4c50738528153b83b0b0194d3a29507dfec16eccd4b342903397d0"
dependencies = [
"log",
"regex",
]
[[package]]
name = "env_logger"
version = "0.11.7"
version = "0.10.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c3716d7a920fb4fac5d84e9d4bce8ceb321e9414b4409da61b07b75c1e3d0697"
checksum = "4cd405aab171cb85d6735e5c8d9db038c17d3ca007a4d2c25f337935c3d90580"
dependencies = [
"humantime",
"is-terminal",
"log",
"regex",
"termcolor",
]
[[package]]
name = "env_logger"
version = "0.11.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6c012a26a7f605efc424dd53697843a72be7dc86ad2d01f7814337794a12231d"
dependencies = [
"anstream",
"anstyle",
"env_filter",
"jiff",
"log",
]
@@ -2167,7 +2157,7 @@ checksum = "3bf679796c0322556351f287a51b49e48f7c4986e727b5dd78c972d30e2e16cc"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -2301,12 +2291,6 @@ version = "0.4.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0ce7134b9999ecaf8bcd65542e436736ef32ddca1b3e06094cb6ec5755203b80"
[[package]]
name = "flagset"
version = "0.4.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b3ea1ec5f8307826a5b71094dd91fc04d4ae75d5709b20ad351c7fb4815c86ec"
[[package]]
name = "flate2"
version = "1.0.26"
@@ -2433,7 +2417,7 @@ checksum = "162ee34ebcb7c64a8abebc059ce0fee27c2262618d7b60ed8faf72fef13c3650"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -2546,7 +2530,7 @@ checksum = "53010ccb100b96a67bc32c0175f0ed1426b31b655d562898e57325f81c023ac0"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -2863,7 +2847,6 @@ version = "0.1.0"
dependencies = [
"anyhow",
"bytes",
"camino",
"fail",
"futures",
"hyper 0.14.30",
@@ -2874,7 +2857,6 @@ dependencies = [
"pprof",
"regex",
"routerify",
"rustls-pemfile 2.1.1",
"serde",
"serde_json",
"serde_path_to_error",
@@ -2904,9 +2886,9 @@ checksum = "c4a1e36c821dbe04574f602848a19f742f4fb3c98d40449f11bcad18d6b17421"
[[package]]
name = "humantime"
version = "2.2.0"
version = "2.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9b112acc8b3adf4b107a8ec20977da0273a8c386765a3ec0229bd500a1443f9f"
checksum = "9a3a5bfb195931eeb336b2a7b4d761daec841b97f947d34394601737a7bba5e4"
[[package]]
name = "humantime-serde"
@@ -3166,7 +3148,7 @@ checksum = "1ec89e9337638ecdc08744df490b221a7399bf8d164eb52a665454e60e075ad6"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -3259,7 +3241,7 @@ dependencies = [
"crossbeam-channel",
"crossbeam-utils",
"dashmap 6.1.0",
"env_logger",
"env_logger 0.11.2",
"indexmap 2.0.1",
"itoa",
"log",
@@ -3272,11 +3254,11 @@ dependencies = [
[[package]]
name = "inotify"
version = "0.9.6"
version = "0.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f8069d3ec154eb856955c1c0fbffefbf5f3c40a104ec912d4797314c1801abff"
checksum = "f37dccff2791ab604f9babef0ba14fbe0be30bd368dc541e2b08d07c8aa908f3"
dependencies = [
"bitflags 1.3.2",
"bitflags 2.8.0",
"inotify-sys",
"libc",
]
@@ -3382,30 +3364,6 @@ dependencies = [
"tracing",
]
[[package]]
name = "jiff"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d699bc6dfc879fb1bf9bdff0d4c56f0884fc6f0d0eb0fba397a6d00cd9a6b85e"
dependencies = [
"jiff-static",
"log",
"portable-atomic",
"portable-atomic-util",
"serde",
]
[[package]]
name = "jiff-static"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8d16e75759ee0aa64c57a56acbf43916987b20c77373cb7e808979e02b93c9f9"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "jobserver"
version = "0.1.32"
@@ -3577,9 +3535,9 @@ dependencies = [
[[package]]
name = "log"
version = "0.4.26"
version = "0.4.20"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "30bde2b3dc3671ae49d8e2e9f044c7c005836e7a023ee57cffa25ab82764bb9e"
checksum = "b5e6163cb8c49088c2c36f57875e58ccd8c87c7427f7fbd50ea6710b2f3f2e8f"
[[package]]
name = "lru"
@@ -3660,7 +3618,7 @@ dependencies = [
"heck",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -3774,18 +3732,6 @@ dependencies = [
"adler2",
]
[[package]]
name = "mio"
version = "0.8.11"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a4a650543ca06a924e8b371db273b2756685faae30f8487da1b56505a8f78b0c"
dependencies = [
"libc",
"log",
"wasi 0.11.0+wasi-snapshot-preview1",
"windows-sys 0.48.0",
]
[[package]]
name = "mio"
version = "1.0.3"
@@ -3793,6 +3739,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2886843bf800fba2e3377cff24abf6379b4c4d5c6681eaf9ea5b0d15090450bd"
dependencies = [
"libc",
"log",
"wasi 0.11.0+wasi-snapshot-preview1",
"windows-sys 0.52.0",
]
@@ -3870,23 +3817,29 @@ checksum = "38bf9645c8b145698bb0b18a4637dcacbc421ea49bef2317e4fd8065a387cf21"
[[package]]
name = "notify"
version = "6.1.1"
version = "8.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6205bd8bb1e454ad2e27422015fb5e4f2bcc7e08fa8f27058670d208324a4d2d"
checksum = "2fee8403b3d66ac7b26aee6e40a897d85dc5ce26f44da36b8b73e987cc52e943"
dependencies = [
"bitflags 2.8.0",
"crossbeam-channel",
"filetime",
"fsevent-sys",
"inotify",
"kqueue",
"libc",
"log",
"mio 0.8.11",
"mio",
"notify-types",
"walkdir",
"windows-sys 0.48.0",
"windows-sys 0.59.0",
]
[[package]]
name = "notify-types"
version = "2.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5e0826a989adedc2a244799e823aece04662b66609d96af8dff7ac6df9a8925d"
[[package]]
name = "ntapi"
version = "0.4.1"
@@ -4329,6 +4282,8 @@ dependencies = [
"reqwest",
"rpds",
"rustls 0.23.18",
"rustls-pemfile 2.1.1",
"rustls-pki-types",
"scopeguard",
"send-future",
"serde",
@@ -4353,7 +4308,6 @@ dependencies = [
"tokio-util",
"toml_edit",
"tracing",
"tracing-utils",
"url",
"utils",
"uuid",
@@ -4531,7 +4485,7 @@ dependencies = [
"parquet",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -4633,7 +4587,7 @@ checksum = "f6e859e6e5bd50440ab63c47e3ebabc90f26251f7c73c3d3e837b74a1cc3fa67"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -4729,15 +4683,6 @@ version = "1.10.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "280dc24453071f1b63954171985a0b0d30058d287960968b9b2aca264c8d4ee6"
[[package]]
name = "portable-atomic-util"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d8a2f0d8d040d7848a709caf78912debcc3f33ee4b3cac47d73d1e1069e83507"
dependencies = [
"portable-atomic",
]
[[package]]
name = "postgres"
version = "0.19.7"
@@ -4845,7 +4790,7 @@ dependencies = [
"bytes",
"crc32c",
"criterion",
"env_logger",
"env_logger 0.10.2",
"log",
"memoffset 0.9.0",
"once_cell",
@@ -4944,7 +4889,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8d3928fb5db768cb86f891ff014f0144589297e3c6a1aba6ed7cecfdace270c7"
dependencies = [
"proc-macro2",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -4958,9 +4903,9 @@ dependencies = [
[[package]]
name = "proc-macro2"
version = "1.0.94"
version = "1.0.92"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a31971752e70b8b2686d7e46ec17fb38dad4051d94024c88df49b667caea9c84"
checksum = "37d3544b3f2748c54e147655edb5025752e2303145b5aefb3c3ea2c78b973bb0"
dependencies = [
"unicode-ident",
]
@@ -5035,7 +4980,7 @@ checksum = "22505a5c94da8e3b7c2996394d1c933236c4d743e81a410bcca4e6989fc066a4"
dependencies = [
"bytes",
"heck",
"itertools 0.12.1",
"itertools 0.10.5",
"log",
"multimap",
"once_cell",
@@ -5044,7 +4989,7 @@ dependencies = [
"prost 0.12.6",
"prost-types 0.12.6",
"regex",
"syn 2.0.100",
"syn 2.0.90",
"tempfile",
]
@@ -5065,7 +5010,7 @@ dependencies = [
"prost 0.13.3",
"prost-types 0.13.3",
"regex",
"syn 2.0.100",
"syn 2.0.90",
"tempfile",
]
@@ -5076,10 +5021,10 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "81bddcdb20abf9501610992b6759a4c888aef7d1a7247ef75e2404275ac24af1"
dependencies = [
"anyhow",
"itertools 0.12.1",
"itertools 0.10.5",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -5092,7 +5037,7 @@ dependencies = [
"itertools 0.12.1",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -5139,7 +5084,7 @@ dependencies = [
"consumption_metrics",
"ecdsa 0.16.9",
"ed25519-dalek",
"env_logger",
"env_logger 0.10.2",
"fallible-iterator",
"flate2",
"framed-websockets",
@@ -5276,9 +5221,9 @@ dependencies = [
[[package]]
name = "quote"
version = "1.0.39"
version = "1.0.37"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c1f1914ce909e1658d9907913b4b91947430c7d9be598b15a1912935b8c04801"
checksum = "b5b9d34b8991d19d98081b46eacdd8eb58c6f2b201139f7c5f643cc155a633af"
dependencies = [
"proc-macro2",
]
@@ -5807,7 +5752,7 @@ dependencies = [
"regex",
"relative-path",
"rustc_version",
"syn 2.0.100",
"syn 2.0.90",
"unicode-ident",
]
@@ -6022,7 +5967,7 @@ dependencies = [
"crc32c",
"criterion",
"desim",
"env_logger",
"env_logger 0.10.2",
"fail",
"futures",
"hex",
@@ -6044,7 +5989,6 @@ dependencies = [
"regex",
"remote_storage",
"reqwest",
"rustls 0.23.18",
"safekeeper_api",
"safekeeper_client",
"scopeguard",
@@ -6061,7 +6005,6 @@ dependencies = [
"tokio",
"tokio-io-timeout",
"tokio-postgres",
"tokio-rustls 0.26.0",
"tokio-stream",
"tokio-tar",
"tokio-util",
@@ -6355,7 +6298,7 @@ checksum = "ad1e866f866923f252f05c889987993144fb74e722403468a4ebd70c3cd756c0"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -6437,7 +6380,7 @@ dependencies = [
"darling",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -6452,9 +6395,9 @@ dependencies = [
[[package]]
name = "sha1"
version = "0.10.6"
version = "0.10.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e3bf829a2d51ab4a5ddf1352d8470c140cadc8301b2ae1789db023f01cedd6ba"
checksum = "f04293dc80c3993519f2d7f6f511707ee7094fe0c6d3406feb330cdb3540eba3"
dependencies = [
"cfg-if",
"cpufeatures",
@@ -6678,7 +6621,6 @@ dependencies = [
"bytes",
"chrono",
"clap",
"clashmap",
"control_plane",
"cron",
"diesel",
@@ -6839,7 +6781,7 @@ dependencies = [
"proc-macro2",
"quote",
"rustversion",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -6890,9 +6832,9 @@ dependencies = [
[[package]]
name = "syn"
version = "2.0.100"
version = "2.0.90"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b09a44accad81e1ba1cd74a32461ba89dee89095ba17b32f5d03683b1b1fc2a0"
checksum = "919d3b74a5dd0ccd15aeb8f93e7006bd9e14c295087c9896a110f490752bcf31"
dependencies = [
"proc-macro2",
"quote",
@@ -6922,7 +6864,7 @@ checksum = "c8af7666ab7b6390ab78131fb5b0fce11d6b7a6951602017c35fa82800708971"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -6973,6 +6915,15 @@ dependencies = [
"serde_json",
]
[[package]]
name = "termcolor"
version = "1.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "be55cf8942feac5c765c2c993422806843c9a9a45d4d5c407ad6dd2ea95eb9b6"
dependencies = [
"winapi-util",
]
[[package]]
name = "test-context"
version = "0.3.0"
@@ -6991,7 +6942,7 @@ checksum = "78ea17a2dc368aeca6f554343ced1b1e31f76d63683fa8016e5844bd7a5144a1"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7020,7 +6971,7 @@ checksum = "4fee6c4efc90059e10f81e6d42c60a18f76588c3d74cb83a0b242a2b6c7504c1"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7031,7 +6982,7 @@ checksum = "26afc1baea8a989337eeb52b6e72a039780ce45c3edfcc9c5b9d112feeb173c2"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7162,27 +7113,6 @@ version = "0.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
[[package]]
name = "tls_codec"
version = "0.4.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0de2e01245e2bb89d6f05801c564fa27624dbd7b1846859876c7dad82e90bf6b"
dependencies = [
"tls_codec_derive",
"zeroize",
]
[[package]]
name = "tls_codec_derive"
version = "0.4.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2d2e76690929402faae40aebdda620a2c0e25dd6d3b9afe48867dfd95991f4bd"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "tokio"
version = "1.43.0"
@@ -7192,7 +7122,7 @@ dependencies = [
"backtrace",
"bytes",
"libc",
"mio 1.0.3",
"mio",
"parking_lot 0.12.1",
"pin-project-lite",
"signal-hook-registry",
@@ -7235,7 +7165,7 @@ checksum = "6e06d43f1345a3bcd39f6a56dbb7dcab2ba47e68e8ac134855e7e2bdbaf8cab8"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7468,7 +7398,7 @@ dependencies = [
"prost-build 0.13.3",
"prost-types 0.13.3",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7583,7 +7513,7 @@ checksum = "395ae124c09f9e6918a2310af6038fba074bcf474ac352496d5910dd59a2226d"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -7914,7 +7844,6 @@ dependencies = [
"tracing",
"tracing-error",
"tracing-subscriber",
"tracing-utils",
"walkdir",
]
@@ -7978,7 +7907,7 @@ dependencies = [
"anyhow",
"camino-tempfile",
"clap",
"env_logger",
"env_logger 0.10.2",
"log",
"postgres",
"postgres_ffi",
@@ -8083,7 +8012,7 @@ dependencies = [
"once_cell",
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
"wasm-bindgen-shared",
]
@@ -8117,7 +8046,7 @@ checksum = "e94f17b526d0a461a191c78ea52bbce64071ed5c04c9ffe424dcb38f74171bb7"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
"wasm-bindgen-backend",
"wasm-bindgen-shared",
]
@@ -8424,7 +8353,6 @@ name = "workspace_hack"
version = "0.1.0"
dependencies = [
"ahash",
"anstream",
"anyhow",
"base64 0.13.1",
"base64 0.21.7",
@@ -8435,17 +8363,12 @@ dependencies = [
"chrono",
"clap",
"clap_builder",
"const-oid",
"crypto-bigint 0.5.5",
"der 0.7.8",
"deranged",
"digest",
"displaydoc",
"ecdsa 0.16.9",
"either",
"elliptic-curve 0.13.8",
"env_filter",
"env_logger",
"fail",
"form_urlencoded",
"futures-channel",
@@ -8463,6 +8386,7 @@ dependencies = [
"hyper-util",
"indexmap 1.9.3",
"indexmap 2.0.1",
"itertools 0.10.5",
"itertools 0.12.1",
"lazy_static",
"libc",
@@ -8478,7 +8402,6 @@ dependencies = [
"num-rational",
"num-traits",
"once_cell",
"p256 0.13.2",
"parquet",
"prettyplease",
"proc-macro2",
@@ -8491,7 +8414,6 @@ dependencies = [
"reqwest",
"rustls 0.23.18",
"scopeguard",
"sec1 0.7.3",
"serde",
"serde_json",
"sha2",
@@ -8500,7 +8422,7 @@ dependencies = [
"spki 0.7.3",
"stable_deref_trait",
"subtle",
"syn 2.0.100",
"syn 2.0.90",
"sync_wrapper 0.1.2",
"tikv-jemalloc-ctl",
"tikv-jemalloc-sys",
@@ -8537,18 +8459,6 @@ version = "0.5.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1e9df38ee2d2c3c5948ea468a8406ff0db0b29ae1ffde1bcf20ef305bcc95c51"
[[package]]
name = "x509-cert"
version = "0.2.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1301e935010a701ae5f8655edc0ad17c44bad3ac5ce8c39185f75453b720ae94"
dependencies = [
"const-oid",
"der 0.7.8",
"spki 0.7.3",
"tls_codec",
]
[[package]]
name = "x509-certificate"
version = "0.23.1"
@@ -8629,7 +8539,7 @@ checksum = "2380878cad4ac9aac1e2435f3eb4020e8374b5f13c296cb75b4620ff8e229154"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
"synstructure",
]
@@ -8651,7 +8561,7 @@ checksum = "b3c129550b3e6de3fd0ba67ba5c81818f9805e58b8d7fee80a3a59d2c9fc601a"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -8671,15 +8581,15 @@ checksum = "595eed982f7d355beb85837f651fa22e90b3c044842dc7f2c2842c086f295808"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
"synstructure",
]
[[package]]
name = "zeroize"
version = "1.8.1"
version = "1.7.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ced3678a2879b30306d323f4542626697a464a97c0a07c9aebf7ebca65cd4dde"
checksum = "525b4ec142c6b68a2d10f01f7bbf6755599ca3f81ea53b8431b7dd348f5fdb2d"
dependencies = [
"serde",
"zeroize_derive",
@@ -8693,7 +8603,7 @@ checksum = "ce36e65b0d2999d2aafac989fb249189a141aee1f53c612c1f37d72631959f69"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]
@@ -8715,7 +8625,7 @@ checksum = "6eafa6dfb17584ea3e2bd6e76e0cc15ad7af12b09abdd1ca55961bed9b1063c6"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
"syn 2.0.90",
]
[[package]]

View File

@@ -106,13 +106,13 @@ hostname = "0.4"
http = {version = "1.1.0", features = ["std"]}
http-types = { version = "2", default-features = false }
http-body-util = "0.1.2"
humantime = "2.2"
humantime = "2.1"
humantime-serde = "1.1.1"
hyper0 = { package = "hyper", version = "0.14" }
hyper = "1.4"
hyper-util = "0.1"
tokio-tungstenite = "0.21.0"
indexmap = { version = "2", features = ["serde"] }
indexmap = "2"
indoc = "2"
ipnet = "2.10.0"
itertools = "0.10"
@@ -126,9 +126,7 @@ measured = { version = "0.0.22", features=["lasso"] }
measured-process = { version = "0.0.22" }
memoffset = "0.9"
nix = { version = "0.27", features = ["dir", "fs", "process", "socket", "signal", "poll"] }
# Do not update to >= 7.0.0, at least. The update will have a significant impact
# on compute startup metrics (start_postgres_ms), >= 25% degradation.
notify = "6.0.0"
notify = "8.0.0"
num_cpus = "1.15"
num-traits = "0.2.15"
once_cell = "1.13"
@@ -221,7 +219,7 @@ zerocopy = { version = "0.7", features = ["derive"] }
json-structural-diff = { version = "0.2.0" }
## TODO replace this with tracing
env_logger = "0.11"
env_logger = "0.10"
log = "0.4"
## Libraries from neondatabase/ git forks, ideally with changes to be upstreamed

View File

@@ -1735,8 +1735,6 @@ RUN set -e \
libevent-dev \
libtool \
pkg-config \
libcurl4-openssl-dev \
libssl-dev \
&& apt clean && rm -rf /var/lib/apt/lists/*
# Use `dist_man_MANS=` to skip manpage generation (which requires python3/pandoc)
@@ -1745,7 +1743,7 @@ RUN set -e \
&& git clone --recurse-submodules --depth 1 --branch ${PGBOUNCER_TAG} https://github.com/pgbouncer/pgbouncer.git pgbouncer \
&& cd pgbouncer \
&& ./autogen.sh \
&& ./configure --prefix=/usr/local/pgbouncer \
&& ./configure --prefix=/usr/local/pgbouncer --without-openssl \
&& make -j $(nproc) dist_man_MANS= \
&& make install dist_man_MANS=

View File

@@ -145,7 +145,7 @@ merge: |
COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
RUN chmod 0666 /etc/compute_rsyslog.conf
RUN mkdir /var/log/rsyslog && chown -R postgres /var/log/rsyslog
RUN chmod 0666 /var/log/
COPY --from=libcgroup-builder /libcgroup-install/bin/* /usr/bin/

View File

@@ -140,7 +140,7 @@ merge: |
COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
RUN chmod 0666 /etc/compute_rsyslog.conf
RUN mkdir /var/log/rsyslog && chown -R postgres /var/log/rsyslog
RUN chmod 0666 /var/log/
COPY --from=libcgroup-builder /libcgroup-install/bin/* /usr/bin/

View File

@@ -26,7 +26,6 @@ fail.workspace = true
flate2.workspace = true
futures.workspace = true
http.workspace = true
indexmap.workspace = true
jsonwebtoken.workspace = true
metrics.workspace = true
nix.workspace = true
@@ -35,19 +34,16 @@ num_cpus.workspace = true
once_cell.workspace = true
opentelemetry.workspace = true
opentelemetry_sdk.workspace = true
p256 = { version = "0.13", features = ["pem"] }
postgres.workspace = true
regex.workspace = true
reqwest = { workspace = true, features = ["json"] }
ring = "0.17"
serde.workspace = true
serde_with.workspace = true
serde_json.workspace = true
signal-hook.workspace = true
spki = { version = "0.7.3", features = ["std"] }
tar.workspace = true
tower.workspace = true
tower-http.workspace = true
reqwest = { workspace = true, features = ["json"] }
tokio = { workspace = true, features = ["rt", "rt-multi-thread"] }
tokio-postgres.workspace = true
tokio-util.workspace = true
@@ -61,7 +57,6 @@ thiserror.workspace = true
url.workspace = true
uuid.workspace = true
walkdir.workspace = true
x509-cert = { version = "0.2.5" }
postgres_initdb.workspace = true
compute_api.workspace = true

View File

@@ -37,11 +37,10 @@ use crate::logger::startup_context_from_env;
use crate::lsn_lease::launch_lsn_lease_bg_task_for_static;
use crate::monitor::launch_monitor;
use crate::pg_helpers::*;
use crate::rsyslog::{configure_audit_rsyslog, launch_pgaudit_gc};
use crate::rsyslog::configure_audit_rsyslog;
use crate::spec::*;
use crate::swap::resize_swap;
use crate::sync_sk::{check_if_synced, ping_safekeeper};
use crate::tls::watch_cert_for_changes;
use crate::{config, extension_server, local_proxy};
pub static SYNC_SAFEKEEPERS_PID: AtomicU32 = AtomicU32::new(0);
@@ -113,7 +112,6 @@ pub struct ComputeNode {
// key: ext_archive_name, value: started download time, download_completed?
pub ext_download_progress: RwLock<HashMap<String, (DateTime<Utc>, bool)>>,
pub compute_ctl_config: ComputeCtlConfig,
}
// store some metrics about download size that might impact startup time
@@ -137,6 +135,8 @@ pub struct ComputeState {
/// passed by the control plane with a /configure HTTP request.
pub pspec: Option<ParsedSpec>,
pub compute_ctl_config: ComputeCtlConfig,
/// If the spec is passed by a /configure request, 'startup_span' is the
/// /configure request's tracing span. The main thread enters it when it
/// processes the compute startup, so that the compute startup is considered
@@ -160,6 +160,7 @@ impl ComputeState {
last_active: None,
error: None,
pspec: None,
compute_ctl_config: ComputeCtlConfig::default(),
startup_span: None,
metrics: ComputeMetrics::default(),
}
@@ -313,6 +314,7 @@ impl ComputeNode {
let pspec = ParsedSpec::try_from(cli_spec).map_err(|msg| anyhow::anyhow!(msg))?;
new_state.pspec = Some(pspec);
}
new_state.compute_ctl_config = compute_ctl_config;
Ok(ComputeNode {
params,
@@ -321,7 +323,6 @@ impl ComputeNode {
state: Mutex::new(new_state),
state_changed: Condvar::new(),
ext_download_progress: RwLock::new(HashMap::new()),
compute_ctl_config,
})
}
@@ -344,7 +345,7 @@ impl ComputeNode {
// requests while configuration is still in progress.
crate::http::server::Server::External {
port: this.params.external_http_port,
config: this.compute_ctl_config.clone(),
jwks: this.state.lock().unwrap().compute_ctl_config.jwks.clone(),
compute_id: this.params.compute_id.clone(),
}
.launch(&this);
@@ -523,16 +524,6 @@ impl ComputeNode {
// Collect all the tasks that must finish here
let mut pre_tasks = tokio::task::JoinSet::new();
// Make sure TLS certificates are properly loaded and in the right place.
if self.compute_ctl_config.tls.is_some() {
let this = self.clone();
pre_tasks.spawn(async move {
this.watch_cert_for_changes().await;
Ok::<(), anyhow::Error>(())
});
}
// If there are any remote extensions in shared_preload_libraries, start downloading them
if pspec.spec.remote_extensions.is_some() {
let (this, spec) = (self.clone(), pspec.spec.clone());
@@ -588,13 +579,11 @@ impl ComputeNode {
if let Some(pgbouncer_settings) = &pspec.spec.pgbouncer_settings {
info!("tuning pgbouncer");
let pgbouncer_settings = pgbouncer_settings.clone();
let tls_config = self.compute_ctl_config.tls.clone();
// Spawn a background task to do the tuning,
// so that we don't block the main thread that starts Postgres.
let pgbouncer_settings = pgbouncer_settings.clone();
let _handle = tokio::spawn(async move {
let res = tune_pgbouncer(pgbouncer_settings, tls_config).await;
let res = tune_pgbouncer(pgbouncer_settings).await;
if let Err(err) = res {
error!("error while tuning pgbouncer: {err:?}");
// Continue with the startup anyway
@@ -625,11 +614,13 @@ impl ComputeNode {
}
let log_directory_path = Path::new(&self.params.pgdata).join("log");
let log_directory_path = log_directory_path.to_string_lossy().to_string();
configure_audit_rsyslog(log_directory_path.clone(), "hipaa", &remote_endpoint)?;
// Launch a background task to clean up the audit logs
launch_pgaudit_gc(log_directory_path);
// TODO: make this more robust
// now rsyslog starts once and there is no monitoring or restart if it fails
configure_audit_rsyslog(
log_directory_path.to_str().unwrap(),
"hipaa",
&remote_endpoint,
)?;
}
// Launch remaining service threads
@@ -654,9 +645,9 @@ impl ComputeNode {
if pspec.spec.mode == ComputeMode::Primary {
self.configure_as_primary(&compute_state)?;
let conf = self.get_tokio_conn_conf(None);
tokio::task::spawn(async {
let res = get_installed_extensions(conf).await;
let conf = self.get_conn_conf(None);
tokio::task::spawn_blocking(|| {
let res = get_installed_extensions(conf);
match res {
Ok(extensions) => {
info!(
@@ -1114,10 +1105,9 @@ impl ComputeNode {
// Remove/create an empty pgdata directory and put configuration there.
self.create_pgdata()?;
config::write_postgres_conf(
pgdata_path,
&pgdata_path.join("postgresql.conf"),
&pspec.spec,
self.params.internal_http_port,
&self.compute_ctl_config.tls,
)?;
// Syncing safekeepers is only safe with primary nodes: if a primary
@@ -1499,13 +1489,11 @@ impl ComputeNode {
if let Some(ref pgbouncer_settings) = spec.pgbouncer_settings {
info!("tuning pgbouncer");
let pgbouncer_settings = pgbouncer_settings.clone();
let tls_config = self.compute_ctl_config.tls.clone();
// Spawn a background task to do the tuning,
// so that we don't block the main thread that starts Postgres.
let pgbouncer_settings = pgbouncer_settings.clone();
tokio::spawn(async move {
let res = tune_pgbouncer(pgbouncer_settings, tls_config).await;
let res = tune_pgbouncer(pgbouncer_settings).await;
if let Err(err) = res {
error!("error while tuning pgbouncer: {err:?}");
}
@@ -1517,8 +1505,7 @@ impl ComputeNode {
// Spawn a background task to do the configuration,
// so that we don't block the main thread that starts Postgres.
let mut local_proxy = local_proxy.clone();
local_proxy.tls = self.compute_ctl_config.tls.clone();
let local_proxy = local_proxy.clone();
tokio::spawn(async move {
if let Err(err) = local_proxy::configure(&local_proxy) {
error!("error while configuring local_proxy: {err:?}");
@@ -1528,12 +1515,8 @@ impl ComputeNode {
// Write new config
let pgdata_path = Path::new(&self.params.pgdata);
config::write_postgres_conf(
pgdata_path,
&spec,
self.params.internal_http_port,
&self.compute_ctl_config.tls,
)?;
let postgresql_conf_path = pgdata_path.join("postgresql.conf");
config::write_postgres_conf(&postgresql_conf_path, &spec, self.params.internal_http_port)?;
if !spec.skip_pg_catalog_updates {
let max_concurrent_connections = spec.reconfigure_concurrency;
@@ -1604,56 +1587,6 @@ impl ComputeNode {
Ok(())
}
pub async fn watch_cert_for_changes(self: Arc<Self>) {
// update status on cert renewal
if let Some(tls_config) = &self.compute_ctl_config.tls {
let tls_config = tls_config.clone();
// wait until the cert exists.
let mut cert_watch = watch_cert_for_changes(tls_config.cert_path.clone()).await;
tokio::task::spawn_blocking(move || {
let handle = tokio::runtime::Handle::current();
'cert_update: loop {
// let postgres/pgbouncer/local_proxy know the new cert/key exists.
// we need to wait until it's configurable first.
let mut state = self.state.lock().unwrap();
'status_update: loop {
match state.status {
// let's update the state to config pending
ComputeStatus::ConfigurationPending | ComputeStatus::Running => {
state.set_status(
ComputeStatus::ConfigurationPending,
&self.state_changed,
);
break 'status_update;
}
// exit loop
ComputeStatus::Failed
| ComputeStatus::TerminationPending
| ComputeStatus::Terminated => break 'cert_update,
// wait
ComputeStatus::Init
| ComputeStatus::Configuration
| ComputeStatus::Empty => {
state = self.state_changed.wait(state).unwrap();
}
}
}
drop(state);
// wait for a new certificate update
if handle.block_on(cert_watch.changed()).is_err() {
break;
}
}
});
}
}
/// Update the `last_active` in the shared state, but ensure that it's a more recent one.
pub fn update_last_active(&self, last_active: Option<DateTime<Utc>>) {
let mut state = self.state.lock().unwrap();

View File

@@ -6,13 +6,11 @@ use std::io::Write;
use std::io::prelude::*;
use std::path::Path;
use compute_api::responses::TlsConfig;
use compute_api::spec::{ComputeAudit, ComputeMode, ComputeSpec, GenericOption};
use crate::pg_helpers::{
GenericOptionExt, GenericOptionsSearch, PgOptionsSerialize, escape_conf_value,
};
use crate::tls::{self, SERVER_CRT, SERVER_KEY};
/// Check that `line` is inside a text file and put it there if it is not.
/// Create file if it doesn't exist.
@@ -40,12 +38,10 @@ pub fn line_in_file(path: &Path, line: &str) -> Result<bool> {
/// Create or completely rewrite configuration file specified by `path`
pub fn write_postgres_conf(
pgdata_path: &Path,
path: &Path,
spec: &ComputeSpec,
extension_server_port: u16,
tls_config: &Option<TlsConfig>,
) -> Result<()> {
let path = pgdata_path.join("postgresql.conf");
// File::create() destroys the file content if it exists.
let mut file = File::create(path)?;
@@ -90,20 +86,6 @@ pub fn write_postgres_conf(
)?;
}
// tls
if let Some(tls_config) = tls_config {
writeln!(file, "ssl = on")?;
// postgres requires the keyfile to be in a secure file,
// currently too complicated to ensure that at the VM level,
// so we just copy them to another file instead. :shrug:
tls::update_key_path_blocking(pgdata_path, tls_config);
// these are the default, but good to be explicit.
writeln!(file, "ssl_cert_file = '{}'", SERVER_CRT)?;
writeln!(file, "ssl_key_file = '{}'", SERVER_KEY)?;
}
// Locales
if cfg!(target_os = "macos") {
writeln!(file, "lc_messages='C'")?;
@@ -167,8 +149,7 @@ pub fn write_postgres_conf(
writeln!(file, "# Managed by compute_ctl audit settings: begin")?;
// This log level is very verbose
// but this is necessary for HIPAA compliance.
// Exclude 'misc' category, because it doesn't contain anythig relevant.
writeln!(file, "pgaudit.log='all, -misc'")?;
writeln!(file, "pgaudit.log='all'")?;
writeln!(file, "pgaudit.log_parameter=on")?;
// Disable logging of catalog queries
// The catalog doesn't contain sensitive data, so we don't need to audit it.

View File

@@ -4,8 +4,7 @@ module(load="imfile")
# Input configuration for log files in the specified directory
# Replace {log_directory} with the directory containing the log files
input(type="imfile" File="{log_directory}/*.log" Tag="{tag}" Severity="info" Facility="local0")
# the directory to store rsyslog state files
global(workDirectory="/var/log/rsyslog")
global(workDirectory="/var/log")
# Forward logs to remote syslog server
*.* @@{remote_endpoint}

View File

@@ -1,2 +1 @@
pub(in crate::http) mod authorize;
pub(in crate::http) mod request_id;

View File

@@ -1,16 +0,0 @@
use axum::{extract::Request, middleware::Next, response::Response};
use uuid::Uuid;
use crate::http::headers::X_REQUEST_ID;
/// This middleware function allows compute_ctl to generate its own request ID
/// if one isn't supplied. The control plane will always send one as a UUID. The
/// neon Postgres extension on the other hand does not send one.
pub async fn maybe_add_request_id_header(mut request: Request, next: Next) -> Response {
let headers = request.headers_mut();
if !headers.contains_key(X_REQUEST_ID) {
headers.append(X_REQUEST_ID, Uuid::new_v4().to_string().parse().unwrap());
}
next.run(request).await
}

View File

@@ -5,19 +5,20 @@ use std::time::Duration;
use anyhow::Result;
use axum::Router;
use axum::middleware::{self};
use axum::response::IntoResponse;
use axum::extract::Request;
use axum::middleware::{self, Next};
use axum::response::{IntoResponse, Response};
use axum::routing::{get, post};
use compute_api::responses::ComputeCtlConfig;
use http::StatusCode;
use jsonwebtoken::jwk::JwkSet;
use tokio::net::TcpListener;
use tower::ServiceBuilder;
use tower_http::{
auth::AsyncRequireAuthorizationLayer, request_id::PropagateRequestIdLayer, trace::TraceLayer,
};
use tracing::{Span, error, info};
use uuid::Uuid;
use super::middleware::request_id::maybe_add_request_id_header;
use super::{
headers::X_REQUEST_ID,
middleware::authorize::Authorize,
@@ -41,7 +42,7 @@ pub enum Server {
},
External {
port: u16,
config: ComputeCtlConfig,
jwks: JwkSet,
compute_id: String,
},
}
@@ -79,7 +80,7 @@ impl From<&Server> for Router<Arc<ComputeNode>> {
router
}
Server::External {
config, compute_id, ..
jwks, compute_id, ..
} => {
let unauthenticated_router =
Router::<Arc<ComputeNode>>::new().route("/metrics", get(metrics::get_metrics));
@@ -95,7 +96,7 @@ impl From<&Server> for Router<Arc<ComputeNode>> {
.route("/terminate", post(terminate::terminate))
.layer(AsyncRequireAuthorizationLayer::new(Authorize::new(
compute_id.clone(),
config.jwks.clone(),
jwks.clone(),
)));
router
@@ -218,3 +219,15 @@ impl Server {
tokio::spawn(self.serve(state));
}
}
/// This middleware function allows compute_ctl to generate its own request ID
/// if one isn't supplied. The control plane will always send one as a UUID. The
/// neon Postgres extension on the other hand does not send one.
async fn maybe_add_request_id_header(mut request: Request, next: Next) -> Response {
let headers = request.headers_mut();
if headers.get(X_REQUEST_ID).is_none() {
headers.append(X_REQUEST_ID, Uuid::new_v4().to_string().parse().unwrap());
}
next.run(request).await
}

View File

@@ -2,7 +2,7 @@ use std::collections::HashMap;
use anyhow::Result;
use compute_api::responses::{InstalledExtension, InstalledExtensions};
use tokio_postgres::{Client, Config, NoTls};
use postgres::{Client, NoTls};
use crate::metrics::INSTALLED_EXTENSIONS;
@@ -10,7 +10,7 @@ use crate::metrics::INSTALLED_EXTENSIONS;
/// and to make database listing query here more explicit.
///
/// Limit the number of databases to 500 to avoid excessive load.
async fn list_dbs(client: &mut Client) -> Result<Vec<String>> {
fn list_dbs(client: &mut Client) -> Result<Vec<String>> {
// `pg_database.datconnlimit = -2` means that the database is in the
// invalid state
let databases = client
@@ -20,8 +20,7 @@ async fn list_dbs(client: &mut Client) -> Result<Vec<String>> {
AND datconnlimit <> - 2
LIMIT 500",
&[],
)
.await?
)?
.iter()
.map(|row| {
let db: String = row.get("datname");
@@ -37,36 +36,20 @@ async fn list_dbs(client: &mut Client) -> Result<Vec<String>> {
/// Same extension can be installed in multiple databases with different versions,
/// so we report a separate metric (number of databases where it is installed)
/// for each extension version.
pub async fn get_installed_extensions(mut conf: Config) -> Result<InstalledExtensions> {
pub fn get_installed_extensions(mut conf: postgres::config::Config) -> Result<InstalledExtensions> {
conf.application_name("compute_ctl:get_installed_extensions");
let databases: Vec<String> = {
let (mut client, connection) = conf.connect(NoTls).await?;
tokio::spawn(async move {
if let Err(e) = connection.await {
eprintln!("connection error: {}", e);
}
});
list_dbs(&mut client).await?
};
let mut client = conf.connect(NoTls)?;
let databases: Vec<String> = list_dbs(&mut client)?;
let mut extensions_map: HashMap<(String, String, String), InstalledExtension> = HashMap::new();
for db in databases.iter() {
conf.dbname(db);
let (client, connection) = conf.connect(NoTls).await?;
tokio::spawn(async move {
if let Err(e) = connection.await {
eprintln!("connection error: {}", e);
}
});
let extensions: Vec<(String, String, i32)> = client
let mut db_client = conf.connect(NoTls)?;
let extensions: Vec<(String, String, i32)> = db_client
.query(
"SELECT extname, extversion, extowner::integer FROM pg_catalog.pg_extension",
&[],
)
.await?
)?
.iter()
.map(|row| {
(

View File

@@ -26,4 +26,3 @@ pub mod spec;
mod spec_apply;
pub mod swap;
pub mod sync_sk;
pub mod tls;

View File

@@ -24,8 +24,7 @@ pub async fn init_tracing_and_logging(default_log_level: &str) -> anyhow::Result
.with_writer(std::io::stderr);
// Initialize OpenTelemetry
let otlp_layer =
tracing_utils::init_tracing("compute_ctl", tracing_utils::ExportConfig::default()).await;
let otlp_layer = tracing_utils::init_tracing("compute_ctl").await;
// Put it all together
tracing_subscriber::registry()

View File

@@ -1,8 +1,6 @@
use metrics::core::{AtomicF64, Collector, GenericGauge};
use metrics::core::Collector;
use metrics::proto::MetricFamily;
use metrics::{
IntCounterVec, UIntGaugeVec, register_gauge, register_int_counter_vec, register_uint_gauge_vec,
};
use metrics::{IntCounterVec, UIntGaugeVec, register_int_counter_vec, register_uint_gauge_vec};
use once_cell::sync::Lazy;
pub(crate) static INSTALLED_EXTENSIONS: Lazy<UIntGaugeVec> = Lazy::new(|| {
@@ -61,20 +59,10 @@ pub(crate) static REMOTE_EXT_REQUESTS_TOTAL: Lazy<IntCounterVec> = Lazy::new(||
.expect("failed to define a metric")
});
// Size of audit log directory in bytes
pub(crate) static AUDIT_LOG_DIR_SIZE: Lazy<GenericGauge<AtomicF64>> = Lazy::new(|| {
register_gauge!(
"compute_audit_log_dir_size",
"Size of audit log directory in bytes",
)
.expect("failed to define a metric")
});
pub fn collect() -> Vec<MetricFamily> {
let mut metrics = INSTALLED_EXTENSIONS.collect();
metrics.extend(CPLANE_REQUESTS_TOTAL.collect());
metrics.extend(REMOTE_EXT_REQUESTS_TOTAL.collect());
metrics.extend(DB_MIGRATION_FAILED.collect());
metrics.extend(AUDIT_LOG_DIR_SIZE.collect());
metrics
}

View File

@@ -10,10 +10,8 @@ use std::str::FromStr;
use std::time::{Duration, Instant};
use anyhow::{Result, bail};
use compute_api::responses::TlsConfig;
use compute_api::spec::{Database, GenericOption, GenericOptions, PgIdent, Role};
use futures::StreamExt;
use indexmap::IndexMap;
use ini::Ini;
use notify::{RecursiveMode, Watcher};
use postgres::config::Config;
@@ -208,8 +206,8 @@ impl Escaping for PgIdent {
/// Here we somewhat mimic the logic of Postgres' `pg_get_functiondef()`,
/// <https://github.com/postgres/postgres/blob/8b49392b270b4ac0b9f5c210e2a503546841e832/src/backend/utils/adt/ruleutils.c#L2924>
fn pg_quote_dollar(&self) -> (String, String) {
let mut tag: String = "x".to_string();
let mut outer_tag = "xx".to_string();
let mut tag: String = "".to_string();
let mut outer_tag = "x".to_string();
// Find the first suitable tag that is not present in the string.
// Postgres' max role/DB name length is 63 bytes, so even in the
@@ -408,7 +406,7 @@ pub fn create_pgdata(pgdata: &str) -> Result<()> {
/// Update pgbouncer.ini with provided options
fn update_pgbouncer_ini(
pgbouncer_config: IndexMap<String, String>,
pgbouncer_config: HashMap<String, String>,
pgbouncer_ini_path: &str,
) -> Result<()> {
let mut conf = Ini::load_from_file(pgbouncer_ini_path)?;
@@ -429,10 +427,7 @@ fn update_pgbouncer_ini(
/// Tune pgbouncer.
/// 1. Apply new config using pgbouncer admin console
/// 2. Add new values to pgbouncer.ini to preserve them after restart
pub async fn tune_pgbouncer(
mut pgbouncer_config: IndexMap<String, String>,
tls_config: Option<TlsConfig>,
) -> Result<()> {
pub async fn tune_pgbouncer(pgbouncer_config: HashMap<String, String>) -> Result<()> {
let pgbouncer_connstr = if std::env::var_os("AUTOSCALING").is_some() {
// for VMs use pgbouncer specific way to connect to
// pgbouncer admin console without password
@@ -478,21 +473,19 @@ pub async fn tune_pgbouncer(
}
};
if let Some(tls_config) = tls_config {
// pgbouncer starts in a half-ok state if it cannot find these files.
// It will default to client_tls_sslmode=deny, which causes proxy to error.
// There is a small window at startup where these files don't yet exist in the VM.
// Best to wait until it exists.
loop {
if let Ok(true) = tokio::fs::try_exists(&tls_config.key_path).await {
break;
}
tokio::time::sleep(Duration::from_millis(500)).await
}
// Apply new config
for (option_name, value) in pgbouncer_config.iter() {
let query = format!("SET {}={}", option_name, value);
// keep this log line for debugging purposes
info!("Applying pgbouncer setting change: {}", query);
pgbouncer_config.insert("client_tls_cert_file".to_string(), tls_config.cert_path);
pgbouncer_config.insert("client_tls_key_file".to_string(), tls_config.key_path);
pgbouncer_config.insert("client_tls_sslmode".to_string(), "allow".to_string());
if let Err(err) = client.simple_query(&query).await {
// Don't fail on error, just print it into log
error!(
"Failed to apply pgbouncer setting change: {}, {}",
query, err
);
};
}
// save values to pgbouncer.ini
@@ -508,13 +501,6 @@ pub async fn tune_pgbouncer(
};
update_pgbouncer_ini(pgbouncer_config, &pgbouncer_ini_path)?;
info!("Applying pgbouncer setting change");
if let Err(err) = client.simple_query("RELOAD").await {
// Don't fail on error, just print it into log
error!("Failed to apply pgbouncer setting change, {err}",);
};
Ok(())
}

View File

@@ -1,11 +1,8 @@
use std::fs;
use std::path::Path;
use std::process::Command;
use std::time::Duration;
use std::{fs::OpenOptions, io::Write};
use anyhow::{Context, Result};
use tracing::{error, info, instrument, warn};
use tracing::info;
fn get_rsyslog_pid() -> Option<String> {
let output = Command::new("pgrep")
@@ -46,7 +43,7 @@ fn restart_rsyslog() -> Result<()> {
}
pub fn configure_audit_rsyslog(
log_directory: String,
log_directory: &str,
tag: &str,
remote_endpoint: &str,
) -> Result<()> {
@@ -78,61 +75,3 @@ pub fn configure_audit_rsyslog(
Ok(())
}
#[instrument(skip_all)]
async fn pgaudit_gc_main_loop(log_directory: String) -> Result<()> {
info!("running pgaudit GC main loop");
loop {
// Check log_directory for old pgaudit logs and delete them.
// New log files are checked every 5 minutes, as set in pgaudit.log_rotation_age
// Find files that were not modified in the last 15 minutes and delete them.
// This should be enough time for rsyslog to process the logs and for us to catch the alerts.
//
// In case of a very high load, we might need to adjust this value and pgaudit.log_rotation_age.
//
// TODO: add some smarter logic to delete the files that are fully streamed according to rsyslog
// imfile-state files, but for now just do a simple GC to avoid filling up the disk.
let _ = Command::new("find")
.arg(&log_directory)
.arg("-name")
.arg("audit*.log")
.arg("-mmin")
.arg("+15")
.arg("-delete")
.output()?;
// also collect the metric for the size of the log directory
async fn get_log_files_size(path: &Path) -> Result<u64> {
let mut total_size = 0;
for entry in fs::read_dir(path)? {
let entry = entry?;
let entry_path = entry.path();
if entry_path.is_file() && entry_path.to_string_lossy().ends_with("log") {
total_size += entry.metadata()?.len();
}
}
Ok(total_size)
}
let log_directory_size = get_log_files_size(Path::new(&log_directory))
.await
.unwrap_or_else(|e| {
warn!("Failed to get log directory size: {}", e);
0
});
crate::metrics::AUDIT_LOG_DIR_SIZE.set(log_directory_size as f64);
tokio::time::sleep(Duration::from_secs(60)).await;
}
}
// launch pgaudit GC thread to clean up the old pgaudit logs stored in the log_directory
pub fn launch_pgaudit_gc(log_directory: String) {
tokio::spawn(async move {
if let Err(e) = pgaudit_gc_main_loop(log_directory).await {
error!("pgaudit GC main loop failed: {}", e);
}
});
}

View File

@@ -1,118 +0,0 @@
use std::{io::Write, os::unix::fs::OpenOptionsExt, path::Path, time::Duration};
use anyhow::{Context, Result, bail};
use compute_api::responses::TlsConfig;
use ring::digest;
use spki::ObjectIdentifier;
use spki::der::{Decode, PemReader};
use x509_cert::Certificate;
#[derive(Clone, Copy)]
pub struct CertDigest(digest::Digest);
pub async fn watch_cert_for_changes(cert_path: String) -> tokio::sync::watch::Receiver<CertDigest> {
let mut digest = compute_digest(&cert_path).await;
let (tx, rx) = tokio::sync::watch::channel(digest);
tokio::spawn(async move {
while !tx.is_closed() {
let new_digest = compute_digest(&cert_path).await;
if digest.0.as_ref() != new_digest.0.as_ref() {
digest = new_digest;
_ = tx.send(digest);
}
tokio::time::sleep(Duration::from_secs(60)).await
}
});
rx
}
async fn compute_digest(cert_path: &str) -> CertDigest {
loop {
match try_compute_digest(cert_path).await {
Ok(d) => break d,
Err(e) => {
tracing::error!("could not read cert file {e:?}");
tokio::time::sleep(Duration::from_secs(1)).await
}
}
}
}
async fn try_compute_digest(cert_path: &str) -> Result<CertDigest> {
let data = tokio::fs::read(cert_path).await?;
// sha256 is extremely collision resistent. can safely assume the digest to be unique
Ok(CertDigest(digest::digest(&digest::SHA256, &data)))
}
pub const SERVER_CRT: &str = "server.crt";
pub const SERVER_KEY: &str = "server.key";
pub fn update_key_path_blocking(pg_data: &Path, tls_config: &TlsConfig) {
loop {
match try_update_key_path_blocking(pg_data, tls_config) {
Ok(()) => break,
Err(e) => {
tracing::error!("could not create key file {e:?}");
std::thread::sleep(Duration::from_secs(1))
}
}
}
}
// Postgres requires the keypath be "secure". This means
// 1. Owned by the postgres user.
// 2. Have permission 600.
fn try_update_key_path_blocking(pg_data: &Path, tls_config: &TlsConfig) -> Result<()> {
let key = std::fs::read_to_string(&tls_config.key_path)?;
let crt = std::fs::read_to_string(&tls_config.cert_path)?;
// to mitigate a race condition during renewal.
verify_key_cert(&key, &crt)?;
let mut key_file = std::fs::OpenOptions::new()
.write(true)
.create(true)
.truncate(true)
.mode(0o600)
.open(pg_data.join(SERVER_KEY))?;
let mut crt_file = std::fs::OpenOptions::new()
.write(true)
.create(true)
.truncate(true)
.mode(0o600)
.open(pg_data.join(SERVER_CRT))?;
key_file.write_all(key.as_bytes())?;
crt_file.write_all(crt.as_bytes())?;
Ok(())
}
fn verify_key_cert(key: &str, cert: &str) -> Result<()> {
const ECDSA_WITH_SHA256: ObjectIdentifier = ObjectIdentifier::new_unwrap("1.2.840.10045.4.3.2");
let cert = Certificate::decode(&mut PemReader::new(cert.as_bytes()).context("pem reader")?)
.context("decode cert")?;
match cert.signature_algorithm.oid {
ECDSA_WITH_SHA256 => {
let key = p256::SecretKey::from_sec1_pem(key).context("parse key")?;
let a = key.public_key().to_sec1_bytes();
let b = cert
.tbs_certificate
.subject_public_key_info
.subject_public_key
.raw_bytes();
if *a != *b {
bail!("private key file does not match certificate")
}
}
_ => bail!("unknown TLS key type"),
}
Ok(())
}

View File

@@ -64,8 +64,7 @@ test.escaping = 'here''s a backslash \\ and a quote '' and a double-quote " hoor
#[test]
fn ident_pg_quote_dollar() {
let test_cases = vec![
("name", ("$x$name$x$", "xx")),
("name$", ("$x$name$$x$", "xx")),
("name", ("$$name$$", "x")),
("name$$", ("$x$name$$$x$", "xx")),
("name$$$", ("$x$name$$$$x$", "xx")),
("name$$$$", ("$x$name$$$$$x$", "xx")),

View File

@@ -36,9 +36,7 @@ use pageserver_api::config::{
use pageserver_api::controller_api::{
NodeAvailabilityWrapper, PlacementPolicy, TenantCreateRequest,
};
use pageserver_api::models::{
ShardParameters, TenantConfigRequest, TimelineCreateRequest, TimelineInfo,
};
use pageserver_api::models::{ShardParameters, TimelineCreateRequest, TimelineInfo};
use pageserver_api::shard::{ShardCount, ShardStripeSize, TenantShardId};
use postgres_backend::AuthType;
use postgres_connection::parse_host_port;
@@ -979,7 +977,7 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
neon_distrib_dir: None,
default_tenant_id: TenantId::from_array(std::array::from_fn(|_| 0)),
storage_controller: None,
control_plane_hooks_api: None,
control_plane_compute_hook_api: None,
generate_local_ssl_certs: false,
}
};
@@ -1131,16 +1129,12 @@ async fn handle_tenant(subcmd: &TenantCmd, env: &mut local_env::LocalEnv) -> any
let tenant_id = get_tenant_id(args.tenant_id, env)?;
let tenant_conf: HashMap<_, _> =
args.config.iter().flat_map(|c| c.split_once(':')).collect();
let config = PageServerNode::parse_config(tenant_conf)?;
let req = TenantConfigRequest { tenant_id, config };
let storage_controller = StorageController::from_env(env);
storage_controller
.set_tenant_config(&req)
pageserver
.tenant_config(tenant_id, tenant_conf)
.await
.with_context(|| format!("Tenant config failed for tenant with id {tenant_id}"))?;
println!("tenant {tenant_id} successfully configured via storcon");
println!("tenant {tenant_id} successfully configured on the pageserver");
}
}
Ok(())

View File

@@ -72,9 +72,9 @@ pub struct LocalEnv {
// be propagated into each pageserver's configuration.
pub control_plane_api: Url,
// Control plane upcall APIs for storage controller. If set, this will be propagated into the
// Control plane upcall API for storage controller. If set, this will be propagated into the
// storage controller's configuration.
pub control_plane_hooks_api: Option<Url>,
pub control_plane_compute_hook_api: Option<Url>,
/// Keep human-readable aliases in memory (and persist them to config), to hide ZId hex strings from the user.
// A `HashMap<String, HashMap<TenantId, TimelineId>>` would be more appropriate here,
@@ -104,7 +104,6 @@ pub struct OnDiskConfig {
pub pageservers: Vec<PageServerConf>,
pub safekeepers: Vec<SafekeeperConf>,
pub control_plane_api: Option<Url>,
pub control_plane_hooks_api: Option<Url>,
pub control_plane_compute_hook_api: Option<Url>,
branch_name_mappings: HashMap<String, Vec<(TenantId, TimelineId)>>,
// Note: skip serializing because in compat tests old storage controller fails
@@ -137,7 +136,7 @@ pub struct NeonLocalInitConf {
pub pageservers: Vec<NeonLocalInitPageserverConf>,
pub safekeepers: Vec<SafekeeperConf>,
pub control_plane_api: Option<Url>,
pub control_plane_hooks_api: Option<Url>,
pub control_plane_compute_hook_api: Option<Option<Url>>,
pub generate_local_ssl_certs: bool,
}
@@ -149,7 +148,7 @@ pub struct NeonBroker {
pub listen_addr: SocketAddr,
}
/// A part of storage controller's config the neon_local knows about.
/// Broker config for cluster internal communication.
#[derive(Serialize, Deserialize, PartialEq, Eq, Clone, Debug)]
#[serde(default)]
pub struct NeonStorageControllerConf {
@@ -176,11 +175,8 @@ pub struct NeonStorageControllerConf {
#[serde(with = "humantime_serde")]
pub long_reconcile_threshold: Option<Duration>,
#[serde(default)]
pub use_https_pageserver_api: bool,
pub timelines_onto_safekeepers: bool,
pub use_https_safekeeper_api: bool,
}
impl NeonStorageControllerConf {
@@ -205,8 +201,6 @@ impl Default for NeonStorageControllerConf {
heartbeat_interval: Self::DEFAULT_HEARTBEAT_INTERVAL,
long_reconcile_threshold: None,
use_https_pageserver_api: false,
timelines_onto_safekeepers: false,
use_https_safekeeper_api: false,
}
}
}
@@ -304,7 +298,6 @@ pub struct SafekeeperConf {
pub pg_port: u16,
pub pg_tenant_only_port: Option<u16>,
pub http_port: u16,
pub https_port: Option<u16>,
pub sync: bool,
pub remote_storage: Option<String>,
pub backup_threads: Option<u32>,
@@ -319,7 +312,6 @@ impl Default for SafekeeperConf {
pg_port: 0,
pg_tenant_only_port: None,
http_port: 0,
https_port: None,
sync: true,
remote_storage: None,
backup_threads: None,
@@ -578,8 +570,7 @@ impl LocalEnv {
pageservers,
safekeepers,
control_plane_api,
control_plane_hooks_api,
control_plane_compute_hook_api: _,
control_plane_compute_hook_api,
branch_name_mappings,
generate_local_ssl_certs,
} = on_disk_config;
@@ -594,7 +585,7 @@ impl LocalEnv {
pageservers,
safekeepers,
control_plane_api: control_plane_api.unwrap(),
control_plane_hooks_api,
control_plane_compute_hook_api,
branch_name_mappings,
generate_local_ssl_certs,
}
@@ -701,8 +692,7 @@ impl LocalEnv {
pageservers: vec![], // it's skip_serializing anyway
safekeepers: self.safekeepers.clone(),
control_plane_api: Some(self.control_plane_api.clone()),
control_plane_hooks_api: self.control_plane_hooks_api.clone(),
control_plane_compute_hook_api: None,
control_plane_compute_hook_api: self.control_plane_compute_hook_api.clone(),
branch_name_mappings: self.branch_name_mappings.clone(),
generate_local_ssl_certs: self.generate_local_ssl_certs,
},
@@ -786,8 +776,8 @@ impl LocalEnv {
pageservers,
safekeepers,
control_plane_api,
control_plane_compute_hook_api,
generate_local_ssl_certs,
control_plane_hooks_api,
} = conf;
// Find postgres binaries.
@@ -834,7 +824,7 @@ impl LocalEnv {
pageservers: pageservers.iter().map(Into::into).collect(),
safekeepers,
control_plane_api: control_plane_api.unwrap(),
control_plane_hooks_api,
control_plane_compute_hook_api: control_plane_compute_hook_api.unwrap_or_default(),
branch_name_mappings: Default::default(),
generate_local_ssl_certs,
};
@@ -849,9 +839,6 @@ impl LocalEnv {
// create safekeeper dirs
for safekeeper in &env.safekeepers {
fs::create_dir_all(SafekeeperNode::datadir_path_by_id(&env, safekeeper.id))?;
SafekeeperNode::from_env(&env, safekeeper)
.initialize()
.context("safekeeper init failed")?;
}
// initialize pageserver state

View File

@@ -111,18 +111,6 @@ impl SafekeeperNode {
.expect("non-Unicode path")
}
/// Initializes a safekeeper node by creating all necessary files,
/// e.g. SSL certificates.
pub fn initialize(&self) -> anyhow::Result<()> {
if self.env.generate_local_ssl_certs {
self.env.generate_ssl_cert(
&self.datadir_path().join("server.crt"),
&self.datadir_path().join("server.key"),
)?;
}
Ok(())
}
pub async fn start(
&self,
extra_opts: &[String],
@@ -208,16 +196,6 @@ impl SafekeeperNode {
]);
}
if let Some(https_port) = self.conf.https_port {
args.extend([
"--listen-https".to_owned(),
format!("{}:{}", self.listen_addr, https_port),
]);
}
if let Some(ssl_ca_file) = self.env.ssl_ca_cert_path() {
args.push(format!("--ssl-ca-file={}", ssl_ca_file.to_str().unwrap()));
}
args.extend_from_slice(extra_opts);
background_process::start_process(

View File

@@ -14,7 +14,7 @@ use pageserver_api::controller_api::{
NodeConfigureRequest, NodeDescribeResponse, NodeRegisterRequest, TenantCreateRequest,
TenantCreateResponse, TenantLocateResponse,
};
use pageserver_api::models::{TenantConfigRequest, TimelineCreateRequest, TimelineInfo};
use pageserver_api::models::{TimelineCreateRequest, TimelineInfo};
use pageserver_api::shard::TenantShardId;
use pageserver_client::mgmt_api::ResponseErrorMessageExt;
use postgres_backend::AuthType;
@@ -538,10 +538,6 @@ impl StorageController {
args.push("--use-https-pageserver-api".to_string());
}
if self.config.use_https_safekeeper_api {
args.push("--use-https-safekeeper-api".to_string());
}
if let Some(ssl_ca_file) = self.env.ssl_ca_cert_path() {
args.push(format!("--ssl-ca-file={}", ssl_ca_file.to_str().unwrap()));
}
@@ -562,8 +558,10 @@ impl StorageController {
args.push(format!("--public-key=\"{public_key}\""));
}
if let Some(control_plane_hooks_api) = &self.env.control_plane_hooks_api {
args.push(format!("--control-plane-url={control_plane_hooks_api}"));
if let Some(control_plane_compute_hook_api) = &self.env.control_plane_compute_hook_api {
args.push(format!(
"--compute-hook-url={control_plane_compute_hook_api}"
));
}
if let Some(split_threshold) = self.config.split_threshold.as_ref() {
@@ -586,10 +584,6 @@ impl StorageController {
self.env.base_data_dir.display()
));
if self.config.timelines_onto_safekeepers {
args.push("--timelines-onto-safekeepers".to_string());
}
background_process::start_process(
COMMAND,
&instance_dir,
@@ -880,9 +874,4 @@ impl StorageController {
)
.await
}
pub async fn set_tenant_config(&self, req: &TenantConfigRequest) -> anyhow::Result<()> {
self.dispatch(Method::PUT, "v1/tenant/config".to_string(), Some(req))
.await
}
}

View File

@@ -14,8 +14,8 @@ use pageserver_api::controller_api::{
TenantShardMigrateRequest, TenantShardMigrateResponse,
};
use pageserver_api::models::{
EvictionPolicy, EvictionPolicyLayerAccessThreshold, ShardParameters, TenantConfig,
TenantConfigPatchRequest, TenantConfigRequest, TenantShardSplitRequest,
EvictionPolicy, EvictionPolicyLayerAccessThreshold, LocationConfigSecondary, ShardParameters,
TenantConfig, TenantConfigPatchRequest, TenantConfigRequest, TenantShardSplitRequest,
TenantShardSplitResponse,
};
use pageserver_api::shard::{ShardStripeSize, TenantShardId};
@@ -158,6 +158,12 @@ enum Command {
#[arg(long)]
tenant_id: TenantId,
},
/// For a tenant which hasn't been onboarded to the storage controller yet, add it in secondary
/// mode so that it can warm up content on a pageserver.
TenantWarmup {
#[arg(long)]
tenant_id: TenantId,
},
TenantSetPreferredAz {
#[arg(long)]
tenant_id: TenantId,
@@ -865,6 +871,94 @@ async fn main() -> anyhow::Result<()> {
)
.await?;
}
Command::TenantWarmup { tenant_id } => {
let describe_response = storcon_client
.dispatch::<(), TenantDescribeResponse>(
Method::GET,
format!("control/v1/tenant/{tenant_id}"),
None,
)
.await;
match describe_response {
Ok(describe) => {
if matches!(describe.policy, PlacementPolicy::Secondary) {
// Fine: it's already known to controller in secondary mode: calling
// again to put it into secondary mode won't cause problems.
} else {
anyhow::bail!("Tenant already present with policy {:?}", describe.policy);
}
}
Err(mgmt_api::Error::ApiError(StatusCode::NOT_FOUND, _)) => {
// Fine: this tenant isn't know to the storage controller yet.
}
Err(e) => {
// Unexpected API error
return Err(e.into());
}
}
vps_client
.location_config(
TenantShardId::unsharded(tenant_id),
pageserver_api::models::LocationConfig {
mode: pageserver_api::models::LocationConfigMode::Secondary,
generation: None,
secondary_conf: Some(LocationConfigSecondary { warm: true }),
shard_number: 0,
shard_count: 0,
shard_stripe_size: ShardParameters::DEFAULT_STRIPE_SIZE.0,
tenant_conf: TenantConfig::default(),
},
None,
true,
)
.await?;
let describe_response = storcon_client
.dispatch::<(), TenantDescribeResponse>(
Method::GET,
format!("control/v1/tenant/{tenant_id}"),
None,
)
.await?;
let secondary_ps_id = describe_response
.shards
.first()
.unwrap()
.node_secondary
.first()
.unwrap();
println!("Tenant {tenant_id} warming up on pageserver {secondary_ps_id}");
loop {
let (status, progress) = vps_client
.tenant_secondary_download(
TenantShardId::unsharded(tenant_id),
Some(Duration::from_secs(10)),
)
.await?;
println!(
"Progress: {}/{} layers, {}/{} bytes",
progress.layers_downloaded,
progress.layers_total,
progress.bytes_downloaded,
progress.bytes_total
);
match status {
StatusCode::OK => {
println!("Download complete");
break;
}
StatusCode::ACCEPTED => {
// Loop
}
_ => {
anyhow::bail!("Unexpected download status: {status}");
}
}
}
}
Command::TenantDrop { tenant_id, unclean } => {
if !unclean {
anyhow::bail!(

View File

@@ -1,7 +1,3 @@
# Neon RFCs
## Overview
This directory contains Request for Comments documents, or RFCs, for
features or concepts that have been proposed. Alternative names:
technical design doc, ERD, one-pager
@@ -63,10 +59,37 @@ RFC lifecycle:
### RFC template
Use template with `YYYY-MM-DD-copy-me.md` as a starting point. Timestamp prefix helps to avoid awkward 'id' collisions.
```sh
cp docs/rfcs/YYYY-MM-DD-copy-me.md docs/rfcs/$(date +"%Y-%m-%d")-<name>.md
```
Note, a lot of the sections are marked as if relevant. They are included into the template as a reminder and to help inspiration.
```
# Name
Created on ..
Implemented on ..
## Summary
## Motivation
## Non Goals (if relevant)
## Impacted components (e.g. pageserver, safekeeper, console, etc)
## Proposed implementation
### Reliability, failure modes and corner cases (if relevant)
### Interaction/Sequence diagram (if relevant)
### Scalability (if relevant)
### Security implications (if relevant)
### Unresolved questions (if relevant)
## Alternative implementation (if relevant)
## Pros/cons of proposed approaches (if relevant)
## Definition of Done (if relevant)
```

View File

@@ -1,30 +0,0 @@
# Name
Created on YYYY-MM-DD
Implemented on _TBD_
## Summary
## Motivation
## Non Goals (if relevant)
## Impacted components (e.g. pageserver, safekeeper, console, etc)
## Proposed implementation
### Reliability, failure modes and corner cases (if relevant)
### Interaction/Sequence diagram (if relevant)
### Scalability (if relevant)
### Security implications (if relevant)
### Unresolved questions (if relevant)
## Alternative implementation (if relevant)
## Pros/cons of proposed approaches (if relevant)
## Definition of Done (if relevant)

View File

@@ -101,25 +101,15 @@ changes such as a pageserver node becoming unavailable, or the tenant's shard co
postgres clients to handle such changes, the storage controller calls an API hook when a tenant's pageserver
location changes.
The hook is configured using the storage controller's `--control-plane-url` CLI option, from which the hook URL is computed.
The hook is configured using the storage controller's `--compute-hook-url` CLI option. If the hook requires
JWT auth, the token may be provided with `--control-plane-jwt-token`. The hook will be invoked with a `PUT` request.
Currently, there is two hooks, each computed by appending the name to the provided control plane URL prefix:
- `notify-attach`, called whenever attachment for pageservers changes
- `notify-safekeepers`, called whenever attachment for safekeepers changes
If the hooks require JWT auth, the token may be provided with `--control-plane-jwt-token`.
The hooks will be invoked with a `PUT` request.
In the Neon cloud service, these hooks are implemented by Neon's internal cloud control plane. In `neon_local` systems,
In the Neon cloud service, this hook is implemented by Neon's internal cloud control plane. In `neon_local` systems
the storage controller integrates directly with neon_local to reconfigure local postgres processes instead of calling
the compute hook.
When implementing an on-premise Neon deployment, you must implement a service that handles the compute hooks. This is not complicated.
### `notify-attach` body
The `notify-attach` request body follows the format of the `ComputeHookNotifyRequest` structure, provided below for convenience.
When implementing an on-premise Neon deployment, you must implement a service that handles the compute hook. This is not complicated:
the request body has format of the `ComputeHookNotifyRequest` structure, provided below for convenience.
```
struct ComputeHookNotifyRequestShard {
@@ -138,15 +128,15 @@ When a notification is received:
1. Modify postgres configuration for this tenant:
- set `neon.pageserver_connstring` to a comma-separated list of postgres connection strings to pageservers according to the `shards` list. The
- set `neon.pageserver_connstr` to a comma-separated list of postgres connection strings to pageservers according to the `shards` list. The
shards identified by `NodeId` must be converted to the address+port of the node.
- if stripe_size is not None, set `neon.shard_stripe_size` to this value
- if stripe_size is not None, set `neon.stripe_size` to this value
2. Send SIGHUP to postgres to reload configuration
3. Respond with 200 to the notification request. Do not return success if postgres was not updated: if an error is returned, the controller
will retry the notification until it succeeds..
Example body:
### Example notification body
```
{
@@ -158,34 +148,3 @@ Example body:
],
}
```
### `notify-safekeepers` body
The `notify-safekeepers` request body forllows the format of the `SafekeepersNotifyRequest` structure, provided below for convenience.
```
pub struct SafekeeperInfo {
pub id: NodeId,
pub hostname: String,
}
pub struct SafekeepersNotifyRequest {
pub tenant_id: TenantId,
pub timeline_id: TimelineId,
pub generation: u32,
pub safekeepers: Vec<SafekeeperInfo>,
}
```
When a notification is received:
1. Modify postgres configuration for this tenant:
- set `neon.safekeeper_connstrings` to an array of postgres connection strings to safekeepers according to the `safekeepers` list. The
safekeepers identified by `NodeId` must be converted to the address+port of the respective safekeeper.
The hostname is provided for debugging purposes, so we reserve changes to how we pass it.
- set `neon.safekeepers_generation` to the provided `generation` value.
2. Send SIGHUP to postgres to reload configuration
3. Respond with 200 to the notification request. Do not return success if postgres was not updated: if an error is returned, the controller
will retry the notification until it succeeds..

View File

@@ -7,7 +7,6 @@ license.workspace = true
[dependencies]
anyhow.workspace = true
chrono.workspace = true
indexmap.workspace = true
jsonwebtoken.workspace = true
serde.workspace = true
serde_json.workspace = true

View File

@@ -139,7 +139,6 @@ pub struct ComputeCtlConfig {
/// Set of JSON web keys that the compute can use to authenticate
/// communication from the control plane.
pub jwks: JwkSet,
pub tls: Option<TlsConfig>,
}
impl Default for ComputeCtlConfig {
@@ -148,17 +147,10 @@ impl Default for ComputeCtlConfig {
jwks: JwkSet {
keys: Vec::default(),
},
tls: None,
}
}
}
#[derive(Clone, Debug, Deserialize, Serialize)]
pub struct TlsConfig {
pub key_path: String,
pub cert_path: String,
}
/// Response of the `/computes/{compute_id}/spec` control-plane API.
#[derive(Deserialize, Debug)]
pub struct ControlPlaneSpecResponse {

View File

@@ -5,15 +5,12 @@
//! and connect it to the storage nodes.
use std::collections::HashMap;
use indexmap::IndexMap;
use regex::Regex;
use remote_storage::RemotePath;
use serde::{Deserialize, Serialize};
use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
use crate::responses::TlsConfig;
/// String type alias representing Postgres identifier and
/// intended to be used for DB / role names.
pub type PgIdent = String;
@@ -128,7 +125,7 @@ pub struct ComputeSpec {
// information about available remote extensions
pub remote_extensions: Option<RemoteExtSpec>,
pub pgbouncer_settings: Option<IndexMap<String, String>>,
pub pgbouncer_settings: Option<HashMap<String, String>>,
// Stripe size for pageserver sharding, in pages
#[serde(default)]
@@ -360,9 +357,6 @@ pub struct LocalProxySpec {
#[serde(default)]
#[serde(skip_serializing_if = "Option::is_none")]
pub jwks: Option<Vec<JwksSettings>>,
#[serde(default)]
#[serde(skip_serializing_if = "Option::is_none")]
pub tls: Option<TlsConfig>,
}
#[derive(Clone, Debug, Deserialize, Serialize)]

View File

@@ -7,7 +7,6 @@ license.workspace = true
[dependencies]
anyhow.workspace = true
bytes.workspace = true
camino.workspace = true
fail.workspace = true
futures.workspace = true
hyper0.workspace = true
@@ -17,7 +16,6 @@ once_cell.workspace = true
pprof.workspace = true
regex.workspace = true
routerify.workspace = true
rustls-pemfile.workspace = true
serde.workspace = true
serde_json.workspace = true
serde_path_to_error.workspace = true

View File

@@ -4,7 +4,6 @@ pub mod failpoints;
pub mod json;
pub mod request;
pub mod server;
pub mod tls_certs;
extern crate hyper0 as hyper;

View File

@@ -1,21 +0,0 @@
use camino::Utf8Path;
use tokio_rustls::rustls::pki_types::{CertificateDer, PrivateKeyDer};
pub fn load_cert_chain(filename: &Utf8Path) -> anyhow::Result<Vec<CertificateDer<'static>>> {
let file = std::fs::File::open(filename)?;
let mut reader = std::io::BufReader::new(file);
Ok(rustls_pemfile::certs(&mut reader).collect::<Result<Vec<_>, _>>()?)
}
pub fn load_private_key(filename: &Utf8Path) -> anyhow::Result<PrivateKeyDer<'static>> {
let file = std::fs::File::open(filename)?;
let mut reader = std::io::BufReader::new(file);
let key = rustls_pemfile::private_key(&mut reader)?;
key.ok_or(anyhow::anyhow!(
"no private key found in {}",
filename.as_str(),
))
}

View File

@@ -272,16 +272,15 @@ pub struct TenantConfigToml {
/// size exceeds `compaction_upper_limit * checkpoint_distance`.
pub compaction_upper_limit: usize,
pub compaction_algorithm: crate::models::CompactionAlgorithmSettings,
/// If true, compact down L0 across all tenant timelines before doing regular compaction. L0
/// compaction must be responsive to avoid read amp during heavy ingestion. Defaults to true.
/// If true, compact down L0 across all tenant timelines before doing regular compaction.
pub compaction_l0_first: bool,
/// If true, use a separate semaphore (i.e. concurrency limit) for the L0 compaction pass. Only
/// has an effect if `compaction_l0_first` is true. Defaults to true.
/// has an effect if `compaction_l0_first` is `true`.
pub compaction_l0_semaphore: bool,
/// Level0 delta layer threshold at which to delay layer flushes such that they take 2x as long,
/// and block on layer flushes during ephemeral layer rolls, for compaction backpressure. This
/// helps compaction keep up with WAL ingestion, and avoids read amplification blowing up.
/// Should be >compaction_threshold. 0 to disable. Defaults to 3x compaction_threshold.
/// Level0 delta layer threshold at which to delay layer flushes for compaction backpressure,
/// such that they take 2x as long, and start waiting for layer flushes during ephemeral layer
/// rolls. This helps compaction keep up with WAL ingestion, and avoids read amplification
/// blowing up. Should be >compaction_threshold. 0 to disable. Disabled by default.
pub l0_flush_delay_threshold: Option<usize>,
/// Level0 delta layer threshold at which to stall layer flushes. Must be >compaction_threshold
/// to avoid deadlock. 0 to disable. Disabled by default.
@@ -289,8 +288,6 @@ pub struct TenantConfigToml {
/// If true, Level0 delta layer flushes will wait for S3 upload before flushing the next
/// layer. This is a temporary backpressure mechanism which should be removed once
/// l0_flush_{delay,stall}_threshold is fully enabled.
///
/// TODO: this is no longer enabled, remove it when the config option is no longer set.
pub l0_flush_wait_upload: bool,
// Determines how much history is retained, to allow
// branching and read replicas at an older point in time.
@@ -570,15 +567,13 @@ pub mod tenant_conf_defaults {
// be reduced later by optimizing L0 hole calculation to avoid loading all keys into memory). So
// with this config, we can get a maximum peak compaction usage of 9 GB.
pub const DEFAULT_COMPACTION_UPPER_LIMIT: usize = 20;
// Enable L0 compaction pass and semaphore by default. L0 compaction must be responsive to avoid
// read amp.
pub const DEFAULT_COMPACTION_L0_FIRST: bool = true;
pub const DEFAULT_COMPACTION_L0_FIRST: bool = false;
pub const DEFAULT_COMPACTION_L0_SEMAPHORE: bool = true;
pub const DEFAULT_COMPACTION_ALGORITHM: crate::models::CompactionAlgorithm =
crate::models::CompactionAlgorithm::Legacy;
pub const DEFAULT_L0_FLUSH_WAIT_UPLOAD: bool = false;
pub const DEFAULT_L0_FLUSH_WAIT_UPLOAD: bool = true;
pub const DEFAULT_GC_HORIZON: u64 = 64 * 1024 * 1024;
@@ -589,8 +584,9 @@ pub mod tenant_conf_defaults {
pub const DEFAULT_GC_PERIOD: &str = "1 hr";
pub const DEFAULT_IMAGE_CREATION_THRESHOLD: usize = 3;
// If there are more than threshold * compaction_threshold (that is 3 * 10 in the default config) L0 layers, image
// layer creation will end immediately. Set to 0 to disable.
pub const DEFAULT_IMAGE_CREATION_PREEMPT_THRESHOLD: usize = 3;
// layer creation will end immediately. Set to 0 to disable. The target default will be 3 once we
// want to enable this feature.
pub const DEFAULT_IMAGE_CREATION_PREEMPT_THRESHOLD: usize = 0;
pub const DEFAULT_PITR_INTERVAL: &str = "7 days";
pub const DEFAULT_WALRECEIVER_CONNECT_TIMEOUT: &str = "10 seconds";
pub const DEFAULT_WALRECEIVER_LAGGING_WAL_TIMEOUT: &str = "10 seconds";

View File

@@ -176,39 +176,6 @@ impl LsnLease {
}
}
/// Controls the detach ancestor behavior.
/// - When set to `NoAncestorAndReparent`, we will only detach a branch if its ancestor is a root branch. It will automatically reparent any children of the ancestor before and at the branch point.
/// - When set to `MultiLevelAndNoReparent`, we will detach a branch from multiple levels of ancestors, and no reparenting will happen at all.
#[derive(Debug, Clone, Copy, Default)]
pub enum DetachBehavior {
#[default]
NoAncestorAndReparent,
MultiLevelAndNoReparent,
}
impl std::str::FromStr for DetachBehavior {
type Err = &'static str;
fn from_str(s: &str) -> Result<Self, Self::Err> {
match s {
"no_ancestor_and_reparent" => Ok(DetachBehavior::NoAncestorAndReparent),
"multi_level_and_no_reparent" => Ok(DetachBehavior::MultiLevelAndNoReparent),
"v1" => Ok(DetachBehavior::NoAncestorAndReparent),
"v2" => Ok(DetachBehavior::MultiLevelAndNoReparent),
_ => Err("cannot parse detach behavior"),
}
}
}
impl std::fmt::Display for DetachBehavior {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
match self {
DetachBehavior::NoAncestorAndReparent => write!(f, "no_ancestor_and_reparent"),
DetachBehavior::MultiLevelAndNoReparent => write!(f, "multi_level_and_no_reparent"),
}
}
}
/// The only [`TenantState`] variants we could be `TenantState::Activating` from.
///
/// XXX: We used to have more variants here, but now it's just one, which makes this rather
@@ -307,31 +274,6 @@ pub struct TimelineCreateRequest {
pub mode: TimelineCreateRequestMode,
}
/// Storage controller specific extensions to [`TimelineInfo`].
#[derive(Serialize, Deserialize, Clone)]
pub struct TimelineCreateResponseStorcon {
#[serde(flatten)]
pub timeline_info: TimelineInfo,
pub safekeepers: Option<SafekeepersInfo>,
}
/// Safekeepers as returned in timeline creation request to storcon or pushed to
/// cplane in the post migration hook.
#[derive(Serialize, Deserialize, Clone)]
pub struct SafekeepersInfo {
pub tenant_id: TenantId,
pub timeline_id: TimelineId,
pub generation: u32,
pub safekeepers: Vec<SafekeeperInfo>,
}
#[derive(Serialize, Deserialize, Clone)]
pub struct SafekeeperInfo {
pub id: NodeId,
pub hostname: String,
}
#[derive(Serialize, Deserialize, Clone)]
#[serde(untagged)]
pub enum TimelineCreateRequestMode {
@@ -1258,10 +1200,9 @@ pub struct TimelineInfo {
pub last_record_lsn: Lsn,
pub prev_record_lsn: Option<Lsn>,
/// Legacy field, retained for one version to enable old storage controller to
/// decode (it was a mandatory field).
#[serde(default, rename = "latest_gc_cutoff_lsn")]
pub _unused: Lsn,
/// Legacy field for compat with control plane. Synonym of `min_readable_lsn`.
/// TODO: remove once control plane no longer reads it.
pub latest_gc_cutoff_lsn: Lsn,
/// The LSN up to which GC has advanced: older data may still exist but it is not available for clients.
/// This LSN is not suitable for deciding where to create branches etc: use [`TimelineInfo::min_readable_lsn`] instead,
@@ -1510,14 +1451,8 @@ pub struct TenantScanRemoteStorageResponse {
#[derive(Serialize, Deserialize, Debug, Clone)]
#[serde(rename_all = "snake_case")]
pub enum TenantSorting {
/// Total size of layers on local disk for all timelines in a shard.
ResidentSize,
/// The logical size of the largest timeline within a _tenant_ (not shard). Only tracked on
/// shard 0, contains the sum across all shards.
MaxLogicalSize,
/// The logical size of the largest timeline within a _tenant_ (not shard), divided by number of
/// shards. Only tracked on shard 0, and estimates the per-shard logical size.
MaxLogicalSizePerShard,
}
impl Default for TenantSorting {
@@ -1547,20 +1482,14 @@ pub struct TopTenantShardsRequest {
pub struct TopTenantShardItem {
pub id: TenantShardId,
/// Total size of layers on local disk for all timelines in this shard.
/// Total size of layers on local disk for all timelines in this tenant
pub resident_size: u64,
/// Total size of layers in remote storage for all timelines in this shard.
/// Total size of layers in remote storage for all timelines in this tenant
pub physical_size: u64,
/// The largest logical size of a timeline within this _tenant_ (not shard). This is only
/// tracked on shard 0, and contains the sum of the logical size across all shards.
/// The largest logical size of a timeline within this tenant
pub max_logical_size: u64,
/// The largest logical size of a timeline within this _tenant_ (not shard) divided by number of
/// shards. This is only tracked on shard 0, and is only an estimate as we divide it evenly by
/// shard count, rounded up.
pub max_logical_size_per_shard: u64,
}
#[derive(Serialize, Deserialize, Debug, Default)]

View File

@@ -112,16 +112,6 @@ impl ShardIdentity {
}
}
/// An unsharded identity with the given stripe size (if non-zero). This is typically used to
/// carry over a stripe size for an unsharded tenant from persistent storage.
pub fn unsharded_with_stripe_size(stripe_size: ShardStripeSize) -> Self {
let mut shard_identity = Self::unsharded();
if stripe_size.0 > 0 {
shard_identity.stripe_size = stripe_size;
}
shard_identity
}
/// A broken instance of this type is only used for `TenantState::Broken` tenants,
/// which are constructed in code paths that don't have access to proper configuration.
///

View File

@@ -135,8 +135,8 @@ impl Type {
pub enum Kind {
/// A simple type like `VARCHAR` or `INTEGER`.
Simple,
/// An enumerated type.
Enum,
/// An enumerated type along with its variants.
Enum(Vec<String>),
/// A pseudo-type.
Pseudo,
/// An array type along with the type of its elements.
@@ -146,9 +146,9 @@ pub enum Kind {
/// A multirange type along with the type of its elements.
Multirange(Type),
/// A domain type along with its underlying type.
Domain(Oid),
/// A composite type.
Composite(Oid),
Domain(Type),
/// A composite type along with information about its fields.
Composite(Vec<Field>),
}
/// Information about a field of a composite type.

View File

@@ -19,10 +19,10 @@ use crate::config::{Host, SslMode};
use crate::connection::{Request, RequestMessages};
use crate::query::RowStream;
use crate::simple_query::SimpleQueryStream;
use crate::types::{Oid, Type};
use crate::types::{Oid, ToSql, Type};
use crate::{
CancelToken, Error, ReadyForQueryStatus, SimpleQueryMessage, Statement, Transaction,
TransactionBuilder, query, simple_query,
CancelToken, Error, ReadyForQueryStatus, Row, SimpleQueryMessage, Statement, Transaction,
TransactionBuilder, query, simple_query, slice_iter,
};
pub struct Responses {
@@ -54,18 +54,26 @@ impl Responses {
/// A cache of type info and prepared statements for fetching type info
/// (corresponding to the queries in the [crate::prepare] module).
#[derive(Default)]
pub(crate) struct CachedTypeInfo {
struct CachedTypeInfo {
/// A statement for basic information for a type from its
/// OID. Corresponds to [TYPEINFO_QUERY](crate::prepare::TYPEINFO_QUERY) (or its
/// fallback).
pub(crate) typeinfo: Option<Statement>,
typeinfo: Option<Statement>,
/// A statement for getting information for a composite type from its OID.
/// Corresponds to [TYPEINFO_QUERY](crate::prepare::TYPEINFO_COMPOSITE_QUERY).
typeinfo_composite: Option<Statement>,
/// A statement for getting information for a composite type from its OID.
/// Corresponds to [TYPEINFO_QUERY](crate::prepare::TYPEINFO_COMPOSITE_QUERY) (or
/// its fallback).
typeinfo_enum: Option<Statement>,
/// Cache of types already looked up.
pub(crate) types: HashMap<Oid, Type>,
types: HashMap<Oid, Type>,
}
pub struct InnerClient {
sender: mpsc::UnboundedSender<Request>,
cached_typeinfo: Mutex<CachedTypeInfo>,
/// A buffer to use when writing out postgres commands.
buffer: Mutex<BytesMut>,
@@ -83,6 +91,38 @@ impl InnerClient {
})
}
pub fn typeinfo(&self) -> Option<Statement> {
self.cached_typeinfo.lock().typeinfo.clone()
}
pub fn set_typeinfo(&self, statement: &Statement) {
self.cached_typeinfo.lock().typeinfo = Some(statement.clone());
}
pub fn typeinfo_composite(&self) -> Option<Statement> {
self.cached_typeinfo.lock().typeinfo_composite.clone()
}
pub fn set_typeinfo_composite(&self, statement: &Statement) {
self.cached_typeinfo.lock().typeinfo_composite = Some(statement.clone());
}
pub fn typeinfo_enum(&self) -> Option<Statement> {
self.cached_typeinfo.lock().typeinfo_enum.clone()
}
pub fn set_typeinfo_enum(&self, statement: &Statement) {
self.cached_typeinfo.lock().typeinfo_enum = Some(statement.clone());
}
pub fn type_(&self, oid: Oid) -> Option<Type> {
self.cached_typeinfo.lock().types.get(&oid).cloned()
}
pub fn set_type(&self, oid: Oid, type_: &Type) {
self.cached_typeinfo.lock().types.insert(oid, type_.clone());
}
/// Call the given function with a buffer to be used when writing out
/// postgres commands.
pub fn with_buf<F, R>(&self, f: F) -> R
@@ -102,6 +142,7 @@ pub struct SocketConfig {
pub host: Host,
pub port: u16,
pub connect_timeout: Option<Duration>,
// pub keepalive: Option<KeepaliveConfig>,
}
/// An asynchronous PostgreSQL client.
@@ -110,7 +151,6 @@ pub struct SocketConfig {
/// through this client object.
pub struct Client {
inner: Arc<InnerClient>,
cached_typeinfo: CachedTypeInfo,
socket_config: SocketConfig,
ssl_mode: SslMode,
@@ -129,9 +169,9 @@ impl Client {
Client {
inner: Arc::new(InnerClient {
sender,
cached_typeinfo: Default::default(),
buffer: Default::default(),
}),
cached_typeinfo: Default::default(),
socket_config,
ssl_mode,
@@ -149,6 +189,55 @@ impl Client {
&self.inner
}
/// Executes a statement, returning a vector of the resulting rows.
///
/// A statement may contain parameters, specified by `$n`, where `n` is the index of the parameter of the list
/// provided, 1-indexed.
///
/// The `statement` argument can either be a `Statement`, or a raw query string. If the same statement will be
/// repeatedly executed (perhaps with different query parameters), consider preparing the statement up front
/// with the `prepare` method.
///
/// # Panics
///
/// Panics if the number of parameters provided does not match the number expected.
pub async fn query(
&self,
statement: Statement,
params: &[&(dyn ToSql + Sync)],
) -> Result<Vec<Row>, Error> {
self.query_raw(statement, slice_iter(params))
.await?
.try_collect()
.await
}
/// The maximally flexible version of [`query`].
///
/// A statement may contain parameters, specified by `$n`, where `n` is the index of the parameter of the list
/// provided, 1-indexed.
///
/// The `statement` argument can either be a `Statement`, or a raw query string. If the same statement will be
/// repeatedly executed (perhaps with different query parameters), consider preparing the statement up front
/// with the `prepare` method.
///
/// # Panics
///
/// Panics if the number of parameters provided does not match the number expected.
///
/// [`query`]: #method.query
pub async fn query_raw<'a, I>(
&self,
statement: Statement,
params: I,
) -> Result<RowStream, Error>
where
I: IntoIterator<Item = &'a (dyn ToSql + Sync)>,
I::IntoIter: ExactSizeIterator,
{
query::query(&self.inner, statement, params).await
}
/// Pass text directly to the Postgres backend to allow it to sort out typing itself and
/// to save a roundtrip
pub async fn query_raw_txt<S, I>(&self, statement: &str, params: I) -> Result<RowStream, Error>
@@ -195,14 +284,6 @@ impl Client {
simple_query::batch_execute(self.inner(), query).await
}
pub async fn discard_all(&mut self) -> Result<ReadyForQueryStatus, Error> {
// clear the prepared statements that are about to be nuked from the postgres session
self.cached_typeinfo.typeinfo = None;
self.batch_execute("discard all").await
}
/// Begins a new database transaction.
///
/// The transaction will roll back by default - use the `commit` method to commit it.
@@ -266,8 +347,8 @@ impl Client {
}
/// Query for type information
pub(crate) async fn get_type_inner(&mut self, oid: Oid) -> Result<Type, Error> {
crate::prepare::get_type(&self.inner, &mut self.cached_typeinfo, oid).await
pub async fn get_type(&self, oid: Oid) -> Result<Type, Error> {
crate::prepare::get_type(&self.inner, oid).await
}
/// Determines if the connection to the server has already closed.

View File

@@ -22,7 +22,7 @@ pub trait GenericClient: private::Sealed {
I::IntoIter: ExactSizeIterator + Sync + Send;
/// Query for type information
async fn get_type(&mut self, oid: Oid) -> Result<Type, Error>;
async fn get_type(&self, oid: Oid) -> Result<Type, Error>;
}
impl private::Sealed for Client {}
@@ -38,8 +38,8 @@ impl GenericClient for Client {
}
/// Query for type information
async fn get_type(&mut self, oid: Oid) -> Result<Type, Error> {
self.get_type_inner(oid).await
async fn get_type(&self, oid: Oid) -> Result<Type, Error> {
crate::prepare::get_type(self.inner(), oid).await
}
}
@@ -56,7 +56,7 @@ impl GenericClient for Transaction<'_> {
}
/// Query for type information
async fn get_type(&mut self, oid: Oid) -> Result<Type, Error> {
self.client_mut().get_type(oid).await
async fn get_type(&self, oid: Oid) -> Result<Type, Error> {
self.client().get_type(oid).await
}
}

View File

@@ -9,10 +9,10 @@ use log::debug;
use postgres_protocol2::message::backend::Message;
use postgres_protocol2::message::frontend;
use crate::client::{CachedTypeInfo, InnerClient};
use crate::client::InnerClient;
use crate::codec::FrontendMessage;
use crate::connection::RequestMessages;
use crate::types::{Kind, Oid, Type};
use crate::types::{Field, Kind, Oid, Type};
use crate::{Column, Error, Statement, query, slice_iter};
pub(crate) const TYPEINFO_QUERY: &str = "\
@@ -23,7 +23,23 @@ INNER JOIN pg_catalog.pg_namespace n ON t.typnamespace = n.oid
WHERE t.oid = $1
";
async fn prepare_typecheck(
const TYPEINFO_ENUM_QUERY: &str = "\
SELECT enumlabel
FROM pg_catalog.pg_enum
WHERE enumtypid = $1
ORDER BY enumsortorder
";
pub(crate) const TYPEINFO_COMPOSITE_QUERY: &str = "\
SELECT attname, atttypid
FROM pg_catalog.pg_attribute
WHERE attrelid = $1
AND NOT attisdropped
AND attnum > 0
ORDER BY attnum
";
pub async fn prepare(
client: &Arc<InnerClient>,
name: &'static str,
query: &str,
@@ -51,7 +67,7 @@ async fn prepare_typecheck(
let mut parameters = vec![];
let mut it = parameter_description.parameters();
while let Some(oid) = it.next().map_err(Error::parse)? {
let type_ = Type::from_oid(oid).ok_or_else(Error::unexpected_message)?;
let type_ = get_type(client, oid).await?;
parameters.push(type_);
}
@@ -59,7 +75,7 @@ async fn prepare_typecheck(
if let Some(row_description) = row_description {
let mut it = row_description.fields();
while let Some(field) = it.next().map_err(Error::parse)? {
let type_ = Type::from_oid(field.type_oid()).ok_or_else(Error::unexpected_message)?;
let type_ = get_type(client, field.type_oid()).await?;
let column = Column::new(field.name().to_string(), type_, field);
columns.push(column);
}
@@ -68,6 +84,15 @@ async fn prepare_typecheck(
Ok(Statement::new(client, name, parameters, columns))
}
fn prepare_rec<'a>(
client: &'a Arc<InnerClient>,
name: &'static str,
query: &'a str,
types: &'a [Type],
) -> Pin<Box<dyn Future<Output = Result<Statement, Error>> + 'a + Send>> {
Box::pin(prepare(client, name, query, types))
}
fn encode(client: &InnerClient, name: &str, query: &str, types: &[Type]) -> Result<Bytes, Error> {
if types.is_empty() {
debug!("preparing query {}: {}", name, query);
@@ -83,20 +108,16 @@ fn encode(client: &InnerClient, name: &str, query: &str, types: &[Type]) -> Resu
})
}
pub async fn get_type(
client: &Arc<InnerClient>,
typecache: &mut CachedTypeInfo,
oid: Oid,
) -> Result<Type, Error> {
pub async fn get_type(client: &Arc<InnerClient>, oid: Oid) -> Result<Type, Error> {
if let Some(type_) = Type::from_oid(oid) {
return Ok(type_);
}
if let Some(type_) = typecache.types.get(&oid) {
return Ok(type_.clone());
};
if let Some(type_) = client.type_(oid) {
return Ok(type_);
}
let stmt = typeinfo_statement(client, typecache).await?;
let stmt = typeinfo_statement(client).await?;
let rows = query::query(client, stmt, slice_iter(&[&oid])).await?;
pin_mut!(rows);
@@ -115,48 +136,100 @@ pub async fn get_type(
let relid: Oid = row.try_get(6)?;
let kind = if type_ == b'e' as i8 {
Kind::Enum
let variants = get_enum_variants(client, oid).await?;
Kind::Enum(variants)
} else if type_ == b'p' as i8 {
Kind::Pseudo
} else if basetype != 0 {
Kind::Domain(basetype)
let type_ = get_type_rec(client, basetype).await?;
Kind::Domain(type_)
} else if elem_oid != 0 {
let type_ = get_type_rec(client, typecache, elem_oid).await?;
let type_ = get_type_rec(client, elem_oid).await?;
Kind::Array(type_)
} else if relid != 0 {
Kind::Composite(relid)
let fields = get_composite_fields(client, relid).await?;
Kind::Composite(fields)
} else if let Some(rngsubtype) = rngsubtype {
let type_ = get_type_rec(client, typecache, rngsubtype).await?;
let type_ = get_type_rec(client, rngsubtype).await?;
Kind::Range(type_)
} else {
Kind::Simple
};
let type_ = Type::new(name, oid, kind, schema);
typecache.types.insert(oid, type_.clone());
client.set_type(oid, &type_);
Ok(type_)
}
fn get_type_rec<'a>(
client: &'a Arc<InnerClient>,
typecache: &'a mut CachedTypeInfo,
oid: Oid,
) -> Pin<Box<dyn Future<Output = Result<Type, Error>> + Send + 'a>> {
Box::pin(get_type(client, typecache, oid))
Box::pin(get_type(client, oid))
}
async fn typeinfo_statement(
client: &Arc<InnerClient>,
typecache: &mut CachedTypeInfo,
) -> Result<Statement, Error> {
if let Some(stmt) = &typecache.typeinfo {
return Ok(stmt.clone());
async fn typeinfo_statement(client: &Arc<InnerClient>) -> Result<Statement, Error> {
if let Some(stmt) = client.typeinfo() {
return Ok(stmt);
}
let typeinfo = "neon_proxy_typeinfo";
let stmt = prepare_typecheck(client, typeinfo, TYPEINFO_QUERY, &[]).await?;
let stmt = prepare_rec(client, typeinfo, TYPEINFO_QUERY, &[]).await?;
typecache.typeinfo = Some(stmt.clone());
client.set_typeinfo(&stmt);
Ok(stmt)
}
async fn get_enum_variants(client: &Arc<InnerClient>, oid: Oid) -> Result<Vec<String>, Error> {
let stmt = typeinfo_enum_statement(client).await?;
query::query(client, stmt, slice_iter(&[&oid]))
.await?
.and_then(|row| async move { row.try_get(0) })
.try_collect()
.await
}
async fn typeinfo_enum_statement(client: &Arc<InnerClient>) -> Result<Statement, Error> {
if let Some(stmt) = client.typeinfo_enum() {
return Ok(stmt);
}
let typeinfo = "neon_proxy_typeinfo_enum";
let stmt = prepare_rec(client, typeinfo, TYPEINFO_ENUM_QUERY, &[]).await?;
client.set_typeinfo_enum(&stmt);
Ok(stmt)
}
async fn get_composite_fields(client: &Arc<InnerClient>, oid: Oid) -> Result<Vec<Field>, Error> {
let stmt = typeinfo_composite_statement(client).await?;
let rows = query::query(client, stmt, slice_iter(&[&oid]))
.await?
.try_collect::<Vec<_>>()
.await?;
let mut fields = vec![];
for row in rows {
let name = row.try_get(0)?;
let oid = row.try_get(1)?;
let type_ = get_type_rec(client, oid).await?;
fields.push(Field::new(name, type_));
}
Ok(fields)
}
async fn typeinfo_composite_statement(client: &Arc<InnerClient>) -> Result<Statement, Error> {
if let Some(stmt) = client.typeinfo_composite() {
return Ok(stmt);
}
let typeinfo = "neon_proxy_typeinfo_composite";
let stmt = prepare_rec(client, typeinfo, TYPEINFO_COMPOSITE_QUERY, &[]).await?;
client.set_typeinfo_composite(&stmt);
Ok(stmt)
}

View File

@@ -72,9 +72,4 @@ impl<'a> Transaction<'a> {
pub fn client(&self) -> &Client {
self.client
}
/// Returns a reference to the underlying `Client`.
pub fn client_mut(&mut self) -> &mut Client {
self.client
}
}

View File

@@ -131,14 +131,6 @@ impl Configuration {
}
}
pub fn new(members: MemberSet) -> Self {
Configuration {
generation: INITIAL_GENERATION,
members,
new_members: None,
}
}
/// Is `sk_id` member of the configuration?
pub fn contains(&self, sk_id: NodeId) -> bool {
self.members.contains(sk_id) || self.new_members.as_ref().is_some_and(|m| m.contains(sk_id))

View File

@@ -18,7 +18,7 @@ pub struct SafekeeperStatus {
pub id: NodeId,
}
#[derive(Serialize, Deserialize, Clone)]
#[derive(Serialize, Deserialize)]
pub struct TimelineCreateRequest {
pub tenant_id: TenantId,
pub timeline_id: TimelineId,
@@ -221,11 +221,6 @@ pub struct TimelineMembershipSwitchResponse {
pub current_conf: Configuration,
}
#[derive(Clone, Copy, Serialize, Deserialize)]
pub struct TimelineDeleteResult {
pub dir_existed: bool,
}
fn lsn_invalid() -> Lsn {
Lsn::INVALID
}
@@ -288,7 +283,7 @@ pub struct SafekeeperUtilization {
}
/// pull_timeline request body.
#[derive(Debug, Clone, Deserialize, Serialize)]
#[derive(Debug, Deserialize, Serialize)]
pub struct PullTimelineRequest {
pub tenant_id: TenantId,
pub timeline_id: TimelineId,

View File

@@ -21,7 +21,7 @@
//! .with_writer(std::io::stderr);
//!
//! // Initialize OpenTelemetry. Exports tracing spans as OpenTelemetry traces
//! let otlp_layer = tracing_utils::init_tracing("my_application", tracing_utils::ExportConfig::default()).await;
//! let otlp_layer = tracing_utils::init_tracing("my_application").await;
//!
//! // Put it all together
//! tracing_subscriber::registry()
@@ -38,12 +38,8 @@ pub mod http;
use opentelemetry::KeyValue;
use opentelemetry::trace::TracerProvider;
use opentelemetry_otlp::WithExportConfig;
pub use opentelemetry_otlp::{ExportConfig, Protocol};
use tracing::level_filters::LevelFilter;
use tracing::{Dispatch, Subscriber};
use tracing::Subscriber;
use tracing_subscriber::Layer;
use tracing_subscriber::layer::SubscriberExt;
use tracing_subscriber::registry::LookupSpan;
/// Set up OpenTelemetry exporter, using configuration from environment variables.
@@ -73,28 +69,19 @@ use tracing_subscriber::registry::LookupSpan;
///
/// This doesn't block, but is marked as 'async' to hint that this must be called in
/// asynchronous execution context.
pub async fn init_tracing<S>(
service_name: &str,
export_config: ExportConfig,
) -> Option<impl Layer<S>>
pub async fn init_tracing<S>(service_name: &str) -> Option<impl Layer<S>>
where
S: Subscriber + for<'span> LookupSpan<'span>,
{
if std::env::var("OTEL_SDK_DISABLED") == Ok("true".to_string()) {
return None;
};
Some(init_tracing_internal(
service_name.to_string(),
export_config,
))
Some(init_tracing_internal(service_name.to_string()))
}
/// Like `init_tracing`, but creates a separate tokio Runtime for the tracing
/// tasks.
pub fn init_tracing_without_runtime<S>(
service_name: &str,
export_config: ExportConfig,
) -> Option<impl Layer<S>>
pub fn init_tracing_without_runtime<S>(service_name: &str) -> Option<impl Layer<S>>
where
S: Subscriber + for<'span> LookupSpan<'span>,
{
@@ -125,22 +112,16 @@ where
));
let _guard = runtime.enter();
Some(init_tracing_internal(
service_name.to_string(),
export_config,
))
Some(init_tracing_internal(service_name.to_string()))
}
fn init_tracing_internal<S>(service_name: String, export_config: ExportConfig) -> impl Layer<S>
fn init_tracing_internal<S>(service_name: String) -> impl Layer<S>
where
S: Subscriber + for<'span> LookupSpan<'span>,
{
// Sets up exporter from the provided [`ExportConfig`] parameter.
// If the endpoint is not specified, it is loaded from the
// OTEL_EXPORTER_OTLP_ENDPOINT environment variable.
// Sets up exporter from the OTEL_EXPORTER_* environment variables.
let exporter = opentelemetry_otlp::SpanExporter::builder()
.with_http()
.with_export_config(export_config)
.build()
.expect("could not initialize opentelemetry exporter");
@@ -170,51 +151,3 @@ where
pub fn shutdown_tracing() {
opentelemetry::global::shutdown_tracer_provider();
}
pub enum OtelEnablement {
Disabled,
Enabled {
service_name: String,
export_config: ExportConfig,
runtime: &'static tokio::runtime::Runtime,
},
}
pub struct OtelGuard {
pub dispatch: Dispatch,
}
impl Drop for OtelGuard {
fn drop(&mut self) {
shutdown_tracing();
}
}
/// Initializes OTEL infrastructure for performance tracing according to the provided configuration
///
/// Performance tracing is handled by a different [`tracing::Subscriber`]. This functions returns
/// an [`OtelGuard`] containing a [`tracing::Dispatch`] associated with a newly created subscriber.
/// Applications should use this dispatch for their performance traces.
///
/// The lifetime of the guard should match taht of the application. On drop, it tears down the
/// OTEL infra.
pub fn init_performance_tracing(otel_enablement: OtelEnablement) -> Option<OtelGuard> {
let otel_subscriber = match otel_enablement {
OtelEnablement::Disabled => None,
OtelEnablement::Enabled {
service_name,
export_config,
runtime,
} => {
let otel_layer = runtime
.block_on(init_tracing(&service_name, export_config))
.with_filter(LevelFilter::INFO);
let otel_subscriber = tracing_subscriber::registry().with(otel_layer);
let otel_dispatch = Dispatch::new(otel_subscriber);
Some(otel_dispatch)
}
};
otel_subscriber.map(|dispatch| OtelGuard { dispatch })
}

View File

@@ -42,7 +42,6 @@ toml_edit = { workspace = true, features = ["serde"] }
tracing.workspace = true
tracing-error.workspace = true
tracing-subscriber = { workspace = true, features = ["json", "registry"] }
tracing-utils.workspace = true
rand.workspace = true
scopeguard.workspace = true
strum.workspace = true

View File

@@ -49,13 +49,7 @@ pub fn bench_log_slow(c: &mut Criterion) {
// performance too. Use a simple noop future that yields once, to avoid any scheduler fast
// paths for a ready future.
if enabled {
b.iter(|| {
runtime.block_on(log_slow(
"ready",
THRESHOLD,
std::pin::pin!(tokio::task::yield_now()),
))
});
b.iter(|| runtime.block_on(log_slow("ready", THRESHOLD, tokio::task::yield_now())));
} else {
b.iter(|| runtime.block_on(tokio::task::yield_now()));
}

View File

@@ -165,7 +165,6 @@ pub fn init(
};
log_layer.with_filter(rust_log_env_filter())
});
let r = r.with(
TracingEventCountLayer(&TRACING_EVENT_COUNT_METRIC).with_filter(rust_log_env_filter()),
);
@@ -331,90 +330,37 @@ impl std::fmt::Debug for SecretString {
///
/// TODO: consider upgrading this to a warning, but currently it fires too often.
#[inline]
pub async fn log_slow<F, O>(name: &str, threshold: Duration, f: std::pin::Pin<&mut F>) -> O
where
F: Future<Output = O>,
{
monitor_slow_future(
threshold,
threshold, // period = threshold
f,
|MonitorSlowFutureCallback {
ready,
is_slow,
elapsed_total,
elapsed_since_last_callback: _,
}| {
if !is_slow {
return;
}
if ready {
info!(
"slow {name} completed after {:.3}s",
elapsed_total.as_secs_f64()
);
} else {
info!(
"slow {name} still running after {:.3}s",
elapsed_total.as_secs_f64()
);
}
},
)
.await
}
pub async fn log_slow<O>(name: &str, threshold: Duration, f: impl Future<Output = O>) -> O {
// TODO: we unfortunately have to pin the future on the heap, since GetPage futures are huge and
// won't fit on the stack.
let mut f = Box::pin(f);
/// Poll future `fut` to completion, invoking callback `cb` at the given `threshold` and every
/// `period` afterwards, and also unconditionally when the future completes.
#[inline]
pub async fn monitor_slow_future<F, O>(
threshold: Duration,
period: Duration,
mut fut: std::pin::Pin<&mut F>,
mut cb: impl FnMut(MonitorSlowFutureCallback),
) -> O
where
F: Future<Output = O>,
{
let started = Instant::now();
let mut attempt = 1;
let mut last_cb = started;
loop {
// NB: use timeout_at() instead of timeout() to avoid an extra clock reading in the common
// case where the timeout doesn't fire.
let deadline = started + threshold + (attempt - 1) * period;
// TODO: still call the callback if the future panics? Copy how we do it for the page_service flush_in_progress counter.
let res = tokio::time::timeout_at(deadline, &mut fut).await;
let now = Instant::now();
let elapsed_total = now - started;
cb(MonitorSlowFutureCallback {
ready: res.is_ok(),
is_slow: elapsed_total >= threshold,
elapsed_total,
elapsed_since_last_callback: now - last_cb,
});
last_cb = now;
if let Ok(output) = res {
let deadline = started + attempt * threshold;
if let Ok(output) = tokio::time::timeout_at(deadline, &mut f).await {
// NB: we check if we exceeded the threshold even if the timeout never fired, because
// scheduling or execution delays may cause the future to succeed even if it exceeds the
// timeout. This costs an extra unconditional clock reading, but seems worth it to avoid
// false negatives.
let elapsed = started.elapsed();
if elapsed >= threshold {
info!("slow {name} completed after {:.3}s", elapsed.as_secs_f64());
}
return output;
}
let elapsed = started.elapsed().as_secs_f64();
info!("slow {name} still running after {elapsed:.3}s",);
attempt += 1;
}
}
/// See [`monitor_slow_future`].
pub struct MonitorSlowFutureCallback {
/// Whether the future completed. If true, there will be no more callbacks.
pub ready: bool,
/// Whether the future is taking `>=` the specififed threshold duration to complete.
/// Monotonic: if true in one callback invocation, true in all subsequent onces.
pub is_slow: bool,
/// The time elapsed since the [`monitor_slow_future`] was first polled.
pub elapsed_total: Duration,
/// The time elapsed since the last callback invocation.
/// For the initial callback invocation, the time elapsed since the [`monitor_slow_future`] was first polled.
pub elapsed_since_last_callback: Duration,
}
#[cfg(test)]
mod tests {
use metrics::IntCounterVec;

View File

@@ -48,6 +48,8 @@ pprof.workspace = true
rand.workspace = true
range-set-blaze = { version = "0.1.16", features = ["alloc"] }
regex.workspace = true
rustls-pemfile.workspace = true
rustls-pki-types.workspace = true
rustls.workspace = true
scopeguard.workspace = true
send-future.workspace = true
@@ -68,7 +70,6 @@ tokio-stream.workspace = true
tokio-util.workspace = true
toml_edit = { workspace = true, features = [ "serde" ] }
tracing.workspace = true
tracing-utils.workspace = true
url.workspace = true
walkdir.workspace = true
metrics.workspace = true
@@ -119,10 +120,6 @@ harness = false
name = "upload_queue"
harness = false
[[bench]]
name = "bench_metrics"
harness = false
[[bin]]
name = "test_helper_slow_client_reads"
required-features = [ "testing" ]

View File

@@ -1,366 +0,0 @@
use criterion::{BenchmarkId, Criterion, criterion_group, criterion_main};
use utils::id::{TenantId, TimelineId};
//
// Demonstrates that repeat label values lookup is a multicore scalability bottleneck
// that is worth avoiding.
//
criterion_group!(
label_values,
label_values::bench_naive_usage,
label_values::bench_cache_label_values_lookup
);
mod label_values {
use super::*;
pub fn bench_naive_usage(c: &mut Criterion) {
let mut g = c.benchmark_group("label_values__naive_usage");
for ntimelines in [1, 4, 8] {
g.bench_with_input(
BenchmarkId::new("ntimelines", ntimelines),
&ntimelines,
|b, ntimelines| {
b.iter_custom(|iters| {
let barrier = std::sync::Barrier::new(*ntimelines + 1);
let timelines = (0..*ntimelines)
.map(|_| {
(
TenantId::generate().to_string(),
"0000".to_string(),
TimelineId::generate().to_string(),
)
})
.collect::<Vec<_>>();
let metric_vec = metrics::UIntGaugeVec::new(
metrics::opts!("testmetric", "testhelp"),
&["tenant_id", "shard_id", "timeline_id"],
)
.unwrap();
std::thread::scope(|s| {
for (tenant_id, shard_id, timeline_id) in &timelines {
s.spawn(|| {
barrier.wait();
for _ in 0..iters {
metric_vec
.with_label_values(&[tenant_id, shard_id, timeline_id])
.inc();
}
barrier.wait();
});
}
barrier.wait();
let start = std::time::Instant::now();
barrier.wait();
start.elapsed()
})
})
},
);
}
g.finish();
}
pub fn bench_cache_label_values_lookup(c: &mut Criterion) {
let mut g = c.benchmark_group("label_values__cache_label_values_lookup");
for ntimelines in [1, 4, 8] {
g.bench_with_input(
BenchmarkId::new("ntimelines", ntimelines),
&ntimelines,
|b, ntimelines| {
b.iter_custom(|iters| {
let barrier = std::sync::Barrier::new(*ntimelines + 1);
let timelines = (0..*ntimelines)
.map(|_| {
(
TenantId::generate().to_string(),
"0000".to_string(),
TimelineId::generate().to_string(),
)
})
.collect::<Vec<_>>();
let metric_vec = metrics::UIntGaugeVec::new(
metrics::opts!("testmetric", "testhelp"),
&["tenant_id", "shard_id", "timeline_id"],
)
.unwrap();
std::thread::scope(|s| {
for (tenant_id, shard_id, timeline_id) in &timelines {
s.spawn(|| {
let metric = metric_vec.with_label_values(&[
tenant_id,
shard_id,
timeline_id,
]);
barrier.wait();
for _ in 0..iters {
metric.inc();
}
barrier.wait();
});
}
barrier.wait();
let start = std::time::Instant::now();
barrier.wait();
start.elapsed()
})
})
},
);
}
g.finish();
}
}
//
// Demonstrates that even a single metric can be a scalability bottleneck
// if multiple threads in it concurrently but there's nothing we can do
// about it without changing the metrics framework to use e.g. sharded counte atomics.
//
criterion_group!(
single_metric_multicore_scalability,
single_metric_multicore_scalability::bench,
);
mod single_metric_multicore_scalability {
use super::*;
pub fn bench(c: &mut Criterion) {
let mut g = c.benchmark_group("single_metric_multicore_scalability");
for nthreads in [1, 4, 8] {
g.bench_with_input(
BenchmarkId::new("nthreads", nthreads),
&nthreads,
|b, nthreads| {
b.iter_custom(|iters| {
let barrier = std::sync::Barrier::new(*nthreads + 1);
let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
std::thread::scope(|s| {
for _ in 0..*nthreads {
s.spawn(|| {
barrier.wait();
for _ in 0..iters {
metric.inc();
}
barrier.wait();
});
}
barrier.wait();
let start = std::time::Instant::now();
barrier.wait();
start.elapsed()
})
})
},
);
}
g.finish();
}
}
//
// Demonstrates that even if we cache label value, the propagation of such a cached metric value
// by Clone'ing it is a scalability bottleneck.
// The reason is that it's an Arc internally and thus there's contention on the reference count atomics.
//
// We can avoid that by having long-lived references per thread (= indirection).
//
criterion_group!(
propagation_of_cached_label_value,
propagation_of_cached_label_value::bench_naive,
propagation_of_cached_label_value::bench_long_lived_reference_per_thread,
);
mod propagation_of_cached_label_value {
use std::sync::Arc;
use super::*;
pub fn bench_naive(c: &mut Criterion) {
let mut g = c.benchmark_group("propagation_of_cached_label_value__naive");
for nthreads in [1, 4, 8] {
g.bench_with_input(
BenchmarkId::new("nthreads", nthreads),
&nthreads,
|b, nthreads| {
b.iter_custom(|iters| {
let barrier = std::sync::Barrier::new(*nthreads + 1);
let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
std::thread::scope(|s| {
for _ in 0..*nthreads {
s.spawn(|| {
barrier.wait();
for _ in 0..iters {
// propagating the metric means we'd clone it into the child RequestContext
let propagated = metric.clone();
// simulate some work
criterion::black_box(propagated);
}
barrier.wait();
});
}
barrier.wait();
let start = std::time::Instant::now();
barrier.wait();
start.elapsed()
})
})
},
);
}
g.finish();
}
pub fn bench_long_lived_reference_per_thread(c: &mut Criterion) {
let mut g =
c.benchmark_group("propagation_of_cached_label_value__long_lived_reference_per_thread");
for nthreads in [1, 4, 8] {
g.bench_with_input(
BenchmarkId::new("nthreads", nthreads),
&nthreads,
|b, nthreads| {
b.iter_custom(|iters| {
let barrier = std::sync::Barrier::new(*nthreads + 1);
let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
std::thread::scope(|s| {
for _ in 0..*nthreads {
s.spawn(|| {
// This is the technique.
let this_threads_metric_reference = Arc::new(metric.clone());
barrier.wait();
for _ in 0..iters {
// propagating the metric means we'd clone it into the child RequestContext
let propagated = Arc::clone(&this_threads_metric_reference);
// simulate some work (include the pointer chase!)
criterion::black_box(&*propagated);
}
barrier.wait();
});
}
barrier.wait();
let start = std::time::Instant::now();
barrier.wait();
start.elapsed()
})
})
},
);
}
}
}
criterion_main!(
label_values,
single_metric_multicore_scalability,
propagation_of_cached_label_value
);
/*
RUST_BACKTRACE=full cargo bench --bench bench_metrics -- --discard-baseline --noplot
Results on an im4gn.2xlarge instance
label_values__naive_usage/ntimelines/1 time: [178.71 ns 178.74 ns 178.76 ns]
label_values__naive_usage/ntimelines/4 time: [532.94 ns 539.59 ns 546.31 ns]
label_values__naive_usage/ntimelines/8 time: [1.1082 µs 1.1109 µs 1.1135 µs]
label_values__cache_label_values_lookup/ntimelines/1 time: [6.4116 ns 6.4119 ns 6.4123 ns]
label_values__cache_label_values_lookup/ntimelines/4 time: [6.3482 ns 6.3819 ns 6.4079 ns]
label_values__cache_label_values_lookup/ntimelines/8 time: [6.4213 ns 6.5279 ns 6.6293 ns]
single_metric_multicore_scalability/nthreads/1 time: [6.0102 ns 6.0104 ns 6.0106 ns]
single_metric_multicore_scalability/nthreads/4 time: [38.127 ns 38.275 ns 38.416 ns]
single_metric_multicore_scalability/nthreads/8 time: [73.698 ns 74.882 ns 75.864 ns]
propagation_of_cached_label_value__naive/nthreads/1 time: [14.424 ns 14.425 ns 14.426 ns]
propagation_of_cached_label_value__naive/nthreads/4 time: [100.71 ns 102.53 ns 104.35 ns]
propagation_of_cached_label_value__naive/nthreads/8 time: [211.50 ns 214.44 ns 216.87 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time: [14.135 ns 14.147 ns 14.160 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time: [14.243 ns 14.255 ns 14.268 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time: [14.470 ns 14.682 ns 14.895 ns]
Results on an i3en.3xlarge instance
label_values__naive_usage/ntimelines/1 time: [117.32 ns 117.53 ns 117.74 ns]
label_values__naive_usage/ntimelines/4 time: [736.58 ns 741.12 ns 745.61 ns]
label_values__naive_usage/ntimelines/8 time: [1.4513 µs 1.4596 µs 1.4665 µs]
label_values__cache_label_values_lookup/ntimelines/1 time: [8.0964 ns 8.0979 ns 8.0995 ns]
label_values__cache_label_values_lookup/ntimelines/4 time: [8.1620 ns 8.2912 ns 8.4491 ns]
label_values__cache_label_values_lookup/ntimelines/8 time: [14.148 ns 14.237 ns 14.324 ns]
single_metric_multicore_scalability/nthreads/1 time: [8.0993 ns 8.1013 ns 8.1046 ns]
single_metric_multicore_scalability/nthreads/4 time: [80.039 ns 80.672 ns 81.297 ns]
single_metric_multicore_scalability/nthreads/8 time: [153.58 ns 154.23 ns 154.90 ns]
propagation_of_cached_label_value__naive/nthreads/1 time: [13.924 ns 13.926 ns 13.928 ns]
propagation_of_cached_label_value__naive/nthreads/4 time: [143.66 ns 145.27 ns 146.59 ns]
propagation_of_cached_label_value__naive/nthreads/8 time: [296.51 ns 297.90 ns 299.30 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time: [14.013 ns 14.149 ns 14.308 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time: [14.311 ns 14.625 ns 14.984 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time: [25.981 ns 26.227 ns 26.476 ns]
Results on an Standard L16s v3 (16 vcpus, 128 GiB memory) Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz
label_values__naive_usage/ntimelines/1 time: [101.63 ns 101.84 ns 102.06 ns]
label_values__naive_usage/ntimelines/4 time: [417.55 ns 424.73 ns 432.63 ns]
label_values__naive_usage/ntimelines/8 time: [874.91 ns 889.51 ns 904.25 ns]
label_values__cache_label_values_lookup/ntimelines/1 time: [5.7724 ns 5.7760 ns 5.7804 ns]
label_values__cache_label_values_lookup/ntimelines/4 time: [7.8878 ns 7.9401 ns 8.0034 ns]
label_values__cache_label_values_lookup/ntimelines/8 time: [7.2621 ns 7.6354 ns 8.0337 ns]
single_metric_multicore_scalability/nthreads/1 time: [5.7710 ns 5.7744 ns 5.7785 ns]
single_metric_multicore_scalability/nthreads/4 time: [66.629 ns 66.994 ns 67.336 ns]
single_metric_multicore_scalability/nthreads/8 time: [130.85 ns 131.98 ns 132.91 ns]
propagation_of_cached_label_value__naive/nthreads/1 time: [11.540 ns 11.546 ns 11.553 ns]
propagation_of_cached_label_value__naive/nthreads/4 time: [131.22 ns 131.90 ns 132.56 ns]
propagation_of_cached_label_value__naive/nthreads/8 time: [260.99 ns 262.75 ns 264.26 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time: [11.544 ns 11.550 ns 11.557 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time: [11.568 ns 11.642 ns 11.763 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time: [13.416 ns 14.121 ns 14.886 ns
Results on an M4 MAX MacBook Pro Total Number of Cores: 14 (10 performance and 4 efficiency)
label_values__naive_usage/ntimelines/1 time: [52.711 ns 53.026 ns 53.381 ns]
label_values__naive_usage/ntimelines/4 time: [323.99 ns 330.40 ns 337.53 ns]
label_values__naive_usage/ntimelines/8 time: [1.1615 µs 1.1998 µs 1.2399 µs]
label_values__cache_label_values_lookup/ntimelines/1 time: [1.6635 ns 1.6715 ns 1.6809 ns]
label_values__cache_label_values_lookup/ntimelines/4 time: [1.7786 ns 1.7876 ns 1.8028 ns]
label_values__cache_label_values_lookup/ntimelines/8 time: [1.8195 ns 1.8371 ns 1.8665 ns]
single_metric_multicore_scalability/nthreads/1 time: [1.7764 ns 1.7909 ns 1.8079 ns]
single_metric_multicore_scalability/nthreads/4 time: [33.875 ns 34.868 ns 35.923 ns]
single_metric_multicore_scalability/nthreads/8 time: [226.85 ns 235.30 ns 244.18 ns]
propagation_of_cached_label_value__naive/nthreads/1 time: [3.4337 ns 3.4491 ns 3.4660 ns]
propagation_of_cached_label_value__naive/nthreads/4 time: [69.486 ns 71.937 ns 74.472 ns]
propagation_of_cached_label_value__naive/nthreads/8 time: [434.87 ns 456.47 ns 477.84 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time: [3.3767 ns 3.3974 ns 3.4220 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time: [3.6105 ns 4.2355 ns 5.1463 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time: [4.0889 ns 4.9714 ns 6.0779 ns]
Results on a Hetzner AX102 AMD Ryzen 9 7950X3D 16-Core Processor
label_values__naive_usage/ntimelines/1 time: [64.510 ns 64.559 ns 64.610 ns]
label_values__naive_usage/ntimelines/4 time: [309.71 ns 326.09 ns 342.32 ns]
label_values__naive_usage/ntimelines/8 time: [776.92 ns 819.35 ns 856.93 ns]
label_values__cache_label_values_lookup/ntimelines/1 time: [1.2855 ns 1.2943 ns 1.3021 ns]
label_values__cache_label_values_lookup/ntimelines/4 time: [1.3865 ns 1.4139 ns 1.4441 ns]
label_values__cache_label_values_lookup/ntimelines/8 time: [1.5311 ns 1.5669 ns 1.6046 ns]
single_metric_multicore_scalability/nthreads/1 time: [1.1927 ns 1.1981 ns 1.2049 ns]
single_metric_multicore_scalability/nthreads/4 time: [24.346 ns 25.439 ns 26.634 ns]
single_metric_multicore_scalability/nthreads/8 time: [58.666 ns 60.137 ns 61.486 ns]
propagation_of_cached_label_value__naive/nthreads/1 time: [2.7067 ns 2.7238 ns 2.7402 ns]
propagation_of_cached_label_value__naive/nthreads/4 time: [62.723 ns 66.214 ns 69.787 ns]
propagation_of_cached_label_value__naive/nthreads/8 time: [164.24 ns 170.10 ns 175.68 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time: [2.2915 ns 2.2960 ns 2.3012 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time: [2.5726 ns 2.6158 ns 2.6624 ns]
propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time: [2.7068 ns 2.8243 ns 2.9824 ns]
*/

View File

@@ -7,7 +7,7 @@ use http_utils::error::HttpErrorBody;
use pageserver_api::models::*;
use pageserver_api::shard::TenantShardId;
pub use reqwest::Body as ReqwestBody;
use reqwest::{Certificate, IntoUrl, Method, StatusCode, Url};
use reqwest::{Certificate, IntoUrl, Method, StatusCode};
use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
@@ -458,21 +458,30 @@ impl Client {
&self,
tenant_shard_id: TenantShardId,
timeline_id: TimelineId,
behavior: Option<DetachBehavior>,
) -> Result<AncestorDetached> {
let uri = format!(
"{}/v1/tenant/{tenant_shard_id}/timeline/{timeline_id}/detach_ancestor",
self.mgmt_api_endpoint
);
let mut uri = Url::parse(&uri)
.map_err(|e| Error::ApiError(StatusCode::INTERNAL_SERVER_ERROR, format!("{e}")))?;
if let Some(behavior) = behavior {
uri.query_pairs_mut()
.append_pair("detach_behavior", &behavior.to_string());
}
self.request(Method::PUT, &uri, ())
.await?
.json()
.await
.map_err(Error::ReceiveBody)
}
self.request(Method::PUT, uri, ())
pub async fn timeline_detach_ancestor_v2(
&self,
tenant_shard_id: TenantShardId,
timeline_id: TimelineId,
) -> Result<AncestorDetached> {
let uri = format!(
"{}/v1/tenant/{tenant_shard_id}/timeline/{timeline_id}/detach_ancestor_v2",
self.mgmt_api_endpoint
);
self.request(Method::PUT, &uri, ())
.await?
.json()
.await

View File

@@ -12,7 +12,7 @@ pub(crate) fn setup_logging() {
logging::TracingErrorLayerEnablement::EnableWithRustLogFilter,
logging::Output::Stdout,
)
.expect("Failed to init test logging");
.expect("Failed to init test logging")
});
}

View File

@@ -30,6 +30,7 @@ use pageserver::{
};
use postgres_backend::AuthType;
use remote_storage::GenericRemoteStorage;
use rustls_pki_types::{CertificateDer, PrivateKeyDer};
use tokio::signal::unix::SignalKind;
use tokio::time::Instant;
use tokio_util::sync::CancellationToken;
@@ -110,7 +111,6 @@ fn main() -> anyhow::Result<()> {
} else {
TracingErrorLayerEnablement::Disabled
};
logging::init(
conf.log_format,
tracing_error_layer_enablement,
@@ -621,8 +621,8 @@ fn start_pageserver(
let https_task = match https_listener {
Some(https_listener) => {
let certs = http_utils::tls_certs::load_cert_chain(&conf.ssl_cert_file)?;
let key = http_utils::tls_certs::load_private_key(&conf.ssl_key_file)?;
let certs = load_certs(&conf.ssl_cert_file)?;
let key = load_private_key(&conf.ssl_key_file)?;
let server_config = rustls::ServerConfig::builder()
.with_no_client_auth()
@@ -734,6 +734,25 @@ fn start_pageserver(
})
}
fn load_certs(filename: &Utf8Path) -> std::io::Result<Vec<CertificateDer<'static>>> {
let file = std::fs::File::open(filename)?;
let mut reader = std::io::BufReader::new(file);
rustls_pemfile::certs(&mut reader).collect()
}
fn load_private_key(filename: &Utf8Path) -> anyhow::Result<PrivateKeyDer<'static>> {
let file = std::fs::File::open(filename)?;
let mut reader = std::io::BufReader::new(file);
let key = rustls_pemfile::private_key(&mut reader)?;
key.ok_or(anyhow::anyhow!(
"no private key found in {}",
filename.as_str(),
))
}
async fn create_remote_storage_client(
conf: &'static PageServerConf,
) -> anyhow::Result<GenericRemoteStorage> {

View File

@@ -89,112 +89,16 @@
//! [`RequestContext`] argument. Functions in the middle of the call chain
//! only need to pass it on.
use std::sync::Arc;
use once_cell::sync::Lazy;
use tracing::warn;
use utils::{id::TimelineId, shard::TenantShardId};
use crate::{
metrics::{StorageIoSizeMetrics, TimelineMetrics},
task_mgr::TaskKind,
tenant::Timeline,
};
use crate::task_mgr::TaskKind;
// The main structure of this module, see module-level comment.
#[derive(Debug)]
pub struct RequestContext {
task_kind: TaskKind,
download_behavior: DownloadBehavior,
access_stats_behavior: AccessStatsBehavior,
page_content_kind: PageContentKind,
read_path_debug: bool,
scope: Scope,
}
#[derive(Clone)]
pub(crate) enum Scope {
Global {
io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
},
SecondaryTenant {
io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
},
SecondaryTimeline {
io_size_metrics: crate::metrics::StorageIoSizeMetrics,
},
Timeline {
// We wrap the `Arc<TimelineMetrics>`s inside another Arc to avoid child
// context creation contending for the ref counters of the Arc<TimelineMetrics>,
// which are shared among all tasks that operate on the timeline, especially
// concurrent page_service connections.
#[allow(clippy::redundant_allocation)]
arc_arc: Arc<Arc<TimelineMetrics>>,
},
#[cfg(test)]
UnitTest {
io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
},
}
static GLOBAL_IO_SIZE_METRICS: Lazy<crate::metrics::StorageIoSizeMetrics> =
Lazy::new(|| crate::metrics::StorageIoSizeMetrics::new("*", "*", "*"));
impl Scope {
pub(crate) fn new_global() -> Self {
Scope::Global {
io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
}
}
/// NB: this allocates, so, use only at relatively long-lived roots, e.g., at start
/// of a compaction iteration.
pub(crate) fn new_timeline(timeline: &Timeline) -> Self {
Scope::Timeline {
arc_arc: Arc::new(Arc::clone(&timeline.metrics)),
}
}
pub(crate) fn new_page_service_pagestream(
timeline_handle: &crate::tenant::timeline::handle::Handle<
crate::page_service::TenantManagerTypes,
>,
) -> Self {
Scope::Timeline {
arc_arc: Arc::clone(&timeline_handle.metrics),
}
}
pub(crate) fn new_secondary_timeline(
tenant_shard_id: &TenantShardId,
timeline_id: &TimelineId,
) -> Self {
// TODO(https://github.com/neondatabase/neon/issues/11156): secondary timelines have no infrastructure for metrics lifecycle.
let tenant_id = tenant_shard_id.tenant_id.to_string();
let shard_id = tenant_shard_id.shard_slug().to_string();
let timeline_id = timeline_id.to_string();
let io_size_metrics =
crate::metrics::StorageIoSizeMetrics::new(&tenant_id, &shard_id, &timeline_id);
Scope::SecondaryTimeline { io_size_metrics }
}
pub(crate) fn new_secondary_tenant(_tenant_shard_id: &TenantShardId) -> Self {
// Before propagating metrics via RequestContext, the labels were inferred from file path.
// The only user of VirtualFile at tenant scope is the heatmap download & read.
// The inferred labels for the path of the heatmap file on local disk were that of the global metric (*,*,*).
// Thus, we do the same here, and extend that for anything secondary-tenant scoped.
//
// If we want to have (tenant_id, shard_id, '*') labels for secondary tenants in the future,
// we will need to think about the metric lifecycle, i.e., remove them during secondary tenant shutdown,
// like we do for attached timelines. (We don't have attached-tenant-scoped usage of VirtualFile
// at this point, so, we were able to completely side-step tenant-scoped stuff there).
Scope::SecondaryTenant {
io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
}
}
#[cfg(test)]
pub(crate) fn new_unit_test() -> Self {
Scope::UnitTest {
io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
}
}
}
/// The kind of access to the page cache.
@@ -253,7 +157,6 @@ impl RequestContextBuilder {
access_stats_behavior: AccessStatsBehavior::Update,
page_content_kind: PageContentKind::Unknown,
read_path_debug: false,
scope: Scope::new_global(),
},
}
}
@@ -268,16 +171,10 @@ impl RequestContextBuilder {
access_stats_behavior: original.access_stats_behavior,
page_content_kind: original.page_content_kind,
read_path_debug: original.read_path_debug,
scope: original.scope.clone(),
},
}
}
pub fn task_kind(mut self, k: TaskKind) -> Self {
self.inner.task_kind = k;
self
}
/// Configure the DownloadBehavior of the context: whether to
/// download missing layers, and/or warn on the download.
pub fn download_behavior(mut self, b: DownloadBehavior) -> Self {
@@ -302,11 +199,6 @@ impl RequestContextBuilder {
self
}
pub(crate) fn scope(mut self, s: Scope) -> Self {
self.inner.scope = s;
self
}
pub fn build(self) -> RequestContext {
self.inner
}
@@ -389,50 +281,7 @@ impl RequestContext {
}
fn child_impl(&self, task_kind: TaskKind, download_behavior: DownloadBehavior) -> Self {
RequestContextBuilder::extend(self)
.task_kind(task_kind)
.download_behavior(download_behavior)
.build()
}
pub fn with_scope_timeline(&self, timeline: &Arc<Timeline>) -> Self {
RequestContextBuilder::extend(self)
.scope(Scope::new_timeline(timeline))
.build()
}
pub(crate) fn with_scope_page_service_pagestream(
&self,
timeline_handle: &crate::tenant::timeline::handle::Handle<
crate::page_service::TenantManagerTypes,
>,
) -> Self {
RequestContextBuilder::extend(self)
.scope(Scope::new_page_service_pagestream(timeline_handle))
.build()
}
pub fn with_scope_secondary_timeline(
&self,
tenant_shard_id: &TenantShardId,
timeline_id: &TimelineId,
) -> Self {
RequestContextBuilder::extend(self)
.scope(Scope::new_secondary_timeline(tenant_shard_id, timeline_id))
.build()
}
pub fn with_scope_secondary_tenant(&self, tenant_shard_id: &TenantShardId) -> Self {
RequestContextBuilder::extend(self)
.scope(Scope::new_secondary_tenant(tenant_shard_id))
.build()
}
#[cfg(test)]
pub fn with_scope_unit_test(&self) -> Self {
RequestContextBuilder::new(TaskKind::UnitTest)
.scope(Scope::new_unit_test())
.build()
Self::new(task_kind, download_behavior)
}
pub fn task_kind(&self) -> TaskKind {
@@ -454,38 +303,4 @@ impl RequestContext {
pub(crate) fn read_path_debug(&self) -> bool {
self.read_path_debug
}
pub(crate) fn io_size_metrics(&self) -> &StorageIoSizeMetrics {
match &self.scope {
Scope::Global { io_size_metrics } => {
let is_unit_test = cfg!(test);
let is_regress_test_build = cfg!(feature = "testing");
if is_unit_test || is_regress_test_build {
panic!("all VirtualFile instances are timeline-scoped");
} else {
use once_cell::sync::Lazy;
use std::sync::Mutex;
use std::time::Duration;
use utils::rate_limit::RateLimit;
static LIMIT: Lazy<Mutex<RateLimit>> =
Lazy::new(|| Mutex::new(RateLimit::new(Duration::from_secs(1))));
let mut guard = LIMIT.lock().unwrap();
guard.call2(|rate_limit_stats| {
warn!(
%rate_limit_stats,
backtrace=%std::backtrace::Backtrace::force_capture(),
"all VirtualFile instances are timeline-scoped",
);
});
io_size_metrics
}
}
Scope::Timeline { arc_arc } => &arc_arc.storage_io_size,
Scope::SecondaryTimeline { io_size_metrics } => io_size_metrics,
Scope::SecondaryTenant { io_size_metrics } => io_size_metrics,
#[cfg(test)]
Scope::UnitTest { io_size_metrics } => io_size_metrics,
}
}
}

View File

@@ -1079,6 +1079,7 @@ components:
- last_record_lsn
- disk_consistent_lsn
- state
- latest_gc_cutoff_lsn
properties:
timeline_id:
type: string
@@ -1122,6 +1123,9 @@ components:
min_readable_lsn:
type: string
format: hex
latest_gc_cutoff_lsn:
type: string
format: hex
applied_gc_cutoff_lsn:
type: string
format: hex

View File

@@ -28,9 +28,9 @@ use hyper::{Body, Request, Response, StatusCode, Uri, header};
use metrics::launch_timestamp::LaunchTimestamp;
use pageserver_api::models::virtual_file::IoMode;
use pageserver_api::models::{
DetachBehavior, DownloadRemoteLayersTaskSpawnRequest, IngestAuxFilesRequest,
ListAuxFilesRequest, LocationConfig, LocationConfigListResponse, LocationConfigMode, LsnLease,
LsnLeaseRequest, OffloadedTimelineInfo, PageTraceEvent, ShardParameters, StatusResponse,
DownloadRemoteLayersTaskSpawnRequest, IngestAuxFilesRequest, ListAuxFilesRequest,
LocationConfig, LocationConfigListResponse, LocationConfigMode, LsnLease, LsnLeaseRequest,
OffloadedTimelineInfo, PageTraceEvent, ShardParameters, StatusResponse,
TenantConfigPatchRequest, TenantConfigRequest, TenantDetails, TenantInfo,
TenantLocationConfigRequest, TenantLocationConfigResponse, TenantScanRemoteStorageResponse,
TenantScanRemoteStorageShard, TenantShardLocation, TenantShardSplitRequest,
@@ -55,7 +55,6 @@ use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
use crate::config::PageServerConf;
use crate::context;
use crate::context::{DownloadBehavior, RequestContext, RequestContextBuilder};
use crate::deletion_queue::DeletionQueueClient;
use crate::pgdatadir_mapping::LsnForTimestamp;
@@ -72,6 +71,7 @@ use crate::tenant::remote_timeline_client::{
use crate::tenant::secondary::SecondaryController;
use crate::tenant::size::ModelInputs;
use crate::tenant::storage_layer::{IoConcurrency, LayerAccessStatsReset, LayerName};
use crate::tenant::timeline::detach_ancestor::DetachBehavior;
use crate::tenant::timeline::offload::{OffloadError, offload_timeline};
use crate::tenant::timeline::{
CompactFlags, CompactOptions, CompactRequest, CompactionError, Timeline, WaitLsnTimeout,
@@ -460,7 +460,10 @@ async fn build_timeline_info_common(
initdb_lsn,
last_record_lsn,
prev_record_lsn: Some(timeline.get_prev_record_lsn()),
_unused: Default::default(), // Unused, for legacy decode only
// Externally, expose the lowest LSN that can be used to create a branch as the "GC cutoff", although internally
// we distinguish between the "planned" GC cutoff (PITR point) and the "latest" GC cutoff (where we
// actually trimmed data to), which can pass each other when PITR is changed.
latest_gc_cutoff_lsn: min_readable_lsn,
min_readable_lsn,
applied_gc_cutoff_lsn: *timeline.get_applied_gc_cutoff_lsn(),
current_logical_size: current_logical_size.size_dont_care_about_accuracy(),
@@ -951,13 +954,12 @@ async fn timeline_detail_handler(
tenant.wait_to_become_active(ACTIVE_TENANT_TIMEOUT).await?;
let timeline = tenant.get_timeline(timeline_id, false)?;
let ctx = &ctx.with_scope_timeline(&timeline);
let timeline_info = build_timeline_info(
&timeline,
include_non_incremental_logical_size.unwrap_or(false),
force_await_initial_logical_size.unwrap_or(false),
ctx,
&ctx,
)
.await
.context("get local timeline info")
@@ -1001,8 +1003,7 @@ async fn get_lsn_by_timestamp_handler(
let timeline =
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let result = timeline
.find_lsn_for_timestamp(timestamp_pg, &cancel, &ctx)
.await?;
@@ -1074,8 +1075,7 @@ async fn get_timestamp_of_lsn_handler(
let timeline =
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let result = timeline.get_timestamp_for_lsn(lsn, &ctx).await?;
match result {
@@ -1430,8 +1430,7 @@ async fn timeline_layer_scan_disposable_keys(
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let guard = timeline.layers.read().await;
let Some(layer) = guard.try_get_from_key(&layer_name.clone().into()) else {
@@ -1517,8 +1516,7 @@ async fn timeline_download_heatmap_layers_handler(
let timeline =
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let max_concurrency = get_config(&request)
.remote_storage_config
@@ -1566,8 +1564,7 @@ async fn layer_download_handler(
let timeline =
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let downloaded = timeline
.download_layer(&layer_name, &ctx)
.await
@@ -2303,8 +2300,8 @@ async fn timeline_compact_handler(
.unwrap_or(false);
async {
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
if scheduled {
let tenant = state
.tenant_manager
@@ -2391,7 +2388,6 @@ async fn timeline_checkpoint_handler(
let state = get_state(&request);
let mut flags = EnumSet::empty();
flags |= CompactFlags::NoYield; // run compaction to completion
if Some(true) == parse_query_param::<_, bool>(&request, "force_l0_compaction")? {
flags |= CompactFlags::ForceL0Compaction;
}
@@ -2412,8 +2408,8 @@ async fn timeline_checkpoint_handler(
parse_query_param::<_, bool>(&request, "wait_until_uploaded")?.unwrap_or(false);
async {
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
if wait_until_flushed {
timeline.freeze_and_flush().await
} else {
@@ -2468,8 +2464,7 @@ async fn timeline_download_remote_layers_handler_post(
let timeline =
active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
.await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
.with_scope_timeline(&timeline);
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
match timeline.spawn_download_all_remote_layers(body, &ctx).await {
Ok(st) => json_response(StatusCode::ACCEPTED, st),
Err(st) => json_response(StatusCode::CONFLICT, st),
@@ -2495,8 +2490,29 @@ async fn timeline_download_remote_layers_handler_get(
json_response(StatusCode::OK, info)
}
async fn timeline_detach_ancestor_handler(
async fn timeline_detach_ancestor_handler_v1(
request: Request<Body>,
cancel: CancellationToken,
) -> Result<Response<Body>, ApiError> {
timeline_detach_ancestor_handler_common(request, DetachBehavior::NoAncestorAndReparent, cancel)
.await
}
async fn timeline_detach_ancestor_handler_v2(
request: Request<Body>,
cancel: CancellationToken,
) -> Result<Response<Body>, ApiError> {
timeline_detach_ancestor_handler_common(
request,
DetachBehavior::MultiLevelAndNoReparent,
cancel,
)
.await
}
async fn timeline_detach_ancestor_handler_common(
request: Request<Body>,
behavior: DetachBehavior,
_cancel: CancellationToken,
) -> Result<Response<Body>, ApiError> {
use pageserver_api::models::detach_ancestor::AncestorDetached;
@@ -2506,9 +2522,6 @@ async fn timeline_detach_ancestor_handler(
let tenant_shard_id: TenantShardId = parse_request_param(&request, "tenant_shard_id")?;
check_permission(&request, Some(tenant_shard_id.tenant_id))?;
let timeline_id: TimelineId = parse_request_param(&request, "timeline_id")?;
let behavior: Option<DetachBehavior> = parse_query_param(&request, "detach_behavior")?;
let behavior = behavior.unwrap_or_default();
let span = tracing::info_span!("detach_ancestor", tenant_id=%tenant_shard_id.tenant_id, shard_id=%tenant_shard_id.shard_slug(), %timeline_id);
@@ -2555,7 +2568,6 @@ async fn timeline_detach_ancestor_handler(
tracing::info!("all timeline upload queues are drained");
let timeline = tenant.get_timeline(timeline_id, true)?;
let ctx = &ctx.with_scope_timeline(&timeline);
let progress = timeline
.prepare_to_detach_from_ancestor(&tenant, options, behavior, ctx)
@@ -2663,9 +2675,8 @@ async fn getpage_at_lsn_handler_inner(
async {
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
// Enable read path debugging
let ctx = RequestContextBuilder::extend(&ctx).read_path_debug(true).build();
let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
let ctx = RequestContextBuilder::extend(&ctx).read_path_debug(true)
.scope(context::Scope::new_timeline(&timeline)).build();
// Use last_record_lsn if no lsn is provided
let lsn = lsn.unwrap_or_else(|| timeline.get_last_record_lsn());
@@ -2699,8 +2710,8 @@ async fn timeline_collect_keyspace(
let at_lsn: Option<Lsn> = parse_query_param(&request, "at_lsn")?;
async {
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
let at_lsn = at_lsn.unwrap_or_else(|| timeline.get_last_record_lsn());
let (dense_ks, sparse_ks) = timeline
.collect_keyspace(at_lsn, &ctx)
@@ -3225,7 +3236,6 @@ async fn post_top_tenants(
match order_by {
TenantSorting::ResidentSize => sizes.resident_size,
TenantSorting::MaxLogicalSize => sizes.max_logical_size,
TenantSorting::MaxLogicalSizePerShard => sizes.max_logical_size_per_shard,
}
}
@@ -3338,7 +3348,7 @@ async fn put_tenant_timeline_import_basebackup(
tenant.wait_to_become_active(ACTIVE_TENANT_TIMEOUT).await?;
let (timeline, timeline_ctx) = tenant
let timeline = tenant
.create_empty_timeline(timeline_id, base_lsn, pg_version, &ctx)
.map_err(ApiError::InternalServerError)
.await?;
@@ -3357,13 +3367,7 @@ async fn put_tenant_timeline_import_basebackup(
info!("importing basebackup");
timeline
.import_basebackup_from_tar(
tenant.clone(),
&mut body,
base_lsn,
broker_client,
&timeline_ctx,
)
.import_basebackup_from_tar(tenant.clone(), &mut body, base_lsn, broker_client, &ctx)
.await
.map_err(ApiError::InternalServerError)?;
@@ -3403,7 +3407,6 @@ async fn put_tenant_timeline_import_wal(
let state = get_state(&request);
let timeline = active_timeline_of_active_tenant(&state.tenant_manager, TenantShardId::unsharded(tenant_id), timeline_id).await?;
let ctx = RequestContextBuilder::extend(&ctx).scope(context::Scope::new_timeline(&timeline)).build();
let mut body = StreamReader::new(request.into_body().map(|res| {
res.map_err(|error| {
@@ -3762,7 +3765,11 @@ pub fn make_router(
)
.put(
"/v1/tenant/:tenant_shard_id/timeline/:timeline_id/detach_ancestor",
|r| api_handler(r, timeline_detach_ancestor_handler),
|r| api_handler(r, timeline_detach_ancestor_handler_v1),
)
.put(
"/v1/tenant/:tenant_shard_id/timeline/:timeline_id/detach_ancestor_v2",
|r| api_handler(r, timeline_detach_ancestor_handler_v2),
)
.delete("/v1/tenant/:tenant_shard_id/timeline/:timeline_id", |r| {
api_handler(r, timeline_delete_handler)

View File

@@ -465,40 +465,12 @@ pub(crate) fn page_cache_errors_inc(error_kind: PageCacheErrorKind) {
pub(crate) static WAIT_LSN_TIME: Lazy<Histogram> = Lazy::new(|| {
register_histogram!(
"pageserver_wait_lsn_seconds",
"Time spent waiting for WAL to arrive. Updated on completion of the wait_lsn operation.",
"Time spent waiting for WAL to arrive",
CRITICAL_OP_BUCKETS.into(),
)
.expect("failed to define a metric")
});
pub(crate) static WAIT_LSN_START_FINISH_COUNTERPAIR: Lazy<IntCounterPairVec> = Lazy::new(|| {
register_int_counter_pair_vec!(
"pageserver_wait_lsn_started_count",
"Number of wait_lsn operations started.",
"pageserver_wait_lsn_finished_count",
"Number of wait_lsn operations finished.",
&["tenant_id", "shard_id", "timeline_id"],
)
.expect("failed to define a metric")
});
pub(crate) static WAIT_LSN_IN_PROGRESS_MICROS: Lazy<IntCounterVec> = Lazy::new(|| {
register_int_counter_vec!(
"pageserver_wait_lsn_in_progress_micros",
"Time spent waiting for WAL to arrive, by timeline_id. Updated periodically while waiting.",
&["tenant_id", "shard_id", "timeline_id"],
)
.expect("failed to define a metric")
});
pub(crate) static WAIT_LSN_IN_PROGRESS_GLOBAL_MICROS: Lazy<IntCounter> = Lazy::new(|| {
register_int_counter!(
"pageserver_wait_lsn_in_progress_micros_global",
"Time spent waiting for WAL to arrive, globally. Updated periodically while waiting."
)
.expect("failed to define a metric")
});
static FLUSH_WAIT_UPLOAD_TIME: Lazy<GaugeVec> = Lazy::new(|| {
register_gauge_vec!(
"pageserver_flush_wait_upload_seconds",
@@ -1255,24 +1227,11 @@ impl StorageIoTime {
pub(crate) static STORAGE_IO_TIME_METRIC: Lazy<StorageIoTime> = Lazy::new(StorageIoTime::new);
#[derive(Clone, Copy)]
#[repr(usize)]
enum StorageIoSizeOperation {
Read,
Write,
}
impl StorageIoSizeOperation {
const VARIANTS: &'static [&'static str] = &["read", "write"];
fn as_str(&self) -> &'static str {
Self::VARIANTS[*self as usize]
}
}
const STORAGE_IO_SIZE_OPERATIONS: &[&str] = &["read", "write"];
// Needed for the https://neonprod.grafana.net/d/5uK9tHL4k/picking-tenant-for-relocation?orgId=1
static STORAGE_IO_SIZE: Lazy<UIntGaugeVec> = Lazy::new(|| {
register_uint_gauge_vec!(
pub(crate) static STORAGE_IO_SIZE: Lazy<IntGaugeVec> = Lazy::new(|| {
register_int_gauge_vec!(
"pageserver_io_operations_bytes_total",
"Total amount of bytes read/written in IO operations",
&["operation", "tenant_id", "shard_id", "timeline_id"]
@@ -1280,34 +1239,6 @@ static STORAGE_IO_SIZE: Lazy<UIntGaugeVec> = Lazy::new(|| {
.expect("failed to define a metric")
});
#[derive(Clone, Debug)]
pub(crate) struct StorageIoSizeMetrics {
pub read: UIntGauge,
pub write: UIntGauge,
}
impl StorageIoSizeMetrics {
pub(crate) fn new(tenant_id: &str, shard_id: &str, timeline_id: &str) -> Self {
let read = STORAGE_IO_SIZE
.get_metric_with_label_values(&[
StorageIoSizeOperation::Read.as_str(),
tenant_id,
shard_id,
timeline_id,
])
.unwrap();
let write = STORAGE_IO_SIZE
.get_metric_with_label_values(&[
StorageIoSizeOperation::Write.as_str(),
tenant_id,
shard_id,
timeline_id,
])
.unwrap();
Self { read, write }
}
}
#[cfg(not(test))]
pub(crate) mod virtual_file_descriptor_cache {
use super::*;
@@ -2858,6 +2789,7 @@ impl StorageTimeMetrics {
}
}
#[derive(Debug)]
pub(crate) struct TimelineMetrics {
tenant_id: String,
shard_id: String,
@@ -2889,9 +2821,6 @@ pub(crate) struct TimelineMetrics {
/// Number of valid LSN leases.
pub valid_lsn_lease_count_gauge: UIntGauge,
pub wal_records_received: IntCounter,
pub storage_io_size: StorageIoSizeMetrics,
pub wait_lsn_in_progress_micros: GlobalAndPerTenantIntCounter,
pub wait_lsn_start_finish_counterpair: IntCounterPair,
shutdown: std::sync::atomic::AtomicBool,
}
@@ -3027,19 +2956,6 @@ impl TimelineMetrics {
.get_metric_with_label_values(&[&tenant_id, &shard_id, &timeline_id])
.unwrap();
let storage_io_size = StorageIoSizeMetrics::new(&tenant_id, &shard_id, &timeline_id);
let wait_lsn_in_progress_micros = GlobalAndPerTenantIntCounter {
global: WAIT_LSN_IN_PROGRESS_GLOBAL_MICROS.clone(),
per_tenant: WAIT_LSN_IN_PROGRESS_MICROS
.get_metric_with_label_values(&[&tenant_id, &shard_id, &timeline_id])
.unwrap(),
};
let wait_lsn_start_finish_counterpair = WAIT_LSN_START_FINISH_COUNTERPAIR
.get_metric_with_label_values(&[&tenant_id, &shard_id, &timeline_id])
.unwrap();
TimelineMetrics {
tenant_id,
shard_id,
@@ -3069,11 +2985,8 @@ impl TimelineMetrics {
evictions_with_low_residence_duration: std::sync::RwLock::new(
evictions_with_low_residence_duration,
),
storage_io_size,
valid_lsn_lease_count_gauge,
wal_records_received,
wait_lsn_in_progress_micros,
wait_lsn_start_finish_counterpair,
shutdown: std::sync::atomic::AtomicBool::default(),
}
}
@@ -3262,19 +3175,10 @@ impl TimelineMetrics {
]);
}
for op in StorageIoSizeOperation::VARIANTS {
for op in STORAGE_IO_SIZE_OPERATIONS {
let _ = STORAGE_IO_SIZE.remove_label_values(&[op, tenant_id, shard_id, timeline_id]);
}
let _ =
WAIT_LSN_IN_PROGRESS_MICROS.remove_label_values(&[tenant_id, shard_id, timeline_id]);
{
let mut res = [Ok(()), Ok(())];
WAIT_LSN_START_FINISH_COUNTERPAIR
.remove_label_values(&mut res, &[tenant_id, shard_id, timeline_id]);
}
let _ = SMGR_QUERY_STARTED_PER_TENANT_TIMELINE.remove_label_values(&[
SmgrQueryType::GetPageAtLsn.into(),
tenant_id,
@@ -3887,29 +3791,27 @@ pub mod tokio_epoll_uring {
});
}
pub(crate) struct GlobalAndPerTenantIntCounter {
global: IntCounter,
per_tenant: IntCounter,
}
impl GlobalAndPerTenantIntCounter {
#[inline(always)]
pub(crate) fn inc(&self) {
self.inc_by(1)
}
#[inline(always)]
pub(crate) fn inc_by(&self, n: u64) {
self.global.inc_by(n);
self.per_tenant.inc_by(n);
}
}
pub(crate) mod tenant_throttling {
use metrics::register_int_counter_vec;
use metrics::{IntCounter, register_int_counter_vec};
use once_cell::sync::Lazy;
use utils::shard::TenantShardId;
use super::GlobalAndPerTenantIntCounter;
pub(crate) struct GlobalAndPerTenantIntCounter {
global: IntCounter,
per_tenant: IntCounter,
}
impl GlobalAndPerTenantIntCounter {
#[inline(always)]
pub(crate) fn inc(&self) {
self.inc_by(1)
}
#[inline(always)]
pub(crate) fn inc_by(&self, n: u64) {
self.global.inc_by(n);
self.per_tenant.inc_by(n);
}
}
pub(crate) struct Metrics<const KIND: usize> {
pub(super) count_accounted_start: GlobalAndPerTenantIntCounter,
@@ -4155,7 +4057,6 @@ pub fn preinitialize_metrics(conf: &'static PageServerConf) {
&CIRCUIT_BREAKERS_BROKEN,
&CIRCUIT_BREAKERS_UNBROKEN,
&PAGE_SERVICE_SMGR_FLUSH_INPROGRESS_MICROS_GLOBAL,
&WAIT_LSN_IN_PROGRESS_GLOBAL_MICROS,
]
.into_iter()
.for_each(|c| {

View File

@@ -56,7 +56,6 @@ use crate::config::PageServerConf;
use crate::context::{DownloadBehavior, RequestContext};
use crate::metrics::{
self, COMPUTE_COMMANDS_COUNTERS, ComputeCommandKind, LIVE_CONNECTIONS, SmgrOpTimer,
TimelineMetrics,
};
use crate::pgdatadir_mapping::Version;
use crate::span::{
@@ -424,9 +423,6 @@ impl timeline::handle::Types for TenantManagerTypes {
pub(crate) struct TenantManagerCacheItem {
pub(crate) timeline: Arc<Timeline>,
// allow() for cheap propagation through RequestContext inside a task
#[allow(clippy::redundant_allocation)]
pub(crate) metrics: Arc<Arc<TimelineMetrics>>,
#[allow(dead_code)] // we store it to keep the gate open
pub(crate) gate_guard: GateGuard,
}
@@ -510,11 +506,8 @@ impl timeline::handle::TenantManager<TenantManagerTypes> for TenantManagerWrappe
}
};
let metrics = Arc::new(Arc::clone(&timeline.metrics));
Ok(TenantManagerCacheItem {
timeline,
metrics,
gate_guard,
})
}
@@ -1106,19 +1099,12 @@ impl PageServerHandler {
};
// Dispatch the batch to the appropriate request handler.
let log_slow_name = batch.as_static_str();
let (mut handler_results, span) = {
// TODO: we unfortunately have to pin the future on the heap, since GetPage futures are huge and
// won't fit on the stack.
let mut boxpinned =
Box::pin(self.pagestream_dispatch_batched_message(batch, io_concurrency, ctx));
log_slow(
log_slow_name,
LOG_SLOW_GETPAGE_THRESHOLD,
boxpinned.as_mut(),
)
.await?
};
let (mut handler_results, span) = log_slow(
batch.as_static_str(),
LOG_SLOW_GETPAGE_THRESHOLD,
self.pagestream_dispatch_batched_message(batch, io_concurrency, ctx),
)
.await?;
// We purposefully don't count flush time into the smgr operation timer.
//
@@ -1252,14 +1238,6 @@ impl PageServerHandler {
),
QueryError,
> {
macro_rules! upgrade_handle_and_set_context {
($shard:ident) => {{
let weak_handle = &$shard;
let handle = weak_handle.upgrade()?;
let ctx = ctx.with_scope_page_service_pagestream(&handle);
(handle, ctx)
}};
}
Ok(match batch {
BatchedFeMessage::Exists {
span,
@@ -1268,10 +1246,9 @@ impl PageServerHandler {
req,
} => {
fail::fail_point!("ps::handle-pagerequest-message::exists");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
vec![
self.handle_get_rel_exists_request(&shard, &req, &ctx)
self.handle_get_rel_exists_request(&*shard.upgrade()?, &req, ctx)
.instrument(span.clone())
.await
.map(|msg| (msg, timer))
@@ -1287,10 +1264,9 @@ impl PageServerHandler {
req,
} => {
fail::fail_point!("ps::handle-pagerequest-message::nblocks");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
vec![
self.handle_get_nblocks_request(&shard, &req, &ctx)
self.handle_get_nblocks_request(&*shard.upgrade()?, &req, ctx)
.instrument(span.clone())
.await
.map(|msg| (msg, timer))
@@ -1306,18 +1282,17 @@ impl PageServerHandler {
pages,
} => {
fail::fail_point!("ps::handle-pagerequest-message::getpage");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
{
let npages = pages.len();
trace!(npages, "handling getpage request");
let res = self
.handle_get_page_at_lsn_request_batched(
&shard,
&*shard.upgrade()?,
effective_request_lsn,
pages,
io_concurrency,
&ctx,
ctx,
)
.instrument(span.clone())
.await;
@@ -1334,10 +1309,9 @@ impl PageServerHandler {
req,
} => {
fail::fail_point!("ps::handle-pagerequest-message::dbsize");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
vec![
self.handle_db_size_request(&shard, &req, &ctx)
self.handle_db_size_request(&*shard.upgrade()?, &req, ctx)
.instrument(span.clone())
.await
.map(|msg| (msg, timer))
@@ -1353,10 +1327,9 @@ impl PageServerHandler {
req,
} => {
fail::fail_point!("ps::handle-pagerequest-message::slrusegment");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
vec![
self.handle_get_slru_segment_request(&shard, &req, &ctx)
self.handle_get_slru_segment_request(&*shard.upgrade()?, &req, ctx)
.instrument(span.clone())
.await
.map(|msg| (msg, timer))
@@ -1372,13 +1345,12 @@ impl PageServerHandler {
requests,
} => {
fail::fail_point!("ps::handle-pagerequest-message::test");
let (shard, ctx) = upgrade_handle_and_set_context!(shard);
(
{
let npages = requests.len();
trace!(npages, "handling getpage request");
let res = self
.handle_test_request_batch(&shard, requests, &ctx)
.handle_test_request_batch(&*shard.upgrade()?, requests, ctx)
.instrument(span.clone())
.await;
assert_eq!(res.len(), npages);
@@ -2154,7 +2126,6 @@ impl PageServerHandler {
.get(tenant_id, timeline_id, ShardSelector::Zero)
.await?;
set_tracing_field_shard_id(&timeline);
let ctx = ctx.with_scope_timeline(&timeline);
if timeline.is_archived() == Some(true) {
tracing::info!(
@@ -2172,7 +2143,7 @@ impl PageServerHandler {
lsn,
crate::tenant::timeline::WaitLsnWaiter::PageService,
crate::tenant::timeline::WaitLsnTimeout::Default,
&ctx,
ctx,
)
.await?;
timeline
@@ -2198,7 +2169,7 @@ impl PageServerHandler {
prev_lsn,
full_backup,
replica,
&ctx,
ctx,
)
.await
.map_err(map_basebackup_error)?;
@@ -2221,7 +2192,7 @@ impl PageServerHandler {
prev_lsn,
full_backup,
replica,
&ctx,
ctx,
)
.await
.map_err(map_basebackup_error)?;
@@ -2238,7 +2209,7 @@ impl PageServerHandler {
prev_lsn,
full_backup,
replica,
&ctx,
ctx,
)
.await
.map_err(map_basebackup_error)?;

View File

@@ -2758,7 +2758,7 @@ mod tests {
TimelineId::from_array(hex!("11223344556677881122334455667788"));
let (tenant, ctx) = harness.load().await;
let (tline, ctx) = tenant
let tline = tenant
.create_empty_timeline(TIMELINE_ID, Lsn(0x10), DEFAULT_PG_VERSION, &ctx)
.await?;
let tline = tline.raw_timeline().unwrap();

View File

@@ -77,8 +77,6 @@ use self::timeline::{
EvictionTaskTenantState, GcCutoffs, TimelineDeleteProgress, TimelineResources, WaitLsnError,
};
use crate::config::PageServerConf;
use crate::context;
use crate::context::RequestContextBuilder;
use crate::context::{DownloadBehavior, RequestContext};
use crate::deletion_queue::{DeletionQueueClient, DeletionQueueError};
use crate::l0_flush::L0FlushGlobalState;
@@ -1116,7 +1114,7 @@ impl Tenant {
}
};
let (timeline, timeline_ctx) = self.create_timeline_struct(
let timeline = self.create_timeline_struct(
timeline_id,
&metadata,
previous_heatmap,
@@ -1126,7 +1124,6 @@ impl Tenant {
idempotency.clone(),
index_part.gc_compaction.clone(),
index_part.rel_size_migration.clone(),
ctx,
)?;
let disk_consistent_lsn = timeline.get_disk_consistent_lsn();
anyhow::ensure!(
@@ -1260,7 +1257,7 @@ impl Tenant {
match activate {
ActivateTimelineArgs::Yes { broker_client } => {
info!("activating timeline after reload from pgdata import task");
timeline.activate(self.clone(), broker_client, None, &timeline_ctx);
timeline.activate(self.clone(), broker_client, None, ctx);
}
ActivateTimelineArgs::No => (),
}
@@ -1768,7 +1765,6 @@ impl Tenant {
import_pgdata,
ActivateTimelineArgs::No,
guard,
ctx.detached_child(TaskKind::ImportPgdata, DownloadBehavior::Warn),
));
}
}
@@ -1786,7 +1782,6 @@ impl Tenant {
timeline_id,
&index_part.metadata,
remote_timeline_client,
ctx,
)
.instrument(tracing::info_span!("timeline_delete", %timeline_id))
.await
@@ -2224,7 +2219,7 @@ impl Tenant {
self.clone(),
broker_client.clone(),
background_jobs_can_start,
&ctx.with_scope_timeline(&timeline),
&ctx,
);
}
@@ -2421,8 +2416,8 @@ impl Tenant {
new_timeline_id: TimelineId,
initdb_lsn: Lsn,
pg_version: u32,
ctx: &RequestContext,
) -> anyhow::Result<(UninitializedTimeline, RequestContext)> {
_ctx: &RequestContext,
) -> anyhow::Result<UninitializedTimeline> {
anyhow::ensure!(
self.is_active(),
"Cannot create empty timelines on inactive tenant"
@@ -2457,7 +2452,6 @@ impl Tenant {
initdb_lsn,
None,
None,
ctx,
)
.await
}
@@ -2475,7 +2469,7 @@ impl Tenant {
pg_version: u32,
ctx: &RequestContext,
) -> anyhow::Result<Arc<Timeline>> {
let (uninit_tl, ctx) = self
let uninit_tl = self
.create_empty_timeline(new_timeline_id, initdb_lsn, pg_version, ctx)
.await?;
let tline = uninit_tl.raw_timeline().expect("we just created it");
@@ -2487,7 +2481,7 @@ impl Tenant {
.init_empty_test_timeline()
.context("init_empty_test_timeline")?;
modification
.commit(&ctx)
.commit(ctx)
.await
.context("commit init_empty_test_timeline modification")?;
@@ -2705,12 +2699,7 @@ impl Tenant {
// doing stuff before the IndexPart is durable in S3, which is done by the previous section.
let activated_timeline = match result {
CreateTimelineResult::Created(timeline) => {
timeline.activate(
self.clone(),
broker_client,
None,
&ctx.with_scope_timeline(&timeline),
);
timeline.activate(self.clone(), broker_client, None, ctx);
timeline
}
CreateTimelineResult::Idempotent(timeline) => {
@@ -2772,9 +2761,10 @@ impl Tenant {
}
};
let (mut uninit_timeline, timeline_ctx) = {
let mut uninit_timeline = {
let this = &self;
let initdb_lsn = Lsn(0);
let _ctx = ctx;
async move {
let new_metadata = TimelineMetadata::new(
// Initialize disk_consistent LSN to 0, The caller must import some data to
@@ -2794,7 +2784,6 @@ impl Tenant {
initdb_lsn,
None,
None,
ctx,
)
.await
}
@@ -2824,7 +2813,6 @@ impl Tenant {
index_part,
activate,
timeline_create_guard,
timeline_ctx.detached_child(TaskKind::ImportPgdata, DownloadBehavior::Warn),
));
// NB: the timeline doesn't exist in self.timelines at this point
@@ -2838,7 +2826,6 @@ impl Tenant {
index_part: import_pgdata::index_part_format::Root,
activate: ActivateTimelineArgs,
timeline_create_guard: TimelineCreateGuard,
ctx: RequestContext,
) {
debug_assert_current_span_has_tenant_and_timeline_id();
info!("starting");
@@ -2850,7 +2837,6 @@ impl Tenant {
index_part,
activate,
timeline_create_guard,
ctx,
)
.await;
if let Err(err) = &res {
@@ -2866,8 +2852,9 @@ impl Tenant {
index_part: import_pgdata::index_part_format::Root,
activate: ActivateTimelineArgs,
timeline_create_guard: TimelineCreateGuard,
ctx: RequestContext,
) -> Result<(), anyhow::Error> {
let ctx = RequestContext::new(TaskKind::ImportPgdata, DownloadBehavior::Warn);
info!("importing pgdata");
import_pgdata::doit(&timeline, index_part, &ctx, self.cancel.clone())
.await
@@ -3076,7 +3063,6 @@ impl Tenant {
let mut has_pending_l0 = false;
for timeline in compact_l0 {
let ctx = &ctx.with_scope_timeline(&timeline);
let outcome = timeline
.compact(cancel, CompactFlags::OnlyL0Compaction.into(), ctx)
.instrument(info_span!("compact_timeline", timeline_id = %timeline.timeline_id))
@@ -3110,7 +3096,6 @@ impl Tenant {
if !timeline.is_active() {
continue;
}
let ctx = &ctx.with_scope_timeline(&timeline);
let mut outcome = timeline
.compact(cancel, EnumSet::default(), ctx)
@@ -3336,7 +3321,7 @@ impl Tenant {
self.clone(),
broker_client.clone(),
background_jobs_can_start,
&ctx.with_scope_timeline(timeline),
ctx,
);
activated_timelines += 1;
}
@@ -3842,7 +3827,6 @@ impl Tenant {
resident_size: 0,
physical_size: 0,
max_logical_size: 0,
max_logical_size_per_shard: 0,
};
for timeline in self.timelines.lock().unwrap().values() {
@@ -3859,10 +3843,6 @@ impl Tenant {
);
}
result.max_logical_size_per_shard = result
.max_logical_size
.div_ceil(self.tenant_shard_id.shard_count.count() as u64);
result
}
}
@@ -4156,8 +4136,7 @@ impl Tenant {
create_idempotency: CreateTimelineIdempotency,
gc_compaction_state: Option<GcCompactionState>,
rel_size_v2_status: Option<RelSizeMigration>,
ctx: &RequestContext,
) -> anyhow::Result<(Arc<Timeline>, RequestContext)> {
) -> anyhow::Result<Arc<Timeline>> {
let state = match cause {
CreateTimelineCause::Load => {
let ancestor_id = new_metadata.ancestor_timeline();
@@ -4193,11 +4172,7 @@ impl Tenant {
self.cancel.child_token(),
);
let timeline_ctx = RequestContextBuilder::extend(ctx)
.scope(context::Scope::new_timeline(&timeline))
.build();
Ok((timeline, timeline_ctx))
Ok(timeline)
}
/// [`Tenant::shutdown`] must be called before dropping the returned [`Tenant`] object
@@ -4613,7 +4588,6 @@ impl Tenant {
// Ensures all timelines use the same start time when computing the time cutoff.
let now_ts_for_pitr_calc = SystemTime::now();
for timeline in timelines.iter() {
let ctx = &ctx.with_scope_timeline(timeline);
let cutoff = timeline
.get_last_record_lsn()
.checked_sub(horizon)
@@ -4787,7 +4761,7 @@ impl Tenant {
src_timeline: &Arc<Timeline>,
dst_id: TimelineId,
start_lsn: Option<Lsn>,
ctx: &RequestContext,
_ctx: &RequestContext,
) -> Result<CreateTimelineResult, CreateTimelineError> {
let src_id = src_timeline.timeline_id;
@@ -4890,7 +4864,7 @@ impl Tenant {
src_timeline.pg_version,
);
let (uninitialized_timeline, _timeline_ctx) = self
let uninitialized_timeline = self
.prepare_new_timeline(
dst_id,
&metadata,
@@ -4898,7 +4872,6 @@ impl Tenant {
start_lsn + 1,
Some(Arc::clone(src_timeline)),
Some(src_timeline.get_rel_size_v2_status()),
ctx,
)
.await?;
@@ -5165,7 +5138,7 @@ impl Tenant {
pgdata_lsn,
pg_version,
);
let (mut raw_timeline, timeline_ctx) = self
let mut raw_timeline = self
.prepare_new_timeline(
timeline_id,
&new_metadata,
@@ -5173,7 +5146,6 @@ impl Tenant {
pgdata_lsn,
None,
None,
ctx,
)
.await?;
@@ -5184,7 +5156,7 @@ impl Tenant {
&unfinished_timeline,
&pgdata_path,
pgdata_lsn,
&timeline_ctx,
ctx,
)
.await
.with_context(|| {
@@ -5245,7 +5217,6 @@ impl Tenant {
/// An empty layer map is initialized, and new data and WAL can be imported starting
/// at 'disk_consistent_lsn'. After any initial data has been imported, call
/// `finish_creation` to insert the Timeline into the timelines map.
#[allow(clippy::too_many_arguments)]
async fn prepare_new_timeline<'a>(
&'a self,
new_timeline_id: TimelineId,
@@ -5254,8 +5225,7 @@ impl Tenant {
start_lsn: Lsn,
ancestor: Option<Arc<Timeline>>,
rel_size_v2_status: Option<RelSizeMigration>,
ctx: &RequestContext,
) -> anyhow::Result<(UninitializedTimeline<'a>, RequestContext)> {
) -> anyhow::Result<UninitializedTimeline<'a>> {
let tenant_shard_id = self.tenant_shard_id;
let resources = self.build_timeline_resources(new_timeline_id);
@@ -5263,7 +5233,7 @@ impl Tenant {
.remote_client
.init_upload_queue_for_empty_remote(new_metadata, rel_size_v2_status.clone())?;
let (timeline_struct, timeline_ctx) = self
let timeline_struct = self
.create_timeline_struct(
new_timeline_id,
new_metadata,
@@ -5274,7 +5244,6 @@ impl Tenant {
create_guard.idempotency.clone(),
None,
rel_size_v2_status,
ctx,
)
.context("Failed to create timeline data structure")?;
@@ -5295,13 +5264,10 @@ impl Tenant {
"Successfully created initial files for timeline {tenant_shard_id}/{new_timeline_id}"
);
Ok((
UninitializedTimeline::new(
self,
new_timeline_id,
Some((timeline_struct, create_guard)),
),
timeline_ctx,
Ok(UninitializedTimeline::new(
self,
new_timeline_id,
Some((timeline_struct, create_guard)),
))
}
@@ -5754,7 +5720,7 @@ pub(crate) mod harness {
logging::TracingErrorLayerEnablement::EnableWithRustLogFilter,
logging::Output::Stdout,
)
.expect("Failed to init test logging");
.expect("Failed to init test logging")
});
}
@@ -5836,8 +5802,7 @@ pub(crate) mod harness {
}
pub(crate) async fn load(&self) -> (Arc<Tenant>, RequestContext) {
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error)
.with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
(
self.do_try_load(&ctx)
.await
@@ -6559,11 +6524,7 @@ mod tests {
tline.freeze_and_flush().await?;
tline
.compact(
&CancellationToken::new(),
CompactFlags::NoYield.into(),
&ctx,
)
.compact(&CancellationToken::new(), EnumSet::empty(), &ctx)
.await?;
let mut writer = tline.writer().await;
@@ -6580,11 +6541,7 @@ mod tests {
tline.freeze_and_flush().await?;
tline
.compact(
&CancellationToken::new(),
CompactFlags::NoYield.into(),
&ctx,
)
.compact(&CancellationToken::new(), EnumSet::empty(), &ctx)
.await?;
let mut writer = tline.writer().await;
@@ -6601,11 +6558,7 @@ mod tests {
tline.freeze_and_flush().await?;
tline
.compact(
&CancellationToken::new(),
CompactFlags::NoYield.into(),
&ctx,
)
.compact(&CancellationToken::new(), EnumSet::empty(), &ctx)
.await?;
let mut writer = tline.writer().await;
@@ -6622,11 +6575,7 @@ mod tests {
tline.freeze_and_flush().await?;
tline
.compact(
&CancellationToken::new(),
CompactFlags::NoYield.into(),
&ctx,
)
.compact(&CancellationToken::new(), EnumSet::empty(), &ctx)
.await?;
assert_eq!(
@@ -6709,9 +6658,7 @@ mod tests {
timeline.freeze_and_flush().await?;
if compact {
// this requires timeline to be &Arc<Timeline>
timeline
.compact(&cancel, CompactFlags::NoYield.into(), ctx)
.await?;
timeline.compact(&cancel, EnumSet::empty(), ctx).await?;
}
// this doesn't really need to use the timeline_id target, but it is closer to what it
@@ -6878,7 +6825,7 @@ mod tests {
let (tenant, ctx) = harness.load().await;
let io_concurrency = IoConcurrency::spawn_for_test();
let (tline, ctx) = tenant
let tline = tenant
.create_empty_timeline(TIMELINE_ID, Lsn(0), DEFAULT_PG_VERSION, &ctx)
.await?;
let tline = tline.raw_timeline().unwrap();
@@ -7038,7 +6985,6 @@ mod tests {
child_timeline.freeze_and_flush().await?;
let mut flags = EnumSet::new();
flags.insert(CompactFlags::ForceRepartition);
flags.insert(CompactFlags::NoYield);
child_timeline
.compact(&CancellationToken::new(), flags, &ctx)
.await?;
@@ -7417,9 +7363,7 @@ mod tests {
// Perform a cycle of flush, compact, and GC
tline.freeze_and_flush().await?;
tline
.compact(&cancel, CompactFlags::NoYield.into(), &ctx)
.await?;
tline.compact(&cancel, EnumSet::empty(), &ctx).await?;
tenant
.gc_iteration(Some(tline.timeline_id), 0, Duration::ZERO, &cancel, &ctx)
.await?;
@@ -7503,7 +7447,7 @@ mod tests {
.await;
let initdb_lsn = Lsn(0x20);
let (utline, ctx) = tenant
let utline = tenant
.create_empty_timeline(TIMELINE_ID, initdb_lsn, DEFAULT_PG_VERSION, &ctx)
.await?;
let tline = utline.raw_timeline().unwrap();
@@ -7570,7 +7514,7 @@ mod tests {
let harness = TenantHarness::create(name).await?;
{
let (tenant, ctx) = harness.load().await;
let (tline, _ctx) = tenant
let tline = tenant
.create_empty_timeline(TIMELINE_ID, Lsn(0), DEFAULT_PG_VERSION, &ctx)
.await?;
// Leave the timeline ID in [`Tenant::timelines_creating`] to exclude attempting to create it again
@@ -7748,7 +7692,6 @@ mod tests {
let mut flags = EnumSet::new();
flags.insert(CompactFlags::ForceImageLayerCreation);
flags.insert(CompactFlags::ForceRepartition);
flags.insert(CompactFlags::NoYield);
flags
} else {
EnumSet::empty()
@@ -7799,9 +7742,7 @@ mod tests {
let before_num_l0_delta_files =
tline.layers.read().await.layer_map()?.level0_deltas().len();
tline
.compact(&cancel, CompactFlags::NoYield.into(), &ctx)
.await?;
tline.compact(&cancel, EnumSet::empty(), &ctx).await?;
let after_num_l0_delta_files = tline.layers.read().await.layer_map()?.level0_deltas().len();
@@ -7917,6 +7858,7 @@ mod tests {
Ok((res, reconstruct_state.get_delta_layers_visited() as usize))
}
#[allow(clippy::needless_range_loop)]
for blknum in 0..NUM_KEYS {
lsn = Lsn(lsn.0 + 0x10);
test_key.field6 = (blknum * STEP) as u32;
@@ -7966,7 +7908,6 @@ mod tests {
let mut flags = EnumSet::new();
flags.insert(CompactFlags::ForceImageLayerCreation);
flags.insert(CompactFlags::ForceRepartition);
flags.insert(CompactFlags::NoYield);
flags
},
&ctx,
@@ -8429,7 +8370,6 @@ mod tests {
let mut flags = EnumSet::new();
flags.insert(CompactFlags::ForceImageLayerCreation);
flags.insert(CompactFlags::ForceRepartition);
flags.insert(CompactFlags::NoYield);
flags
},
&ctx,
@@ -8497,7 +8437,6 @@ mod tests {
let mut flags = EnumSet::new();
flags.insert(CompactFlags::ForceImageLayerCreation);
flags.insert(CompactFlags::ForceRepartition);
flags.insert(CompactFlags::NoYield);
flags
},
&ctx,

View File

@@ -471,8 +471,7 @@ pub(crate) mod tests {
blobs: &[Vec<u8>],
compression: bool,
) -> Result<(), Error> {
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let (_temp_dir, pathbuf, offsets) =
write_maybe_compressed::<BUFFERED>(blobs, compression, &ctx).await?;

View File

@@ -219,11 +219,7 @@ impl LocationConf {
};
let shard = if conf.shard_count == 0 {
// NB: carry over the persisted stripe size instead of using the default. This doesn't
// matter for most practical purposes, since unsharded tenants don't use the stripe
// size, but can cause inconsistencies between storcon and Pageserver and cause manual
// splits without `new_stripe_size` to use an unintended stripe size.
ShardIdentity::unsharded_with_stripe_size(ShardStripeSize(conf.shard_stripe_size))
ShardIdentity::unsharded()
} else {
ShardIdentity::new(
ShardNumber(conf.shard_number),

View File

@@ -32,7 +32,8 @@ use hex;
use thiserror::Error;
use tracing::error;
use crate::context::RequestContext;
use crate::context::{DownloadBehavior, RequestContext};
use crate::task_mgr::TaskKind;
use crate::tenant::block_io::{BlockReader, BlockWriter};
// The maximum size of a value stored in the B-tree. 5 bytes is enough currently.
@@ -476,15 +477,16 @@ where
}
#[allow(dead_code)]
pub async fn dump(&self, ctx: &RequestContext) -> Result<()> {
pub async fn dump(&self) -> Result<()> {
let mut stack = Vec::new();
let ctx = RequestContext::new(TaskKind::DebugTool, DownloadBehavior::Error);
stack.push((self.root_blk, String::new(), 0, 0, 0));
let block_cursor = self.reader.block_cursor();
while let Some((blknum, path, depth, child_idx, key_off)) = stack.pop() {
let blk = block_cursor.read_blk(self.start_blk + blknum, ctx).await?;
let blk = block_cursor.read_blk(self.start_blk + blknum, &ctx).await?;
let buf: &[u8] = blk.as_ref();
let node = OnDiskNode::<L>::deparse(buf)?;
@@ -833,8 +835,6 @@ pub(crate) mod tests {
use rand::Rng;
use super::*;
use crate::context::DownloadBehavior;
use crate::task_mgr::TaskKind;
use crate::tenant::block_io::{BlockCursor, BlockLease, BlockReaderRef};
#[derive(Clone, Default)]
@@ -869,8 +869,7 @@ pub(crate) mod tests {
let mut disk = TestDisk::new();
let mut writer = DiskBtreeBuilder::<_, 6>::new(&mut disk);
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let all_keys: Vec<&[u8; 6]> = vec![
b"xaaaaa", b"xaaaba", b"xaaaca", b"xabaaa", b"xababa", b"xabaca", b"xabada", b"xabadb",
@@ -888,7 +887,7 @@ pub(crate) mod tests {
let reader = DiskBtreeReader::new(0, root_offset, disk);
reader.dump(&ctx).await?;
reader.dump().await?;
// Test the `get` function on all the keys.
for (key, val) in all_data.iter() {
@@ -980,8 +979,7 @@ pub(crate) mod tests {
async fn lots_of_keys() -> Result<()> {
let mut disk = TestDisk::new();
let mut writer = DiskBtreeBuilder::<_, 8>::new(&mut disk);
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
const NUM_KEYS: u64 = 1000;
@@ -999,7 +997,7 @@ pub(crate) mod tests {
let reader = DiskBtreeReader::new(0, root_offset, disk);
reader.dump(&ctx).await?;
reader.dump().await?;
use std::sync::Mutex;
@@ -1169,8 +1167,7 @@ pub(crate) mod tests {
// Build a tree from it
let mut disk = TestDisk::new();
let mut writer = DiskBtreeBuilder::<_, 26>::new(&mut disk);
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
for (key, val) in disk_btree_test_data::TEST_DATA {
writer.append(&key, val)?;
@@ -1201,7 +1198,7 @@ pub(crate) mod tests {
.await?;
assert_eq!(count, disk_btree_test_data::TEST_DATA.len());
reader.dump(&ctx).await?;
reader.dump().await?;
Ok(())
}

View File

@@ -9,7 +9,7 @@ use camino::Utf8PathBuf;
use num_traits::Num;
use pageserver_api::shard::TenantShardId;
use tokio_epoll_uring::{BoundedBuf, Slice};
use tracing::{error, info_span};
use tracing::error;
use utils::id::TimelineId;
use crate::assert_u64_eq_usize::{U64IsUsize, UsizeIsU64};
@@ -76,7 +76,6 @@ impl EphemeralFile {
|| IoBufferMut::with_capacity(TAIL_SZ),
gate.enter()?,
ctx,
info_span!(parent: None, "ephemeral_file_buffered_writer", tenant_id=%tenant_shard_id.tenant_id, shard_id=%tenant_shard_id.shard_slug(), timeline_id=%timeline_id, path = %filename),
),
_gate_guard: gate.enter()?,
})
@@ -352,8 +351,7 @@ mod tests {
let timeline_id = TimelineId::from_str("22000000000000000000000000000000").unwrap();
fs::create_dir_all(conf.timeline_path(&tenant_shard_id, &timeline_id))?;
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
Ok((conf, tenant_shard_id, timeline_id, ctx))
}

View File

@@ -14,7 +14,7 @@ use futures::StreamExt;
use itertools::Itertools;
use once_cell::sync::Lazy;
use pageserver_api::key::Key;
use pageserver_api::models::{DetachBehavior, LocationConfigMode};
use pageserver_api::models::LocationConfigMode;
use pageserver_api::shard::{
ShardCount, ShardIdentity, ShardIndex, ShardNumber, ShardStripeSize, TenantShardId,
};
@@ -1914,7 +1914,7 @@ impl TenantManager {
tenant_shard_id: TenantShardId,
timeline_id: TimelineId,
prepared: PreparedTimelineDetach,
behavior: DetachBehavior,
behavior: detach_ancestor::DetachBehavior,
mut attempt: detach_ancestor::Attempt,
ctx: &RequestContext,
) -> Result<HashSet<TimelineId>, detach_ancestor::Error> {

View File

@@ -229,7 +229,6 @@ async fn download_object(
|| IoBufferMut::with_capacity(super::BUFFER_SIZE),
gate.enter().map_err(|_| DownloadError::Cancelled)?,
ctx,
tracing::info_span!(parent: None, "download_object_buffered_writer", %dst_path),
);
// TODO: use vectored write (writev) once supported by tokio-epoll-uring.

View File

@@ -491,10 +491,7 @@ impl JobGenerator<PendingDownload, RunningDownload, CompleteDownload, DownloadCo
let remote_storage = self.remote_storage.clone();
let conf = self.tenant_manager.get_conf();
let tenant_shard_id = *secondary_state.get_tenant_shard_id();
let download_ctx = self
.root_ctx
.attached_child()
.with_scope_secondary_tenant(&tenant_shard_id);
let download_ctx = self.root_ctx.attached_child();
(RunningDownload { barrier }, Box::pin(async move {
let _completion = completion;
@@ -774,7 +771,6 @@ impl<'a> TenantDownloader<'a> {
// Download the layers in the heatmap
for timeline in heatmap.timelines {
let ctx = &ctx.with_scope_secondary_timeline(tenant_shard_id, &timeline.timeline_id);
let timeline_state = timeline_states
.remove(&timeline.timeline_id)
.expect("Just populated above");

View File

@@ -474,7 +474,7 @@ async fn fill_logical_sizes(
if cached_size.is_none() {
let timeline = Arc::clone(timeline_hash.get(&timeline_id).unwrap());
let parallel_size_calcs = Arc::clone(limit);
let ctx = ctx.attached_child().with_scope_timeline(&timeline);
let ctx = ctx.attached_child();
joinset.spawn(
calculate_logical_size(parallel_size_calcs, timeline, lsn, cause, ctx)
.in_current_span(),

View File

@@ -1334,7 +1334,7 @@ impl DeltaLayerInner {
block_reader,
);
tree_reader.dump(ctx).await?;
tree_reader.dump().await?;
let keys = self.index_entries(ctx).await?;
@@ -1972,7 +1972,6 @@ pub(crate) mod test {
.create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, ctx)
.await
.unwrap();
let ctx = &ctx.with_scope_timeline(&timeline);
let initdb_layer = timeline
.layers

View File

@@ -199,7 +199,7 @@ impl ImageLayerInner {
block_reader,
);
tree_reader.dump(ctx).await?;
tree_reader.dump().await?;
tree_reader
.visit(

View File

@@ -8,6 +8,7 @@ use utils::id::TimelineId;
use super::failpoints::{Failpoint, FailpointKind};
use super::*;
use crate::context::DownloadBehavior;
use crate::task_mgr::TaskKind;
use crate::tenant::harness::{TenantHarness, test_img};
use crate::tenant::storage_layer::{IoConcurrency, LayerVisibilityHint};
@@ -26,9 +27,11 @@ async fn smoke_test() {
let h = TenantHarness::create("smoke_test").await.unwrap();
let span = h.span();
let download_span = span.in_scope(|| tracing::info_span!("downloading", timeline_id = 1));
let (tenant, ctx) = h.load().await;
let (tenant, _) = h.load().await;
let io_concurrency = IoConcurrency::spawn_for_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Download);
let image_layers = vec![(
Lsn(0x40),
vec![(
@@ -53,7 +56,6 @@ async fn smoke_test() {
)
.await
.unwrap();
let ctx = &ctx.with_scope_timeline(&timeline);
// Grab one of the timeline's layers to exercise in the test, and the other layer that is just
// there to avoid the timeline being illegally empty
@@ -92,7 +94,7 @@ async fn smoke_test() {
controlfile_keyspace.clone(),
Lsn(0x10)..Lsn(0x11),
&mut data,
ctx,
&ctx,
)
.await
.unwrap();
@@ -127,7 +129,7 @@ async fn smoke_test() {
controlfile_keyspace.clone(),
Lsn(0x10)..Lsn(0x11),
&mut data,
ctx,
&ctx,
)
.instrument(download_span.clone())
.await
@@ -177,7 +179,7 @@ async fn smoke_test() {
// plain downloading is rarely needed
layer
.download_and_keep_resident(ctx)
.download_and_keep_resident(&ctx)
.instrument(download_span)
.await
.unwrap();
@@ -339,7 +341,6 @@ fn read_wins_pending_eviction() {
.create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, &ctx)
.await
.unwrap();
let ctx = ctx.with_scope_timeline(&timeline);
let layer = {
let mut layers = {
@@ -472,7 +473,6 @@ fn multiple_pending_evictions_scenario(name: &'static str, in_order: bool) {
.create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, &ctx)
.await
.unwrap();
let ctx = ctx.with_scope_timeline(&timeline);
let layer = {
let mut layers = {
@@ -642,12 +642,12 @@ async fn cancelled_get_or_maybe_download_does_not_cancel_eviction() {
.create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, &ctx)
.await
.unwrap();
let ctx = ctx.with_scope_timeline(&timeline);
// This test does downloads
let ctx = RequestContextBuilder::extend(&ctx)
.download_behavior(DownloadBehavior::Download)
.build();
let layer = {
let mut layers = {
let layers = timeline.layers.read().await;
@@ -727,7 +727,6 @@ async fn evict_and_wait_does_not_wait_for_download() {
.create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, &ctx)
.await
.unwrap();
let ctx = ctx.with_scope_timeline(&timeline);
// This test does downloads
let ctx = RequestContextBuilder::extend(&ctx)

View File

@@ -45,9 +45,8 @@ use pageserver_api::key::{
use pageserver_api::keyspace::{KeySpaceAccum, KeySpaceRandomAccum, SparseKeyPartitioning};
use pageserver_api::models::{
CompactKeyRange, CompactLsnRange, CompactionAlgorithm, CompactionAlgorithmSettings,
DetachBehavior, DownloadRemoteLayersTaskInfo, DownloadRemoteLayersTaskSpawnRequest,
EvictionPolicy, InMemoryLayerInfo, LayerMapInfo, LsnLease, PageTraceEvent, RelSizeMigration,
TimelineState,
DownloadRemoteLayersTaskInfo, DownloadRemoteLayersTaskSpawnRequest, EvictionPolicy,
InMemoryLayerInfo, LayerMapInfo, LsnLease, PageTraceEvent, RelSizeMigration, TimelineState,
};
use pageserver_api::reltag::{BlockNumber, RelTag};
use pageserver_api::shard::{ShardIdentity, ShardIndex, ShardNumber, TenantShardId};
@@ -68,7 +67,6 @@ use tracing::*;
use utils::generation::Generation;
use utils::guard_arc_swap::GuardArcSwap;
use utils::id::TimelineId;
use utils::logging::{MonitorSlowFutureCallback, monitor_slow_future};
use utils::lsn::{AtomicLsn, Lsn, RecordLsn};
use utils::postgres_client::PostgresClientProtocol;
use utils::rate_limit::RateLimit;
@@ -289,7 +287,7 @@ pub struct Timeline {
// The LSN of gc-compaction that was last applied to this timeline.
gc_compaction_state: ArcSwap<Option<GcCompactionState>>,
pub(crate) metrics: Arc<TimelineMetrics>,
pub(super) metrics: TimelineMetrics,
// `Timeline` doesn't write these metrics itself, but it manages the lifetime. Code
// in `crate::page_service` writes these metrics.
@@ -441,8 +439,6 @@ pub struct Timeline {
heatmap_layers_downloader: Mutex<Option<heatmap_layers_downloader::HeatmapLayersDownloader>>,
pub(crate) rel_size_v2_status: ArcSwapOption<RelSizeMigration>,
wait_lsn_log_slow: tokio::sync::Semaphore,
}
pub(crate) enum PreviousHeatmap {
@@ -1483,67 +1479,17 @@ impl Timeline {
WaitLsnTimeout::Default => self.conf.wait_lsn_timeout,
};
let timer = crate::metrics::WAIT_LSN_TIME.start_timer();
let start_finish_counterpair_guard = self.metrics.wait_lsn_start_finish_counterpair.guard();
let _timer = crate::metrics::WAIT_LSN_TIME.start_timer();
let wait_for_timeout = self.last_record_lsn.wait_for_timeout(lsn, timeout);
let wait_for_timeout = std::pin::pin!(wait_for_timeout);
// Use threshold of 1 because even 1 second of wait for ingest is very much abnormal.
let log_slow_threshold = Duration::from_secs(1);
// Use period of 10 to avoid flooding logs during an outage that affects all timelines.
let log_slow_period = Duration::from_secs(10);
let mut logging_permit = None;
let wait_for_timeout = monitor_slow_future(
log_slow_threshold,
log_slow_period,
wait_for_timeout,
|MonitorSlowFutureCallback {
ready,
is_slow,
elapsed_total,
elapsed_since_last_callback,
}| {
self.metrics
.wait_lsn_in_progress_micros
.inc_by(u64::try_from(elapsed_since_last_callback.as_micros()).unwrap());
if !is_slow {
return;
}
// It's slow, see if we should log it.
// (We limit the logging to one per invocation per timeline to avoid excessive
// logging during an extended broker / networking outage that affects all timelines.)
if logging_permit.is_none() {
logging_permit = self.wait_lsn_log_slow.try_acquire().ok();
}
if logging_permit.is_none() {
return;
}
// We log it.
if ready {
info!(
"slow wait_lsn completed after {:.3}s",
elapsed_total.as_secs_f64()
);
} else {
info!(
"slow wait_lsn still running for {:.3}s",
elapsed_total.as_secs_f64()
);
}
},
);
let res = wait_for_timeout.await;
// don't count the time spent waiting for lock below, and also in walreceiver.status(), towards the wait_lsn_time_histo
drop(logging_permit);
drop(start_finish_counterpair_guard);
drop(timer);
match res {
match self.last_record_lsn.wait_for_timeout(lsn, timeout).await {
Ok(()) => Ok(()),
Err(e) => {
use utils::seqwait::SeqWaitError::*;
match e {
Shutdown => Err(WaitLsnError::Shutdown),
Timeout => {
// don't count the time spent waiting for lock below, and also in walreceiver.status(), towards the wait_lsn_time_histo
drop(_timer);
let walreceiver_status = self.walreceiver_status();
Err(WaitLsnError::Timeout(format!(
"Timed out while waiting for WAL record at LSN {} to arrive, last_record_lsn {} disk consistent LSN={}, WalReceiver status: {}",
@@ -2477,9 +2423,8 @@ impl Timeline {
}
fn get_l0_flush_delay_threshold(&self) -> Option<usize> {
// By default, delay L0 flushes at 3x the compaction threshold. The compaction threshold
// defaults to 10, and L0 compaction is generally able to keep L0 counts below 30.
const DEFAULT_L0_FLUSH_DELAY_FACTOR: usize = 3;
// Disable L0 flushes by default. This and compaction needs further tuning.
const DEFAULT_L0_FLUSH_DELAY_FACTOR: usize = 0; // TODO: default to e.g. 3
// If compaction is disabled, don't delay.
if self.get_compaction_period() == Duration::ZERO {
@@ -2507,9 +2452,8 @@ impl Timeline {
}
fn get_l0_flush_stall_threshold(&self) -> Option<usize> {
// Disable L0 stalls by default. Stalling can cause unavailability if L0 compaction isn't
// responsive, and it can e.g. block on other compaction via the compaction semaphore or
// sibling timelines. We need more confidence before enabling this.
// Disable L0 stalls by default. In ingest benchmarks, we see image compaction take >10
// minutes, blocking L0 compaction, and we can't stall L0 flushes for that long.
const DEFAULT_L0_FLUSH_STALL_FACTOR: usize = 0; // TODO: default to e.g. 5
// If compaction is disabled, don't stall.
@@ -2741,14 +2685,14 @@ impl Timeline {
}
Arc::new_cyclic(|myself| {
let metrics = Arc::new(TimelineMetrics::new(
let metrics = TimelineMetrics::new(
&tenant_shard_id,
&timeline_id,
crate::metrics::EvictionsWithLowResidenceDurationBuilder::new(
"mtime",
evictions_low_residence_duration_metric_threshold,
),
));
);
let aux_file_metrics = metrics.aux_file_size_gauge.clone();
let mut result = Timeline {
@@ -2877,8 +2821,6 @@ impl Timeline {
heatmap_layers_downloader: Mutex::new(None),
rel_size_v2_status: ArcSwapOption::from_pointee(rel_size_v2_status),
wait_lsn_log_slow: tokio::sync::Semaphore::new(1),
};
result.repartition_threshold =
@@ -2934,7 +2876,7 @@ impl Timeline {
"layer flush task",
async move {
let _guard = guard;
let background_ctx = RequestContext::todo_child(TaskKind::LayerFlushTask, DownloadBehavior::Error).with_scope_timeline(&self_clone);
let background_ctx = RequestContext::todo_child(TaskKind::LayerFlushTask, DownloadBehavior::Error);
self_clone.flush_loop(layer_flush_start_rx, &background_ctx).await;
let mut flush_loop_state = self_clone.flush_loop_state.lock().unwrap();
assert!(matches!(*flush_loop_state, FlushLoopState::Running{..}));
@@ -5446,7 +5388,7 @@ impl Timeline {
self: &Arc<Timeline>,
tenant: &crate::tenant::Tenant,
options: detach_ancestor::Options,
behavior: DetachBehavior,
behavior: detach_ancestor::DetachBehavior,
ctx: &RequestContext,
) -> Result<detach_ancestor::Progress, detach_ancestor::Error> {
detach_ancestor::prepare(self, tenant, behavior, options, ctx).await
@@ -5467,7 +5409,7 @@ impl Timeline {
prepared: detach_ancestor::PreparedTimelineDetach,
ancestor_timeline_id: TimelineId,
ancestor_lsn: Lsn,
behavior: DetachBehavior,
behavior: detach_ancestor::DetachBehavior,
ctx: &RequestContext,
) -> Result<detach_ancestor::DetachingAndReparenting, detach_ancestor::Error> {
detach_ancestor::detach_and_reparent(
@@ -7198,7 +7140,6 @@ mod tests {
)
.await
.unwrap();
let ctx = &ctx.with_scope_timeline(&timeline);
// Layer visibility is an input to heatmap generation, so refresh it first
timeline.update_layer_visibility().await.unwrap();
@@ -7264,7 +7205,7 @@ mod tests {
eprintln!("Downloading {layer} and re-generating heatmap");
let ctx = &RequestContextBuilder::extend(ctx)
let ctx = &RequestContextBuilder::extend(&ctx)
.download_behavior(crate::context::DownloadBehavior::Download)
.build();

View File

@@ -393,9 +393,6 @@ impl GcCompactionQueue {
if job.dry_run {
flags |= CompactFlags::DryRun;
}
if options.flags.contains(CompactFlags::NoYield) {
flags |= CompactFlags::NoYield;
}
let options = CompactOptions {
flags,
sub_compaction: false,
@@ -1094,7 +1091,7 @@ impl Timeline {
let latest_gc_cutoff = self.get_applied_gc_cutoff_lsn();
tracing::info!(
"starting shard ancestor compaction, latest_gc_cutoff: {}, pitr cutoff {}",
"latest_gc_cutoff: {}, pitr cutoff {}",
*latest_gc_cutoff,
self.gc_info.read().unwrap().cutoffs.time
);
@@ -1123,7 +1120,6 @@ impl Timeline {
// Expensive, exhaustive check of keys in this layer: this guards against ShardedRange's calculations being
// wrong. If ShardedRange claims the local page count is zero, then no keys in this layer
// should be !is_key_disposable()
// TODO: exclude sparse keyspace from this check, otherwise it will infinitely loop.
let range = layer_desc.get_key_range();
let mut key = range.start;
while key < range.end {
@@ -2620,7 +2616,6 @@ impl Timeline {
) -> Result<CompactionOutcome, CompactionError> {
let sub_compaction = options.sub_compaction;
let job = GcCompactJob::from_compact_options(options.clone());
let no_yield = options.flags.contains(CompactFlags::NoYield);
if sub_compaction {
info!(
"running enhanced gc bottom-most compaction with sub-compaction, splitting compaction jobs"
@@ -2635,15 +2630,14 @@ impl Timeline {
idx + 1,
jobs_len
);
self.compact_with_gc_inner(cancel, job, ctx, no_yield)
.await?;
self.compact_with_gc_inner(cancel, job, ctx).await?;
}
if jobs_len == 0 {
info!("no jobs to run, skipping gc bottom-most compaction");
}
return Ok(CompactionOutcome::Done);
}
self.compact_with_gc_inner(cancel, job, ctx, no_yield).await
self.compact_with_gc_inner(cancel, job, ctx).await
}
async fn compact_with_gc_inner(
@@ -2651,7 +2645,6 @@ impl Timeline {
cancel: &CancellationToken,
job: GcCompactJob,
ctx: &RequestContext,
no_yield: bool,
) -> Result<CompactionOutcome, CompactionError> {
// Block other compaction/GC tasks from running for now. GC-compaction could run along
// with legacy compaction tasks in the future. Always ensure the lock order is compaction -> gc.
@@ -2921,18 +2914,14 @@ impl Timeline {
if cancel.is_cancelled() {
return Err(CompactionError::ShuttingDown);
}
if !no_yield {
let should_yield = self
.l0_compaction_trigger
.notified()
.now_or_never()
.is_some();
if should_yield {
tracing::info!(
"preempt gc-compaction when downloading layers: too many L0 layers"
);
return Ok(CompactionOutcome::YieldForL0);
}
let should_yield = self
.l0_compaction_trigger
.notified()
.now_or_never()
.is_some();
if should_yield {
tracing::info!("preempt gc-compaction when downloading layers: too many L0 layers");
return Ok(CompactionOutcome::YieldForL0);
}
let resident_layer = layer
.download_and_keep_resident(ctx)
@@ -3065,21 +3054,16 @@ impl Timeline {
if cancel.is_cancelled() {
return Err(CompactionError::ShuttingDown);
}
if !no_yield {
keys_processed += 1;
if keys_processed % 1000 == 0 {
let should_yield = self
.l0_compaction_trigger
.notified()
.now_or_never()
.is_some();
if should_yield {
tracing::info!(
"preempt gc-compaction in the main loop: too many L0 layers"
);
return Ok(CompactionOutcome::YieldForL0);
}
keys_processed += 1;
if keys_processed % 1000 == 0 {
let should_yield = self
.l0_compaction_trigger
.notified()
.now_or_never()
.is_some();
if should_yield {
tracing::info!("preempt gc-compaction in the main loop: too many L0 layers");
return Ok(CompactionOutcome::YieldForL0);
}
}
if self.shard_identity.is_key_disposable(&key) {
@@ -3189,11 +3173,7 @@ impl Timeline {
}
// TODO: move the below part to the loop body
let Some(last_key) = last_key else {
return Err(CompactionError::Other(anyhow!(
"no keys produced during compaction"
)));
};
let last_key = last_key.expect("no keys produced during compaction");
stat.on_unique_key_visited();
let retention = self

View File

@@ -11,7 +11,6 @@ use utils::id::TimelineId;
use utils::{crashsafe, fs_ext, pausable_failpoint};
use crate::config::PageServerConf;
use crate::context::RequestContext;
use crate::task_mgr::{self, TaskKind};
use crate::tenant::metadata::TimelineMetadata;
use crate::tenant::remote_timeline_client::{
@@ -292,11 +291,10 @@ impl DeleteTimelineFlow {
timeline_id: TimelineId,
local_metadata: &TimelineMetadata,
remote_client: RemoteTimelineClient,
ctx: &RequestContext,
) -> anyhow::Result<()> {
// Note: here we even skip populating layer map. Timeline is essentially uninitialized.
// RemoteTimelineClient is the only functioning part.
let (timeline, _timeline_ctx) = tenant
let timeline = tenant
.create_timeline_struct(
timeline_id,
local_metadata,
@@ -309,7 +307,6 @@ impl DeleteTimelineFlow {
crate::tenant::CreateTimelineIdempotency::FailWithConflict, // doesn't matter what we put here
None, // doesn't matter what we put here
None, // doesn't matter what we put here
ctx,
)
.context("create_timeline_struct")?;

View File

@@ -3,7 +3,6 @@ use std::sync::Arc;
use anyhow::Context;
use http_utils::error::ApiError;
use pageserver_api::models::DetachBehavior;
use pageserver_api::models::detach_ancestor::AncestorDetached;
use pageserver_api::shard::ShardIdentity;
use tokio::sync::Semaphore;
@@ -140,6 +139,15 @@ pub(crate) struct Options {
pub(crate) copy_concurrency: std::num::NonZeroUsize,
}
/// Controls the detach ancestor behavior.
/// - When set to `NoAncestorAndReparent`, we will only detach a branch if its ancestor is a root branch. It will automatically reparent any children of the ancestor before and at the branch point.
/// - When set to `MultiLevelAndNoReparent`, we will detach a branch from multiple levels of ancestors, and no reparenting will happen at all.
#[derive(Debug, Clone, Copy)]
pub enum DetachBehavior {
NoAncestorAndReparent,
MultiLevelAndNoReparent,
}
impl Default for Options {
fn default() -> Self {
Self {

View File

@@ -93,8 +93,7 @@ impl Timeline {
}
}
let ctx = RequestContext::new(TaskKind::Eviction, DownloadBehavior::Warn)
.with_scope_timeline(&self);
let ctx = RequestContext::new(TaskKind::Eviction, DownloadBehavior::Warn);
loop {
let policy = self.get_eviction_policy();
let cf = self

View File

@@ -961,8 +961,7 @@ mod tests {
}
async fn round_trip_test_compressed(blobs: &[Vec<u8>], compression: bool) -> Result<(), Error> {
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let (_temp_dir, pathbuf, offsets) =
write_maybe_compressed::<true>(blobs, compression, &ctx).await?;

View File

@@ -26,14 +26,15 @@ use owned_buffers_io::io_buf_aligned::{IoBufAligned, IoBufAlignedMut};
use owned_buffers_io::io_buf_ext::FullSlice;
use pageserver_api::config::defaults::DEFAULT_IO_BUFFER_ALIGNMENT;
pub use pageserver_api::models::virtual_file as api;
use pageserver_api::shard::TenantShardId;
use tokio::sync::{RwLock, RwLockReadGuard, RwLockWriteGuard};
use tokio::time::Instant;
use tokio_epoll_uring::{BoundedBuf, IoBuf, IoBufMut, Slice};
use crate::assert_u64_eq_usize::UsizeIsU64;
use crate::context::RequestContext;
use crate::metrics::{STORAGE_IO_TIME_METRIC, StorageIoOperation};
use crate::metrics::{STORAGE_IO_SIZE, STORAGE_IO_TIME_METRIC, StorageIoOperation};
use crate::page_cache::{PAGE_SZ, PageWriteGuard};
use crate::tenant::TENANTS_SEGMENT_NAME;
pub(crate) mod io_engine;
pub use io_engine::{
FeatureTestResult as IoEngineFeatureTestResult, feature_test as io_engine_feature_test,
@@ -120,7 +121,7 @@ impl VirtualFile {
pub async fn open_with_options<P: AsRef<Utf8Path>>(
path: P,
open_options: &OpenOptions,
ctx: &RequestContext,
ctx: &RequestContext, /* TODO: carry a pointer to the metrics in the RequestContext instead of the parsing https://github.com/neondatabase/neon/issues/6107 */
) -> Result<Self, std::io::Error> {
let inner = VirtualFileInner::open_with_options(path, open_options, ctx).await?;
Ok(VirtualFile {
@@ -132,7 +133,7 @@ impl VirtualFile {
pub async fn open_with_options_v2<P: AsRef<Utf8Path>>(
path: P,
open_options: &OpenOptions,
ctx: &RequestContext,
ctx: &RequestContext, /* TODO: carry a pointer to the metrics in the RequestContext instead of the parsing https://github.com/neondatabase/neon/issues/6107 */
) -> Result<Self, std::io::Error> {
let file = match get_io_mode() {
IoMode::Buffered => {
@@ -303,6 +304,13 @@ pub struct VirtualFileInner {
/// storing it here.
pub path: Utf8PathBuf,
open_options: OpenOptions,
// These are strings becase we only use them for metrics, and those expect strings.
// It makes no sense for us to constantly turn the `TimelineId` and `TenantId` into
// strings.
tenant_id: String,
shard_id: String,
timeline_id: String,
}
#[derive(Debug, PartialEq, Clone, Copy)]
@@ -584,16 +592,36 @@ impl VirtualFileInner {
pub async fn open_with_options<P: AsRef<Utf8Path>>(
path: P,
open_options: &OpenOptions,
_ctx: &RequestContext,
_ctx: &RequestContext, /* TODO: carry a pointer to the metrics in the RequestContext instead of the parsing https://github.com/neondatabase/neon/issues/6107 */
) -> Result<VirtualFileInner, std::io::Error> {
let path = path.as_ref();
let path_ref = path.as_ref();
let path_str = path_ref.to_string();
let parts = path_str.split('/').collect::<Vec<&str>>();
let (tenant_id, shard_id, timeline_id) =
if parts.len() > 5 && parts[parts.len() - 5] == TENANTS_SEGMENT_NAME {
let tenant_shard_part = parts[parts.len() - 4];
let (tenant_id, shard_id) = match tenant_shard_part.parse::<TenantShardId>() {
Ok(tenant_shard_id) => (
tenant_shard_id.tenant_id.to_string(),
format!("{}", tenant_shard_id.shard_slug()),
),
Err(_) => {
// Malformed path: this ID is just for observability, so tolerate it
// and pass through
(tenant_shard_part.to_string(), "*".to_string())
}
};
(tenant_id, shard_id, parts[parts.len() - 2].to_string())
} else {
("*".to_string(), "*".to_string(), "*".to_string())
};
let (handle, mut slot_guard) = get_open_files().find_victim_slot().await;
// NB: there is also StorageIoOperation::OpenAfterReplace which is for the case
// where our caller doesn't get to use the returned VirtualFile before its
// slot gets re-used by someone else.
let file = observe_duration!(StorageIoOperation::Open, {
open_options.open(path.as_std_path()).await?
open_options.open(path_ref.as_std_path()).await?
});
// Strip all options other than read and write.
@@ -609,8 +637,11 @@ impl VirtualFileInner {
let vfile = VirtualFileInner {
handle: RwLock::new(handle),
pos: 0,
path: path.to_owned(),
path: path_ref.to_path_buf(),
open_options: reopen_options,
tenant_id,
shard_id,
timeline_id,
};
// TODO: Under pressure, it's likely the slot will get re-used and
@@ -913,7 +944,7 @@ impl VirtualFileInner {
&self,
buf: tokio_epoll_uring::Slice<Buf>,
offset: u64,
ctx: &RequestContext,
_ctx: &RequestContext, /* TODO: use for metrics: https://github.com/neondatabase/neon/issues/6107 */
) -> (tokio_epoll_uring::Slice<Buf>, Result<usize, Error>)
where
Buf: tokio_epoll_uring::IoBufMut + Send,
@@ -931,7 +962,14 @@ impl VirtualFileInner {
let ((_file_guard, buf), res) = io_engine::get().read_at(file_guard, offset, buf).await;
let res = res.maybe_fatal_err("io_engine read_at inside VirtualFileInner::read_at");
if let Ok(size) = res {
ctx.io_size_metrics().read.add(size.into_u64());
STORAGE_IO_SIZE
.with_label_values(&[
"read",
&self.tenant_id,
&self.shard_id,
&self.timeline_id,
])
.add(size as i64);
}
(buf, res)
})
@@ -942,9 +980,9 @@ impl VirtualFileInner {
&self,
buf: FullSlice<B>,
offset: u64,
ctx: &RequestContext,
_ctx: &RequestContext, /* TODO: use for metrics: https://github.com/neondatabase/neon/issues/6107 */
) -> (FullSlice<B>, Result<usize, Error>) {
let (slice, result) = self.write_at_inner(buf, offset, ctx).await;
let (slice, result) = self.write_at_inner(buf, offset, _ctx).await;
let result = result.maybe_fatal_err("write_at");
(slice, result)
}
@@ -953,7 +991,7 @@ impl VirtualFileInner {
&self,
buf: FullSlice<B>,
offset: u64,
ctx: &RequestContext,
_ctx: &RequestContext, /* TODO: use for metrics: https://github.com/neondatabase/neon/issues/6107 */
) -> (FullSlice<B>, Result<usize, Error>) {
let file_guard = match self.lock_file().await {
Ok(file_guard) => file_guard,
@@ -963,7 +1001,14 @@ impl VirtualFileInner {
let ((_file_guard, buf), result) =
io_engine::get().write_at(file_guard, offset, buf).await;
if let Ok(size) = result {
ctx.io_size_metrics().write.add(size.into_u64());
STORAGE_IO_SIZE
.with_label_values(&[
"write",
&self.tenant_id,
&self.shard_id,
&self.timeline_id,
])
.add(size as i64);
}
(buf, result)
})
@@ -1299,8 +1344,9 @@ impl OwnedAsyncWriter for VirtualFile {
buf: FullSlice<Buf>,
offset: u64,
ctx: &RequestContext,
) -> (FullSlice<Buf>, std::io::Result<()>) {
VirtualFile::write_all_at(self, buf, offset, ctx).await
) -> std::io::Result<FullSlice<Buf>> {
let (buf, res) = VirtualFile::write_all_at(self, buf, offset, ctx).await;
res.map(|_| buf)
}
}
@@ -1547,8 +1593,7 @@ mod tests {
where
A: Adapter,
{
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let testdir = crate::config::PageServerConf::test_repo_dir(testname);
std::fs::create_dir_all(&testdir)?;
@@ -1675,8 +1720,7 @@ mod tests {
const THREADS: usize = 100;
const SAMPLE: [u8; SIZE] = [0xADu8; SIZE];
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let testdir = crate::config::PageServerConf::test_repo_dir("vfile_concurrency");
std::fs::create_dir_all(&testdir)?;
@@ -1735,8 +1779,7 @@ mod tests {
#[tokio::test]
async fn test_atomic_overwrite_basic() {
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let testdir = crate::config::PageServerConf::test_repo_dir("test_atomic_overwrite_basic");
std::fs::create_dir_all(&testdir).unwrap();
@@ -1764,8 +1807,7 @@ mod tests {
#[tokio::test]
async fn test_atomic_overwrite_preexisting_tmp() {
let ctx =
RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error);
let testdir =
crate::config::PageServerConf::test_repo_dir("test_atomic_overwrite_preexisting_tmp");
std::fs::create_dir_all(&testdir).unwrap();

View File

@@ -31,7 +31,7 @@ pub trait OwnedAsyncWriter {
buf: FullSlice<Buf>,
offset: u64,
ctx: &RequestContext,
) -> impl std::future::Future<Output = (FullSlice<Buf>, std::io::Result<()>)> + Send;
) -> impl std::future::Future<Output = std::io::Result<FullSlice<Buf>>> + Send;
}
/// A wrapper aorund an [`OwnedAsyncWriter`] that uses a [`Buffer`] to batch
@@ -66,7 +66,6 @@ where
buf_new: impl Fn() -> B,
gate_guard: utils::sync::gate::GateGuard,
ctx: &RequestContext,
flush_task_span: tracing::Span,
) -> Self {
Self {
writer: writer.clone(),
@@ -76,7 +75,6 @@ where
buf_new(),
gate_guard,
ctx.attached_child(),
flush_task_span,
),
bytes_submitted: 0,
}
@@ -271,12 +269,12 @@ mod tests {
buf: FullSlice<Buf>,
offset: u64,
_: &RequestContext,
) -> (FullSlice<Buf>, std::io::Result<()>) {
) -> std::io::Result<FullSlice<Buf>> {
self.writes
.lock()
.unwrap()
.push((Vec::from(&buf[..]), offset));
(buf, Ok(()))
Ok(buf)
}
}
@@ -295,7 +293,6 @@ mod tests {
|| IoBufferMut::with_capacity(2),
gate.enter()?,
ctx,
tracing::Span::none(),
);
writer.write_buffered_borrowed(b"abc", ctx).await?;

View File

@@ -1,14 +1,9 @@
use std::ops::ControlFlow;
use std::sync::Arc;
use once_cell::sync::Lazy;
use tokio_util::sync::CancellationToken;
use tracing::{Instrument, info, info_span, warn};
use utils::sync::duplex;
use super::{Buffer, CheapCloneForRead, OwnedAsyncWriter};
use crate::context::RequestContext;
use crate::virtual_file::MaybeFatalIo;
use crate::virtual_file::owned_buffers_io::io_buf_aligned::IoBufAligned;
use crate::virtual_file::owned_buffers_io::io_buf_ext::FullSlice;
@@ -123,7 +118,6 @@ where
buf: B,
gate_guard: utils::sync::gate::GateGuard,
ctx: RequestContext,
span: tracing::Span,
) -> Self
where
B: Buffer<IoBuf = Buf> + Send + 'static,
@@ -131,14 +125,11 @@ where
// It is fine to buffer up to only 1 message. We only 1 message in-flight at a time.
let (front, back) = duplex::mpsc::channel(1);
let join_handle = tokio::spawn(
async move {
FlushBackgroundTask::new(back, file, gate_guard, ctx)
.run(buf.flush())
.await
}
.instrument(span),
);
let join_handle = tokio::spawn(async move {
FlushBackgroundTask::new(back, file, gate_guard, ctx)
.run(buf.flush())
.await
});
FlushHandle {
inner: Some(FlushHandleInner {
@@ -245,7 +236,6 @@ where
/// The passed in slice is immediately sent back to the flush handle through the duplex channel.
async fn run(mut self, slice: FullSlice<Buf>) -> std::io::Result<Arc<W>> {
// Sends the extra buffer back to the handle.
// TODO: can this ever await and or fail? I think not.
self.channel.send(slice).await.map_err(|_| {
std::io::Error::new(std::io::ErrorKind::BrokenPipe, "flush handle closed early")
})?;
@@ -261,47 +251,10 @@ where
}
// Write slice to disk at `offset`.
//
// Error handling happens according to the current policy of crashing
// on fatal IO errors and retrying in place otherwise (deeming all other errors retryable).
// (The upper layers of the Pageserver write path are not equipped to retry write errors
// becasuse they often deallocate the buffers that were already written).
//
// TODO: cancellation sensitiity.
// Without it, if we hit a bug where retrying is never successful,
// then we can't shut down the timeline/tenant/pageserver cleanly because
// layers of the Pageserver write path are holding the gate open for EphemeralFile.
//
// TODO: use utils::backoff::retry once async closures are actually usable
//
let mut slice_storage = Some(request.slice);
for attempt in 1.. {
let result = async {
if attempt > 1 {
info!("retrying flush");
}
let slice = slice_storage.take().expect(
"likely previous invocation of this future didn't get polled to completion",
);
let (slice, res) = self.writer.write_all_at(slice, request.offset, &self.ctx).await;
slice_storage = Some(slice);
let res = res.maybe_fatal_err("owned_buffers_io flush");
let Err(err) = res else {
return ControlFlow::Break(());
};
warn!(%err, "error flushing buffered writer buffer to disk, retrying after backoff");
static NO_CANCELLATION: Lazy<CancellationToken> = Lazy::new(CancellationToken::new);
utils::backoff::exponential_backoff(attempt, 1.0, 10.0, &NO_CANCELLATION).await;
ControlFlow::Continue(())
}
.instrument(info_span!("flush_attempt", %attempt))
.await;
match result {
ControlFlow::Break(()) => break,
ControlFlow::Continue(()) => continue,
}
}
let slice = slice_storage.expect("loop must have run at least once");
let slice = self
.writer
.write_all_at(request.slice, request.offset, &self.ctx)
.await?;
#[cfg(test)]
{

View File

@@ -76,10 +76,6 @@
#include "access/xlogrecovery.h"
#endif
#if PG_VERSION_NUM < 160000
typedef PGAlignedBlock PGIOAlignedBlock;
#endif
/*
* If DEBUG_COMPARE_LOCAL is defined, we pass through all the SMGR API
* calls to md.c, and *also* do the calls to the Page Server. On every
@@ -1807,7 +1803,7 @@ static XLogRecPtr
log_newpage_copy(NRelFileInfo * rinfo, ForkNumber forkNum, BlockNumber blkno,
Page page, bool page_std)
{
PGIOAlignedBlock copied_buffer;
PGAlignedBlock copied_buffer;
memcpy(copied_buffer.data, page, BLCKSZ);
return log_newpage(rinfo, forkNum, blkno, copied_buffer.data, page_std);
@@ -1824,7 +1820,7 @@ static XLogRecPtr
log_newpages_copy(NRelFileInfo * rinfo, ForkNumber forkNum, BlockNumber blkno,
BlockNumber nblocks, Page *pages, bool page_std)
{
PGIOAlignedBlock copied_buffer[XLR_MAX_BLOCK_ID];
PGAlignedBlock copied_buffer[XLR_MAX_BLOCK_ID];
BlockNumber blknos[XLR_MAX_BLOCK_ID];
Page pageptrs[XLR_MAX_BLOCK_ID];
int nregistered = 0;
@@ -1862,7 +1858,7 @@ log_newpages_copy(NRelFileInfo * rinfo, ForkNumber forkNum, BlockNumber blkno,
static bool
PageIsEmptyHeapPage(char *buffer)
{
PGIOAlignedBlock empty_page;
PGAlignedBlock empty_page;
PageInit((Page) empty_page.data, BLCKSZ, 0);
@@ -2851,7 +2847,7 @@ static void
neon_zeroextend(SMgrRelation reln, ForkNumber forkNum, BlockNumber blocknum,
int nblocks, bool skipFsync)
{
const PGIOAlignedBlock buffer = {0};
const PGAlignedBlock buffer = {0};
int remblocks = nblocks;
XLogRecPtr lsn = 0;
@@ -2898,11 +2894,6 @@ neon_zeroextend(SMgrRelation reln, ForkNumber forkNum, BlockNumber blocknum,
relpath(reln->smgr_rlocator, forkNum),
InvalidBlockNumber)));
#ifdef DEBUG_COMPARE_LOCAL
if (IS_LOCAL_REL(reln))
mdzeroextend(reln, forkNum, blocknum, nblocks, skipFsync);
#endif
/* Don't log any pages if we're not allowed to do so. */
if (!XLogInsertAllowed())
return;
@@ -3398,16 +3389,15 @@ neon_read(SMgrRelation reln, ForkNumber forkNum, BlockNumber blkno, void *buffer
if (forkNum == MAIN_FORKNUM && IS_LOCAL_REL(reln))
{
char pageserver_masked[BLCKSZ];
PGIOAlignedBlock mdbuf;
PGIOAlignedBlock mdbuf_masked;
XLogRecPtr request_lsn = request_lsns.request_lsn;
char mdbuf[BLCKSZ];
char mdbuf_masked[BLCKSZ];
mdread(reln, forkNum, blkno, mdbuf.data);
mdread(reln, forkNum, blkno, mdbuf);
memcpy(pageserver_masked, buffer, BLCKSZ);
memcpy(mdbuf_masked.data, mdbuf.data, BLCKSZ);
memcpy(mdbuf_masked, mdbuf, BLCKSZ);
if (PageIsNew((Page) mdbuf.data))
if (PageIsNew((Page) mdbuf))
{
if (!PageIsNew((Page) pageserver_masked))
{
@@ -3426,41 +3416,41 @@ neon_read(SMgrRelation reln, ForkNumber forkNum, BlockNumber blkno, void *buffer
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf.data));
hexdump_page(mdbuf));
}
else if (PageGetSpecialSize(mdbuf.data) == 0)
else if (PageGetSpecialSize(mdbuf) == 0)
{
/* assume heap */
RmgrTable[RM_HEAP_ID].rm_mask(mdbuf_masked.data, blkno);
RmgrTable[RM_HEAP_ID].rm_mask(mdbuf_masked, blkno);
RmgrTable[RM_HEAP_ID].rm_mask(pageserver_masked, blkno);
if (memcmp(mdbuf_masked.data, pageserver_masked, BLCKSZ) != 0)
if (memcmp(mdbuf_masked, pageserver_masked, BLCKSZ) != 0)
{
neon_log(PANIC, "heap buffers differ at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n------ MD ------\n%s\n------ Page Server ------\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf_masked.data),
hexdump_page(mdbuf_masked),
hexdump_page(pageserver_masked));
}
}
else if (PageGetSpecialSize(mdbuf.data) == MAXALIGN(sizeof(BTPageOpaqueData)))
else if (PageGetSpecialSize(mdbuf) == MAXALIGN(sizeof(BTPageOpaqueData)))
{
if (((BTPageOpaqueData *) PageGetSpecialPointer(mdbuf.data))->btpo_cycleid < MAX_BT_CYCLE_ID)
if (((BTPageOpaqueData *) PageGetSpecialPointer(mdbuf))->btpo_cycleid < MAX_BT_CYCLE_ID)
{
/* assume btree */
RmgrTable[RM_BTREE_ID].rm_mask(mdbuf_masked.data, blkno);
RmgrTable[RM_BTREE_ID].rm_mask(mdbuf_masked, blkno);
RmgrTable[RM_BTREE_ID].rm_mask(pageserver_masked, blkno);
if (memcmp(mdbuf_masked.data, pageserver_masked, BLCKSZ) != 0)
if (memcmp(mdbuf_masked, pageserver_masked, BLCKSZ) != 0)
{
neon_log(PANIC, "btree buffers differ at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n------ MD ------\n%s\n------ Page Server ------\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf_masked.data),
hexdump_page(mdbuf_masked),
hexdump_page(pageserver_masked));
}
}
@@ -3552,85 +3542,77 @@ neon_readv(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum,
prefetch_pump_state(false);
#ifdef DEBUG_COMPARE_LOCAL
if (forknum == MAIN_FORKNUM && IS_LOCAL_REL(reln))
if (forkNum == MAIN_FORKNUM && IS_LOCAL_REL(reln))
{
char pageserver_masked[BLCKSZ];
PGIOAlignedBlock mdbuf;
PGIOAlignedBlock mdbuf_masked;
XLogRecPtr request_lsn = request_lsns->request_lsn;
char mdbuf[BLCKSZ];
char mdbuf_masked[BLCKSZ];
for (int i = 0; i < nblocks; i++)
{
BlockNumber blkno = blocknum + i;
if (!BITMAP_ISSET(read, i))
continue;
#if PG_MAJORVERSION_NUM >= 17
{
void* mdbuffers[1] = { mdbuf.data };
mdreadv(reln, forknum, blkno, mdbuffers, 1);
}
mdreadv(reln, forkNum, blkno + i, &mdbuf, 1);
#else
mdread(reln, forknum, blkno, mdbuf.data);
mdread(reln, forkNum, blkno + i, mdbuf);
#endif
memcpy(pageserver_masked, buffers[i], BLCKSZ);
memcpy(mdbuf_masked.data, mdbuf.data, BLCKSZ);
memcpy(pageserver_masked, buffer, BLCKSZ);
memcpy(mdbuf_masked, mdbuf, BLCKSZ);
if (PageIsNew((Page) mdbuf.data))
if (PageIsNew((Page) mdbuf))
{
if (!PageIsNew((Page) pageserver_masked))
{
neon_log(PANIC, "page is new in MD but not in Page Server at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forknum,
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(buffers[i]));
hexdump_page(buffer));
}
}
else if (PageIsNew((Page) buffers[i]))
else if (PageIsNew((Page) buffer))
{
neon_log(PANIC, "page is new in Page Server but not in MD at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forknum,
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf.data));
hexdump_page(mdbuf));
}
else if (PageGetSpecialSize(mdbuf.data) == 0)
else if (PageGetSpecialSize(mdbuf) == 0)
{
/* assume heap */
RmgrTable[RM_HEAP_ID].rm_mask(mdbuf_masked.data, blkno);
RmgrTable[RM_HEAP_ID].rm_mask(mdbuf_masked, blkno);
RmgrTable[RM_HEAP_ID].rm_mask(pageserver_masked, blkno);
if (memcmp(mdbuf_masked.data, pageserver_masked, BLCKSZ) != 0)
if (memcmp(mdbuf_masked, pageserver_masked, BLCKSZ) != 0)
{
neon_log(PANIC, "heap buffers differ at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n------ MD ------\n%s\n------ Page Server ------\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forknum,
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf_masked.data),
hexdump_page(mdbuf_masked),
hexdump_page(pageserver_masked));
}
}
else if (PageGetSpecialSize(mdbuf.data) == MAXALIGN(sizeof(BTPageOpaqueData)))
else if (PageGetSpecialSize(mdbuf) == MAXALIGN(sizeof(BTPageOpaqueData)))
{
if (((BTPageOpaqueData *) PageGetSpecialPointer(mdbuf.data))->btpo_cycleid < MAX_BT_CYCLE_ID)
if (((BTPageOpaqueData *) PageGetSpecialPointer(mdbuf))->btpo_cycleid < MAX_BT_CYCLE_ID)
{
/* assume btree */
RmgrTable[RM_BTREE_ID].rm_mask(mdbuf_masked.data, blkno);
RmgrTable[RM_BTREE_ID].rm_mask(mdbuf_masked, blkno);
RmgrTable[RM_BTREE_ID].rm_mask(pageserver_masked, blkno);
if (memcmp(mdbuf_masked.data, pageserver_masked, BLCKSZ) != 0)
if (memcmp(mdbuf_masked, pageserver_masked, BLCKSZ) != 0)
{
neon_log(PANIC, "btree buffers differ at blk %u in rel %u/%u/%u fork %u (request LSN %X/%08X):\n------ MD ------\n%s\n------ Page Server ------\n%s\n",
blkno,
RelFileInfoFmt(InfoFromSMgrRel(reln)),
forknum,
forkNum,
(uint32) (request_lsn >> 32), (uint32) request_lsn,
hexdump_page(mdbuf_masked.data),
hexdump_page(mdbuf_masked),
hexdump_page(pageserver_masked));
}
}
@@ -3682,7 +3664,6 @@ neon_write(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, const vo
switch (reln->smgr_relpersistence)
{
case 0:
#ifndef DEBUG_COMPARE_LOCAL
/* This is a bit tricky. Check if the relation exists locally */
if (mdexists(reln, forknum))
{
@@ -3701,7 +3682,6 @@ neon_write(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, const vo
*/
return;
}
#endif
break;
case RELPERSISTENCE_PERMANENT:
@@ -3752,7 +3732,6 @@ neon_writev(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno,
switch (reln->smgr_relpersistence)
{
case 0:
#ifndef DEBUG_COMPARE_LOCAL
/* This is a bit tricky. Check if the relation exists locally */
if (mdexists(reln, forknum))
{
@@ -3768,7 +3747,6 @@ neon_writev(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno,
*/
return;
}
#endif
break;
case RELPERSISTENCE_PERMANENT:
@@ -3790,7 +3768,7 @@ neon_writev(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno,
#ifdef DEBUG_COMPARE_LOCAL
if (IS_LOCAL_REL(reln))
mdwritev(reln, forknum, blkno, buffers, nblocks, skipFsync);
mdwritev(reln, forknum, blocknum, &buffer, 1, skipFsync);
#endif
}
@@ -4176,10 +4154,8 @@ neon_start_unlogged_build(SMgrRelation reln)
* FIXME: should we pass isRedo true to create the tablespace dir if it
* doesn't exist? Is it needed?
*/
#ifndef DEBUG_COMPARE_LOCAL
if (!IsParallelWorker())
if (!IsParallelWorker())
mdcreate(reln, MAIN_FORKNUM, false);
#endif
}
/*
@@ -4254,10 +4230,8 @@ neon_end_unlogged_build(SMgrRelation reln)
forget_cached_relsize(InfoFromNInfoB(rinfob), forknum);
mdclose(reln, forknum);
#ifndef DEBUG_COMPARE_LOCAL
/* use isRedo == true, so that we drop it immediately */
mdunlink(rinfob, forknum, true);
#endif
}
}

View File

@@ -57,11 +57,10 @@ static void SendProposerGreeting(Safekeeper *sk);
static void RecvAcceptorGreeting(Safekeeper *sk);
static void SendVoteRequest(Safekeeper *sk);
static void RecvVoteResponse(Safekeeper *sk);
static bool VotesCollected(WalProposer *wp);
static void HandleElectedProposer(WalProposer *wp);
static term_t GetHighestTerm(TermHistory *th);
static term_t GetLastLogTerm(Safekeeper *sk);
static void ProcessPropStartPos(WalProposer *wp);
static term_t GetEpoch(Safekeeper *sk);
static void DetermineEpochStartLsn(WalProposer *wp);
static void SendProposerElected(Safekeeper *sk);
static void StartStreaming(Safekeeper *sk);
static void SendMessageToNode(Safekeeper *sk);
@@ -98,7 +97,6 @@ WalProposerCreate(WalProposerConfig *config, walproposer_api api)
wp = palloc0(sizeof(WalProposer));
wp->config = config;
wp->api = api;
wp->state = WPS_COLLECTING_TERMS;
wp_log(LOG, "neon.safekeepers=%s", wp->config->safekeepers_list);
@@ -520,7 +518,7 @@ AdvancePollState(Safekeeper *sk, uint32 events)
* nodes are transferred from SS_VOTING to sending actual vote
* requests.
*/
case SS_WAIT_VOTING:
case SS_VOTING:
wp_log(WARNING, "EOF from node %s:%s in %s state", sk->host,
sk->port, FormatSafekeeperState(sk));
ResetConnection(sk);
@@ -549,7 +547,7 @@ AdvancePollState(Safekeeper *sk, uint32 events)
/*
* Idle state for waiting votes from quorum.
*/
case SS_WAIT_ELECTED:
case SS_IDLE:
wp_log(WARNING, "EOF from node %s:%s in %s state", sk->host,
sk->port, FormatSafekeeperState(sk));
ResetConnection(sk);
@@ -723,15 +721,6 @@ SendProposerGreeting(Safekeeper *sk)
BlockingWrite(sk, sk->outbuf.data, sk->outbuf.len, SS_HANDSHAKE_RECV);
}
/*
* Have we received greeting from enough (quorum) safekeepers to start voting?
*/
static bool
TermsCollected(WalProposer *wp)
{
return wp->n_connected >= wp->quorum;
}
static void
RecvAcceptorGreeting(Safekeeper *sk)
{
@@ -765,7 +754,7 @@ RecvAcceptorGreeting(Safekeeper *sk)
}
/* Protocol is all good, move to voting. */
sk->state = SS_WAIT_VOTING;
sk->state = SS_VOTING;
/*
* Note: it would be better to track the counter on per safekeeper basis,
@@ -773,18 +762,17 @@ RecvAcceptorGreeting(Safekeeper *sk)
* as is for now.
*/
++wp->n_connected;
if (wp->state == WPS_COLLECTING_TERMS)
if (wp->n_connected <= wp->quorum)
{
/* We're still collecting terms from the majority. */
wp->propTerm = Max(sk->greetResponse.term, wp->propTerm);
/* Quorum is acquried, prepare the vote request. */
if (TermsCollected(wp))
if (wp->n_connected == wp->quorum)
{
wp->propTerm++;
wp_log(LOG, "proposer connected to quorum (%d) safekeepers, propTerm=" INT64_FORMAT, wp->quorum, wp->propTerm);
wp->state = WPS_CAMPAIGN;
wp->voteRequest.pam.tag = 'v';
wp->voteRequest.generation = wp->mconf.generation;
wp->voteRequest.term = wp->propTerm;
@@ -799,10 +787,12 @@ RecvAcceptorGreeting(Safekeeper *sk)
}
/*
* If we have quorum, start (or just send vote request to newly connected
* node) election, otherwise wait until we have more greetings.
* Check if we have quorum. If there aren't enough safekeepers, wait and
* do nothing. We'll eventually get a task when the election starts.
*
* If we do have quorum, we can start an election.
*/
if (wp->state == WPS_COLLECTING_TERMS)
if (wp->n_connected < wp->quorum)
{
/*
* SS_VOTING is an idle state; read-ready indicates the connection
@@ -817,7 +807,11 @@ RecvAcceptorGreeting(Safekeeper *sk)
*/
for (int j = 0; j < wp->n_safekeepers; j++)
{
if (wp->safekeeper[j].state == SS_WAIT_VOTING)
/*
* Remember: SS_VOTING indicates that the safekeeper is
* participating in voting, but hasn't sent anything yet.
*/
if (wp->safekeeper[j].state == SS_VOTING)
SendVoteRequest(&wp->safekeeper[j]);
}
}
@@ -844,8 +838,6 @@ RecvVoteResponse(Safekeeper *sk)
{
WalProposer *wp = sk->wp;
Assert(wp->state >= WPS_CAMPAIGN);
sk->voteResponse.apm.tag = 'v';
if (!AsyncReadMessage(sk, (AcceptorProposerMessage *) &sk->voteResponse))
return;
@@ -864,7 +856,7 @@ RecvVoteResponse(Safekeeper *sk)
* we are not elected yet and thus need the vote.
*/
if ((!sk->voteResponse.voteGiven) &&
(sk->voteResponse.term > wp->propTerm || wp->state == WPS_CAMPAIGN))
(sk->voteResponse.term > wp->propTerm || wp->n_votes < wp->quorum))
{
wp_log(FATAL, "WAL acceptor %s:%s with term " INT64_FORMAT " rejects our connection request with term " INT64_FORMAT "",
sk->host, sk->port,
@@ -872,83 +864,38 @@ RecvVoteResponse(Safekeeper *sk)
}
Assert(sk->voteResponse.term == wp->propTerm);
/* ready for elected message */
sk->state = SS_WAIT_ELECTED;
/* Handshake completed, do we have quorum? */
wp->n_votes++;
/* Are we already elected? */
if (wp->state == WPS_CAMPAIGN)
if (wp->n_votes < wp->quorum)
{
/* no; check if this vote makes us elected */
if (VotesCollected(wp))
{
wp->state = WPS_ELECTED;
HandleElectedProposer(wp);
}
else
{
/* can't do much yet, no quorum */
return;
}
sk->state = SS_IDLE; /* can't do much yet, no quorum */
}
else if (wp->n_votes > wp->quorum)
{
/* already elected, start streaming */
SendProposerElected(sk);
}
else
{
Assert(wp->state == WPS_ELECTED);
/* send elected only to this sk */
SendProposerElected(sk);
sk->state = SS_IDLE;
/* Idle state waits for read-ready events */
wp->api.update_event_set(sk, WL_SOCKET_READABLE);
HandleElectedProposer(sk->wp);
}
}
/*
* Checks if enough votes has been collected to get elected and if that's the
* case finds the highest vote, setting donor, donorLastLogTerm,
* propTermStartLsn fields. Also sets truncateLsn.
*/
static bool
VotesCollected(WalProposer *wp)
{
int n_ready = 0;
/* assumed to be called only when not elected yet */
Assert(wp->state == WPS_CAMPAIGN);
wp->propTermStartLsn = InvalidXLogRecPtr;
wp->donorLastLogTerm = 0;
wp->truncateLsn = InvalidXLogRecPtr;
for (int i = 0; i < wp->n_safekeepers; i++)
{
if (wp->safekeeper[i].state == SS_WAIT_ELECTED)
{
n_ready++;
if (GetLastLogTerm(&wp->safekeeper[i]) > wp->donorLastLogTerm ||
(GetLastLogTerm(&wp->safekeeper[i]) == wp->donorLastLogTerm &&
wp->safekeeper[i].voteResponse.flushLsn > wp->propTermStartLsn))
{
wp->donorLastLogTerm = GetLastLogTerm(&wp->safekeeper[i]);
wp->propTermStartLsn = wp->safekeeper[i].voteResponse.flushLsn;
wp->donor = i;
}
wp->truncateLsn = Max(wp->safekeeper[i].voteResponse.truncateLsn, wp->truncateLsn);
}
}
return n_ready >= wp->quorum;
}
/*
* Called once a majority of acceptors have voted for us and current proposer
* has been elected.
*
* Sends ProposerElected message to all acceptors in SS_WAIT_ELECTED state and starts
* Sends ProposerElected message to all acceptors in SS_IDLE state and starts
* replication from walsender.
*/
static void
HandleElectedProposer(WalProposer *wp)
{
ProcessPropStartPos(wp);
Assert(wp->propTermStartLsn != InvalidXLogRecPtr);
DetermineEpochStartLsn(wp);
/*
* Synchronously download WAL from the most advanced safekeeper. We do
@@ -960,24 +907,40 @@ HandleElectedProposer(WalProposer *wp)
wp_log(FATAL, "failed to download WAL for logical replicaiton");
}
if (wp->truncateLsn == wp->propTermStartLsn && wp->config->syncSafekeepers)
/*
* Zero propEpochStartLsn means majority of safekeepers doesn't have any
* WAL, timeline was just created. Compute bumps it to basebackup LSN,
* otherwise we must be sync-safekeepers and we have nothing to do then.
*
* Proceeding is not only pointless but harmful, because we'd give
* safekeepers term history starting with 0/0. These hacks will go away
* once we disable implicit timeline creation on safekeepers and create it
* with non zero LSN from the start.
*/
if (wp->propEpochStartLsn == InvalidXLogRecPtr)
{
Assert(wp->config->syncSafekeepers);
wp_log(LOG, "elected with zero propEpochStartLsn in sync-safekeepers, exiting");
wp->api.finish_sync_safekeepers(wp, wp->propEpochStartLsn);
}
if (wp->truncateLsn == wp->propEpochStartLsn && wp->config->syncSafekeepers)
{
/* Sync is not needed: just exit */
wp->api.finish_sync_safekeepers(wp, wp->propTermStartLsn);
wp->api.finish_sync_safekeepers(wp, wp->propEpochStartLsn);
/* unreachable */
}
for (int i = 0; i < wp->n_safekeepers; i++)
{
if (wp->safekeeper[i].state == SS_WAIT_ELECTED)
if (wp->safekeeper[i].state == SS_IDLE)
SendProposerElected(&wp->safekeeper[i]);
}
/*
* The proposer has been elected, and there will be no quorum waiting
* after this point. There will be no safekeeper with state
* SS_WAIT_ELECTED also, because that state is used only for quorum
* waiting.
* after this point. There will be no safekeeper with state SS_IDLE also,
* because that state is used only for quorum waiting.
*/
if (wp->config->syncSafekeepers)
@@ -994,7 +957,7 @@ HandleElectedProposer(WalProposer *wp)
return;
}
wp->api.start_streaming(wp, wp->propTermStartLsn);
wp->api.start_streaming(wp, wp->propEpochStartLsn);
/* Should not return here */
}
@@ -1007,7 +970,7 @@ GetHighestTerm(TermHistory *th)
/* safekeeper's epoch is the term of the highest entry in the log */
static term_t
GetLastLogTerm(Safekeeper *sk)
GetEpoch(Safekeeper *sk)
{
return GetHighestTerm(&sk->voteResponse.termHistory);
}
@@ -1028,52 +991,72 @@ SkipXLogPageHeader(WalProposer *wp, XLogRecPtr lsn)
}
/*
* Called after quorum gave votes and proposer starting position (highest vote
* term + flush LSN) -- is determined (VotesCollected true), this function
* adopts it: pushes LSN to shmem, sets wp term history, verifies that the
* basebackup matches.
* Called after majority of acceptors gave votes, it calculates the most
* advanced safekeeper (who will be the donor) and epochStartLsn -- LSN since
* which we'll write WAL in our term.
*
* Sets truncateLsn along the way (though it is not of much use at this point --
* only for skipping recovery).
*/
static void
ProcessPropStartPos(WalProposer *wp)
DetermineEpochStartLsn(WalProposer *wp)
{
TermHistory *dth;
int n_ready = 0;
WalproposerShmemState *walprop_shared;
/* must have collected votes */
Assert(wp->state == WPS_ELECTED);
wp->propEpochStartLsn = InvalidXLogRecPtr;
wp->donorEpoch = 0;
wp->truncateLsn = InvalidXLogRecPtr;
/*
* If propTermStartLsn is 0, it means flushLsn is 0 everywhere, we are
* bootstrapping and nothing was committed yet. Start streaming from the
* basebackup LSN then.
*
* In case of sync-safekeepers just exit: proceeding is not only pointless
* but harmful, because we'd give safekeepers term history starting with
* 0/0. These hacks will go away once we disable implicit timeline
* creation on safekeepers and create it with non zero LSN from the start.
*/
if (wp->propTermStartLsn == InvalidXLogRecPtr)
for (int i = 0; i < wp->n_safekeepers; i++)
{
if (!wp->config->syncSafekeepers)
if (wp->safekeeper[i].state == SS_IDLE)
{
wp->propTermStartLsn = wp->truncateLsn = wp->api.get_redo_start_lsn(wp);
wp_log(LOG, "bumped epochStartLsn to the first record %X/%X", LSN_FORMAT_ARGS(wp->propTermStartLsn));
}
else
{
wp_log(LOG, "elected with zero propTermStartLsn in sync-safekeepers, exiting");
wp->api.finish_sync_safekeepers(wp, wp->propTermStartLsn);
n_ready++;
if (GetEpoch(&wp->safekeeper[i]) > wp->donorEpoch ||
(GetEpoch(&wp->safekeeper[i]) == wp->donorEpoch &&
wp->safekeeper[i].voteResponse.flushLsn > wp->propEpochStartLsn))
{
wp->donorEpoch = GetEpoch(&wp->safekeeper[i]);
wp->propEpochStartLsn = wp->safekeeper[i].voteResponse.flushLsn;
wp->donor = i;
}
wp->truncateLsn = Max(wp->safekeeper[i].voteResponse.truncateLsn, wp->truncateLsn);
}
}
pg_atomic_write_u64(&wp->api.get_shmem_state(wp)->propEpochStartLsn, wp->propTermStartLsn);
if (n_ready < wp->quorum)
{
/*
* This is a rare case that can be triggered if safekeeper has voted
* and disconnected. In this case, its state will not be SS_IDLE and
* its vote cannot be used, because we clean up `voteResponse` in
* `ShutdownConnection`.
*/
wp_log(FATAL, "missing majority of votes, collected %d, expected %d, got %d", wp->n_votes, wp->quorum, n_ready);
}
/*
* If propEpochStartLsn is 0, it means flushLsn is 0 everywhere, we are
* bootstrapping and nothing was committed yet. Start streaming then from
* the basebackup LSN.
*/
if (wp->propEpochStartLsn == InvalidXLogRecPtr && !wp->config->syncSafekeepers)
{
wp->propEpochStartLsn = wp->truncateLsn = wp->api.get_redo_start_lsn(wp);
wp_log(LOG, "bumped epochStartLsn to the first record %X/%X", LSN_FORMAT_ARGS(wp->propEpochStartLsn));
}
pg_atomic_write_u64(&wp->api.get_shmem_state(wp)->propEpochStartLsn, wp->propEpochStartLsn);
Assert(wp->truncateLsn != InvalidXLogRecPtr || wp->config->syncSafekeepers);
/*
* We will be generating WAL since propTermStartLsn, so we should set
* We will be generating WAL since propEpochStartLsn, so we should set
* availableLsn to mark this LSN as the latest available position.
*/
wp->availableLsn = wp->propTermStartLsn;
wp->availableLsn = wp->propEpochStartLsn;
/*
* Proposer's term history is the donor's + its own entry.
@@ -1084,12 +1067,12 @@ ProcessPropStartPos(WalProposer *wp)
if (dth->n_entries > 0)
memcpy(wp->propTermHistory.entries, dth->entries, sizeof(TermSwitchEntry) * dth->n_entries);
wp->propTermHistory.entries[wp->propTermHistory.n_entries - 1].term = wp->propTerm;
wp->propTermHistory.entries[wp->propTermHistory.n_entries - 1].lsn = wp->propTermStartLsn;
wp->propTermHistory.entries[wp->propTermHistory.n_entries - 1].lsn = wp->propEpochStartLsn;
wp_log(LOG, "got votes from majority (%d) of nodes, term " UINT64_FORMAT ", epochStartLsn %X/%X, donor %s:%s, truncate_lsn %X/%X",
wp->quorum,
wp->propTerm,
LSN_FORMAT_ARGS(wp->propTermStartLsn),
LSN_FORMAT_ARGS(wp->propEpochStartLsn),
wp->safekeeper[wp->donor].host, wp->safekeeper[wp->donor].port,
LSN_FORMAT_ARGS(wp->truncateLsn));
@@ -1107,7 +1090,7 @@ ProcessPropStartPos(WalProposer *wp)
* Safekeepers don't skip header as they need continious stream of
* data, so correct LSN for comparison.
*/
if (SkipXLogPageHeader(wp, wp->propTermStartLsn) != wp->api.get_redo_start_lsn(wp))
if (SkipXLogPageHeader(wp, wp->propEpochStartLsn) != wp->api.get_redo_start_lsn(wp))
{
/*
* However, allow to proceed if last_log_term on the node which
@@ -1128,8 +1111,8 @@ ProcessPropStartPos(WalProposer *wp)
*/
disable_core_dump();
wp_log(PANIC,
"collected propTermStartLsn %X/%X, but basebackup LSN %X/%X",
LSN_FORMAT_ARGS(wp->propTermStartLsn),
"collected propEpochStartLsn %X/%X, but basebackup LSN %X/%X",
LSN_FORMAT_ARGS(wp->propEpochStartLsn),
LSN_FORMAT_ARGS(wp->api.get_redo_start_lsn(wp)));
}
}
@@ -1640,7 +1623,7 @@ GetAcknowledgedByQuorumWALPosition(WalProposer *wp)
* Like in Raft, we aren't allowed to commit entries from previous
* terms, so ignore reported LSN until it gets to epochStartLsn.
*/
responses[i] = wp->safekeeper[i].appendResponse.flushLsn >= wp->propTermStartLsn ? wp->safekeeper[i].appendResponse.flushLsn : 0;
responses[i] = wp->safekeeper[i].appendResponse.flushLsn >= wp->propEpochStartLsn ? wp->safekeeper[i].appendResponse.flushLsn : 0;
}
qsort(responses, wp->n_safekeepers, sizeof(XLogRecPtr), CompareLsn);
@@ -1673,10 +1656,10 @@ UpdateDonorShmem(WalProposer *wp)
* about its position immediately after election before any feedbacks are
* sent.
*/
if (wp->safekeeper[wp->donor].state >= SS_WAIT_ELECTED)
if (wp->safekeeper[wp->donor].state >= SS_IDLE)
{
donor = &wp->safekeeper[wp->donor];
donor_lsn = wp->propTermStartLsn;
donor_lsn = wp->propEpochStartLsn;
}
/*
@@ -1765,7 +1748,7 @@ HandleSafekeeperResponse(WalProposer *wp, Safekeeper *fromsk)
for (int i = 0; i < wp->n_safekeepers; i++)
{
Safekeeper *sk = &wp->safekeeper[i];
bool synced = sk->appendResponse.commitLsn >= wp->propTermStartLsn;
bool synced = sk->appendResponse.commitLsn >= wp->propEpochStartLsn;
/* alive safekeeper which is not synced yet; wait for it */
if (sk->state != SS_OFFLINE && !synced)
@@ -1789,7 +1772,7 @@ HandleSafekeeperResponse(WalProposer *wp, Safekeeper *fromsk)
*/
BroadcastAppendRequest(wp);
wp->api.finish_sync_safekeepers(wp, wp->propTermStartLsn);
wp->api.finish_sync_safekeepers(wp, wp->propEpochStartLsn);
/* unreachable */
}
}
@@ -2395,7 +2378,7 @@ FormatSafekeeperState(Safekeeper *sk)
case SS_HANDSHAKE_RECV:
return_val = "handshake (receiving)";
break;
case SS_WAIT_VOTING:
case SS_VOTING:
return_val = "voting";
break;
case SS_WAIT_VERDICT:
@@ -2404,7 +2387,7 @@ FormatSafekeeperState(Safekeeper *sk)
case SS_SEND_ELECTED_FLUSH:
return_val = "send-announcement-flush";
break;
case SS_WAIT_ELECTED:
case SS_IDLE:
return_val = "idle";
break;
case SS_ACTIVE:
@@ -2493,8 +2476,8 @@ SafekeeperStateDesiredEvents(Safekeeper *sk, uint32 *sk_events, uint32 *nwr_even
* Idle states use read-readiness as a sign that the connection
* has been disconnected.
*/
case SS_WAIT_VOTING:
case SS_WAIT_ELECTED:
case SS_VOTING:
case SS_IDLE:
*sk_events = WL_SOCKET_READABLE;
return;

View File

@@ -73,12 +73,12 @@ typedef enum
* Moved externally by execution of SS_HANDSHAKE_RECV, when we received a
* quorum of handshakes.
*/
SS_WAIT_VOTING,
SS_VOTING,
/*
* Already sent voting information, waiting to receive confirmation from
* the node. After receiving, moves to SS_WAIT_ELECTED, if the quorum
* isn't reached yet.
* the node. After receiving, moves to SS_IDLE, if the quorum isn't
* reached yet.
*/
SS_WAIT_VERDICT,
@@ -91,7 +91,7 @@ typedef enum
*
* Moves to SS_ACTIVE only by call to StartStreaming.
*/
SS_WAIT_ELECTED,
SS_IDLE,
/*
* Active phase, when we acquired quorum and have WAL to send or feedback
@@ -751,15 +751,6 @@ typedef struct WalProposerConfig
#endif
} WalProposerConfig;
typedef enum
{
/* collecting greetings to determine term to campaign for */
WPS_COLLECTING_TERMS,
/* campaing started, waiting for votes */
WPS_CAMPAIGN,
/* successfully elected */
WPS_ELECTED,
} WalProposerState;
/*
* WAL proposer state.
@@ -767,7 +758,6 @@ typedef enum
typedef struct WalProposer
{
WalProposerConfig *config;
WalProposerState state;
/* Current walproposer membership configuration */
MembershipConfiguration mconf;
@@ -823,10 +813,10 @@ typedef struct WalProposer
TermHistory propTermHistory;
/* epoch start lsn of the proposer */
XLogRecPtr propTermStartLsn;
XLogRecPtr propEpochStartLsn;
/* Most advanced acceptor epoch */
term_t donorLastLogTerm;
term_t donorEpoch;
/* Most advanced acceptor */
int donor;

View File

@@ -1496,7 +1496,7 @@ walprop_pg_wal_reader_allocate(Safekeeper *sk)
snprintf(log_prefix, sizeof(log_prefix), WP_LOG_PREFIX "sk %s:%s nwr: ", sk->host, sk->port);
Assert(!sk->xlogreader);
sk->xlogreader = NeonWALReaderAllocate(wal_segment_size, sk->wp->propTermStartLsn, log_prefix);
sk->xlogreader = NeonWALReaderAllocate(wal_segment_size, sk->wp->propEpochStartLsn, log_prefix);
if (sk->xlogreader == NULL)
wpg_log(FATAL, "failed to allocate xlog reader");
}

34
poetry.lock generated
View File

@@ -1491,38 +1491,14 @@ files = [
[[package]]
name = "jsonnet"
version = "0.21.0rc2"
description = "Python bindings for Jsonnet - The data templating language "
version = "0.20.0"
description = "Python bindings for Jsonnet - The data templating language"
optional = false
python-versions = "*"
groups = ["main"]
markers = "python_version < \"3.13\""
files = [
{file = "jsonnet-0.21.0rc2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:8779ac6820fee44ef736df2baedc3ae93e8cd5d672ee105015c2a47fe627a727"},
{file = "jsonnet-0.21.0rc2-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:99affe8c71e2551465064a8039bb3d1cba27a0b73b2b9ff1b652e06f17d4ea8b"},
{file = "jsonnet-0.21.0rc2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4a9dffb9aa01013d100ddfb7230d1eeb80f2a8eef712b1825a60cad57106d8bd"},
{file = "jsonnet-0.21.0rc2-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:cca6c95f2879dcab52650b7aa09a4e82a139b084931b1f6f8c840f834fecc08a"},
{file = "jsonnet-0.21.0rc2-cp310-cp310-win_amd64.whl", hash = "sha256:016d6afdb302a6d00bf3bce6a0c3d9c093b992e33f9bc67c64a868035892258e"},
{file = "jsonnet-0.21.0rc2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:e893ab2c9bf10d8ec9e9b0cee8961879c88d0619cc6d8f75ea284a78e06ae32b"},
{file = "jsonnet-0.21.0rc2-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:c06b353cd3daa2781e6cd308e05f2f116396376994bcb5f59aaadbc6a752c7f2"},
{file = "jsonnet-0.21.0rc2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:9eb2bc8e62b73101329072da322f7e2a1bdb3ac530b94669128d1b480e311e55"},
{file = "jsonnet-0.21.0rc2-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:113766fd0c25620807bcf04d4c739f461c971a4f0e4aece9ba62b4e762de9598"},
{file = "jsonnet-0.21.0rc2-cp311-cp311-win_amd64.whl", hash = "sha256:8dab208c2c2760be60f87d1ceb8b28c86b51ed0e31129a7d90cd5fe890b41225"},
{file = "jsonnet-0.21.0rc2-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:95f5b9dd26a41d6f258d1baa8d22e557051beeed8c52a6202584f1becca9dcb5"},
{file = "jsonnet-0.21.0rc2-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:cecc6d76e2b377260fae0a060097c113e6ac361b8f739903ea7f3f5f64cdebdf"},
{file = "jsonnet-0.21.0rc2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aaa2d18224af7e63872ef4a101e93962505456cf5f5439c3cfc25dad6845f8b1"},
{file = "jsonnet-0.21.0rc2-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:2a9063f811554487ed552445e964aeec969cafb266b965029c8d6b091ce47950"},
{file = "jsonnet-0.21.0rc2-cp312-cp312-win_amd64.whl", hash = "sha256:80d171182c169761f744ba50068a4ad35d48e52b91d25bf4c7bb9a72f0a04f71"},
{file = "jsonnet-0.21.0rc2-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:f3657938f87cb6bc6da20ca631d437b5faf469ca060a7c7def9c8fd2f25a5e06"},
{file = "jsonnet-0.21.0rc2-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:3dcebc30cb991b58bc416ee05e9387004d04716d5c0b89714ff042bd069af5c8"},
{file = "jsonnet-0.21.0rc2-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3ac52c95482df3ed93c908468ca2f40d4825b6baba284b395ddc47bd663b8c3a"},
{file = "jsonnet-0.21.0rc2-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:8b34450823a7a1861de892fef9f29de1b4c19e1a79e27d81ffe7e57646cc89d6"},
{file = "jsonnet-0.21.0rc2-cp313-cp313-win_amd64.whl", hash = "sha256:573fd2580e46f4875ec505f1732f9e804b7063cba790342ed6fdafe9a6b30556"},
{file = "jsonnet-0.21.0rc2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:871ca1411de3626499bda60b330d37f85a592918f99ba4809089bbb8d4f5bfe4"},
{file = "jsonnet-0.21.0rc2-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:5d33b25a9c5bf9099100b9b16cb385a2876d891fbe639ee9d476fc75c861903a"},
{file = "jsonnet-0.21.0rc2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b2bac374565c7f89a4675f19fd2b624ed1376519267f4e444f49b6fc0368f6e5"},
{file = "jsonnet-0.21.0rc2-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:fab7bbd88f9159f88a7350701a97bda24de9e3b9eef14c2501ba8b9224160d60"},
{file = "jsonnet-0.21.0rc2-cp39-cp39-win_amd64.whl", hash = "sha256:ed71ffba0fd233a1bca7b0f7be79730792c5383e562a9dc7da152478d9ee1612"},
{file = "jsonnet-0.21.0rc2.tar.gz", hash = "sha256:2b83ec4b5a771c3732e0972be23a71f042ad2940db6918d3a52aade69bc394fb"},
{file = "jsonnet-0.20.0.tar.gz", hash = "sha256:7e770c7bf3a366b97b650a39430450f77612e74406731eb75c5bd59f3f104d4f"},
]
[[package]]
@@ -3844,4 +3820,4 @@ cffi = ["cffi (>=1.11)"]
[metadata]
lock-version = "2.1"
python-versions = "^3.11"
content-hash = "715fc8c896dcfa1b15054deeddcdec557ef93af91b26e1c8e4688fe4dbef5296"
content-hash = "010ffce959bb256880ab5a267048c182e4612b3151f9a94e3bf5d3a7807962fe"

View File

@@ -5,7 +5,6 @@ use std::sync::Arc;
use std::time::Duration;
use anyhow::{Context, bail, ensure};
use arc_swap::ArcSwapOption;
use camino::{Utf8Path, Utf8PathBuf};
use clap::Parser;
use compute_api::spec::LocalProxySpec;
@@ -28,7 +27,6 @@ use crate::config::{
};
use crate::control_plane::locks::ApiLocks;
use crate::control_plane::messages::{EndpointJwksResponse, JwksSettings};
use crate::ext::TaskExt;
use crate::http::health_server::AppMetrics;
use crate::intern::RoleNameInt;
use crate::metrics::{Metrics, ThreadPoolMetrics};
@@ -192,11 +190,7 @@ pub async fn run() -> anyhow::Result<()> {
// 2. The config file is written but the signal hook is not yet received
// 3. local_proxy completes startup but has no config loaded, despite there being a registerd config.
refresh_config_notify.notify_one();
tokio::spawn(refresh_config_loop(
config,
args.config_path,
refresh_config_notify,
));
tokio::spawn(refresh_config_loop(args.config_path, refresh_config_notify));
maintenance_tasks.spawn(crate::http::health_server::task_main(
metrics_listener,
@@ -275,7 +269,7 @@ fn build_config(args: &LocalProxyCliArgs) -> anyhow::Result<&'static ProxyConfig
};
Ok(Box::leak(Box::new(ProxyConfig {
tls_config: ArcSwapOption::from(None),
tls_config: None,
metric_collection: None,
http_config,
authentication_config: AuthenticationConfig {
@@ -317,16 +311,14 @@ enum RefreshConfigError {
Parse(#[from] serde_json::Error),
#[error(transparent)]
Validate(anyhow::Error),
#[error(transparent)]
Tls(anyhow::Error),
}
async fn refresh_config_loop(config: &ProxyConfig, path: Utf8PathBuf, rx: Arc<Notify>) {
async fn refresh_config_loop(path: Utf8PathBuf, rx: Arc<Notify>) {
let mut init = true;
loop {
rx.notified().await;
match refresh_config_inner(config, &path).await {
match refresh_config_inner(&path).await {
Ok(()) => {}
// don't log for file not found errors if this is the first time we are checking
// for computes that don't use local_proxy, this is not an error.
@@ -335,9 +327,6 @@ async fn refresh_config_loop(config: &ProxyConfig, path: Utf8PathBuf, rx: Arc<No
{
debug!(error=?e, ?path, "could not read config file");
}
Err(RefreshConfigError::Tls(e)) => {
error!(error=?e, ?path, "could not read TLS certificates");
}
Err(e) => {
error!(error=?e, ?path, "could not read config file");
}
@@ -347,10 +336,7 @@ async fn refresh_config_loop(config: &ProxyConfig, path: Utf8PathBuf, rx: Arc<No
}
}
async fn refresh_config_inner(
config: &ProxyConfig,
path: &Utf8Path,
) -> Result<(), RefreshConfigError> {
async fn refresh_config_inner(path: &Utf8Path) -> Result<(), RefreshConfigError> {
let bytes = tokio::fs::read(&path).await?;
let data: LocalProxySpec = serde_json::from_slice(&bytes)?;
@@ -420,20 +406,5 @@ async fn refresh_config_inner(
info!("successfully loaded new config");
JWKS_ROLE_MAP.store(Some(Arc::new(EndpointJwksResponse { jwks: jwks_set })));
if let Some(tls_config) = data.tls {
let tls_config = tokio::task::spawn_blocking(move || {
crate::tls::server_config::configure_tls(
&tls_config.key_path,
&tls_config.cert_path,
None,
false,
)
})
.await
.propagate_task_panic()
.map_err(RefreshConfigError::Tls)?;
config.tls_config.store(Some(Arc::new(tls_config)));
}
Ok(())
}

View File

@@ -4,7 +4,6 @@ use std::sync::Arc;
use std::time::Duration;
use anyhow::bail;
use arc_swap::ArcSwapOption;
use futures::future::Either;
use remote_storage::RemoteStorageConfig;
use tokio::net::TcpListener;
@@ -564,7 +563,6 @@ fn build_config(args: &ProxyCliArgs) -> anyhow::Result<&'static ProxyConfig> {
(None, None) => None,
_ => bail!("either both or neither tls-key and tls-cert must be specified"),
};
let tls_config = ArcSwapOption::from(tls_config.map(Arc::new));
let backup_metric_collection_config = config::MetricBackupCollectionConfig {
remote_storage_config: args.metric_backup_collection_remote_storage.clone(),

Some files were not shown because too many files have changed in this diff Show More