chore: cherry pick #5625 to v0.12 branch (#6831)

* ci: update 0.12 release ci

Signed-off-by: evenyag <realevenyag@gmail.com>

* fix: out of bound during bloom search (#5625)

Signed-off-by: Zhenchi <zhongzc_arch@outlook.com>

---------

Signed-off-by: evenyag <realevenyag@gmail.com>
Signed-off-by: Zhenchi <zhongzc_arch@outlook.com>
Co-authored-by: Zhenchi <zhongzc_arch@outlook.com>
This commit is contained in:
Yingwen
2025-08-27 16:12:25 +08:00
committed by GitHub
parent 8da5949fc5
commit f4c527cddf
9 changed files with 418 additions and 19 deletions

42
.github/scripts/check-version.sh vendored Executable file
View File

@@ -0,0 +1,42 @@
#!/bin/bash
# Get current version
CURRENT_VERSION=$1
if [ -z "$CURRENT_VERSION" ]; then
echo "Error: Failed to get current version"
exit 1
fi
# Get the latest version from GitHub Releases
API_RESPONSE=$(curl -s "https://api.github.com/repos/GreptimeTeam/greptimedb/releases/latest")
if [ -z "$API_RESPONSE" ] || [ "$(echo "$API_RESPONSE" | jq -r '.message')" = "Not Found" ]; then
echo "Error: Failed to fetch latest version from GitHub"
exit 1
fi
# Get the latest version
LATEST_VERSION=$(echo "$API_RESPONSE" | jq -r '.tag_name')
if [ -z "$LATEST_VERSION" ] || [ "$LATEST_VERSION" = "null" ]; then
echo "Error: No valid version found in GitHub releases"
exit 1
fi
# Cleaned up version number format (removed possible 'v' prefix and -nightly suffix)
CLEAN_CURRENT=$(echo "$CURRENT_VERSION" | sed 's/^v//' | sed 's/-nightly-.*//')
CLEAN_LATEST=$(echo "$LATEST_VERSION" | sed 's/^v//' | sed 's/-nightly-.*//')
echo "Current version: $CLEAN_CURRENT"
echo "Latest release version: $CLEAN_LATEST"
# Use sort -V to compare versions
HIGHER_VERSION=$(printf "%s\n%s" "$CLEAN_CURRENT" "$CLEAN_LATEST" | sort -V | tail -n1)
if [ "$HIGHER_VERSION" = "$CLEAN_CURRENT" ]; then
echo "Current version ($CLEAN_CURRENT) is NEWER than or EQUAL to latest ($CLEAN_LATEST)"
echo "should-push-latest-tag=true" >> $GITHUB_OUTPUT
else
echo "Current version ($CLEAN_CURRENT) is OLDER than latest ($CLEAN_LATEST)"
echo "should-push-latest-tag=false" >> $GITHUB_OUTPUT
fi

View File

@@ -706,7 +706,7 @@ jobs:
- name: Install toolchain
uses: actions-rust-lang/setup-rust-toolchain@v1
with:
cache: false
cache: false
- name: Rust Cache
uses: Swatinem/rust-cache@v2
with:

View File

@@ -110,6 +110,8 @@ jobs:
# The 'version' use as the global tag name of the release workflow.
version: ${{ steps.create-version.outputs.version }}
should-push-latest-tag: ${{ steps.check-version.outputs.should-push-latest-tag }}
steps:
- name: Checkout
uses: actions/checkout@v4
@@ -135,6 +137,11 @@ jobs:
GITHUB_REF_NAME: ${{ github.ref_name }}
NIGHTLY_RELEASE_PREFIX: ${{ env.NIGHTLY_RELEASE_PREFIX }}
- name: Check version
id: check-version
run: |
./.github/scripts/check-version.sh "${{ steps.create-version.outputs.version }}"
- name: Allocate linux-amd64 runner
if: ${{ inputs.build_linux_amd64_artifacts || github.event_name == 'push' || github.event_name == 'schedule' }}
uses: ./.github/actions/start-runner
@@ -314,7 +321,7 @@ jobs:
image-registry-username: ${{ secrets.DOCKERHUB_USERNAME }}
image-registry-password: ${{ secrets.DOCKERHUB_TOKEN }}
version: ${{ needs.allocate-runners.outputs.version }}
push-latest-tag: ${{ github.ref_type == 'tag' && !contains(github.ref_name, 'nightly') && github.event_name != 'schedule' }}
push-latest-tag: ${{ needs.allocate-runners.outputs.should-push-latest-tag == 'true' && github.ref_type == 'tag' && !contains(github.ref_name, 'nightly') && github.event_name != 'schedule' }}
- name: Set build image result
id: set-build-image-result
@@ -332,7 +339,7 @@ jobs:
build-windows-artifacts,
release-images-to-dockerhub,
]
runs-on: ubuntu-latest
runs-on: ubuntu-latest-16-cores
# When we push to ACR, it's easy to fail due to some unknown network issues.
# However, we don't want to fail the whole workflow because of this.
# The ACR have daily sync with DockerHub, so don't worry about the image not being updated.
@@ -361,7 +368,7 @@ jobs:
dev-mode: false
upload-to-s3: true
update-version-info: true
push-latest-tag: ${{ github.ref_type == 'tag' && !contains(github.ref_name, 'nightly') && github.event_name != 'schedule' }}
push-latest-tag: ${{ needs.allocate-runners.outputs.should-push-latest-tag == 'true' && github.ref_type == 'tag' && !contains(github.ref_name, 'nightly') && github.event_name != 'schedule' }}
publish-github-release:
name: Create GitHub release and upload artifacts

View File

@@ -11,17 +11,17 @@ concurrency:
group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
cancel-in-progress: true
permissions:
issues: write
contents: write
pull-requests: write
jobs:
check:
runs-on: ubuntu-latest
permissions:
pull-requests: write # Add permissions to modify PRs
issues: write
timeout-minutes: 10
steps:
- uses: actions/checkout@v4
with:
persist-credentials: false
- uses: ./.github/actions/setup-cyborg
- name: Check Pull Request
working-directory: cyborg

156
cyborg/bin/bump-versions.ts Normal file
View File

@@ -0,0 +1,156 @@
/*
* Copyright 2023 Greptime Team
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import * as core from "@actions/core";
import {obtainClient} from "@/common";
interface RepoConfig {
tokenEnv: string;
repo: string;
workflowLogic: (version: string) => [string, string] | null;
}
const REPO_CONFIGS: Record<string, RepoConfig> = {
website: {
tokenEnv: "WEBSITE_REPO_TOKEN",
repo: "website",
workflowLogic: (version: string) => {
// Skip nightly versions for website
if (version.includes('nightly')) {
console.log('Nightly version detected for website, skipping workflow trigger.');
return null;
}
return ['bump-patch-version.yml', version];
}
},
demo: {
tokenEnv: "DEMO_REPO_TOKEN",
repo: "demo-scene",
workflowLogic: (version: string) => {
// Skip nightly versions for demo
if (version.includes('nightly')) {
console.log('Nightly version detected for demo, skipping workflow trigger.');
return null;
}
return ['bump-patch-version.yml', version];
}
},
docs: {
tokenEnv: "DOCS_REPO_TOKEN",
repo: "docs",
workflowLogic: (version: string) => {
// Check if it's a nightly version
if (version.includes('nightly')) {
return ['bump-nightly-version.yml', version];
}
const parts = version.split('.');
if (parts.length !== 3) {
throw new Error('Invalid version format');
}
// If patch version (last number) is 0, it's a major version
// Return only major.minor version
if (parts[2] === '0') {
return ['bump-version.yml', `${parts[0]}.${parts[1]}`];
}
// Otherwise it's a patch version, use full version
return ['bump-patch-version.yml', version];
}
}
};
async function triggerWorkflow(repoConfig: RepoConfig, workflowId: string, version: string) {
const client = obtainClient(repoConfig.tokenEnv);
try {
await client.rest.actions.createWorkflowDispatch({
owner: "GreptimeTeam",
repo: repoConfig.repo,
workflow_id: workflowId,
ref: "main",
inputs: {
version,
},
});
console.log(`Successfully triggered ${workflowId} workflow for ${repoConfig.repo} with version ${version}`);
} catch (error) {
core.setFailed(`Failed to trigger workflow for ${repoConfig.repo}: ${error.message}`);
throw error;
}
}
async function processRepo(repoName: string, version: string) {
const repoConfig = REPO_CONFIGS[repoName];
if (!repoConfig) {
throw new Error(`Unknown repository: ${repoName}`);
}
try {
const workflowResult = repoConfig.workflowLogic(version);
if (workflowResult === null) {
// Skip this repo (e.g., nightly version for website)
return;
}
const [workflowId, apiVersion] = workflowResult;
await triggerWorkflow(repoConfig, workflowId, apiVersion);
} catch (error) {
core.setFailed(`Error processing ${repoName} with version ${version}: ${error.message}`);
throw error;
}
}
async function main() {
const version = process.env.VERSION;
if (!version) {
core.setFailed("VERSION environment variable is required");
process.exit(1);
}
// Remove 'v' prefix if exists
const cleanVersion = version.startsWith('v') ? version.slice(1) : version;
// Get target repositories from environment variable
// Default to both if not specified
const targetRepos = process.env.TARGET_REPOS?.split(',').map(repo => repo.trim()) || ['website', 'docs'];
console.log(`Processing version ${cleanVersion} for repositories: ${targetRepos.join(', ')}`);
const errors: string[] = [];
// Process each repository
for (const repo of targetRepos) {
try {
await processRepo(repo, cleanVersion);
} catch (error) {
errors.push(`${repo}: ${error.message}`);
}
}
if (errors.length > 0) {
core.setFailed(`Failed to process some repositories: ${errors.join('; ')}`);
process.exit(1);
}
console.log('All repositories processed successfully');
}
// Execute main function
main().catch((error) => {
core.setFailed(`Unexpected error: ${error.message}`);
process.exit(1);
});

View File

@@ -55,12 +55,25 @@ async function main() {
await client.rest.issues.addLabels({
owner, repo, issue_number: number, labels: [labelDocsRequired],
})
// Get available assignees for the docs repo
const assigneesResponse = await docsClient.rest.issues.listAssignees({
owner: 'GreptimeTeam',
repo: 'docs',
})
const validAssignees = assigneesResponse.data.map(assignee => assignee.login)
core.info(`Available assignees: ${validAssignees.join(', ')}`)
// Check if the actor is a valid assignee, otherwise fallback to fengjiachun
const assignee = validAssignees.includes(actor) ? actor : 'fengjiachun'
core.info(`Assigning issue to: ${assignee}`)
await docsClient.rest.issues.create({
owner: 'GreptimeTeam',
repo: 'docs',
title: `Update docs for ${title}`,
body: `A document change request is generated from ${html_url}`,
assignee: actor,
assignee: assignee,
}).then((res) => {
core.info(`Created issue ${res.data}`)
})

View File

@@ -53,6 +53,54 @@ get_arch_type() {
esac
}
# Verify SHA256 checksum
verify_sha256() {
file="$1"
expected_sha256="$2"
if command -v sha256sum >/dev/null 2>&1; then
actual_sha256=$(sha256sum "$file" | cut -d' ' -f1)
elif command -v shasum >/dev/null 2>&1; then
actual_sha256=$(shasum -a 256 "$file" | cut -d' ' -f1)
else
echo "Warning: No SHA256 verification tool found (sha256sum or shasum). Skipping checksum verification."
return 0
fi
if [ "$actual_sha256" = "$expected_sha256" ]; then
echo "SHA256 checksum verified successfully."
return 0
else
echo "Error: SHA256 checksum verification failed!"
echo "Expected: $expected_sha256"
echo "Actual: $actual_sha256"
return 1
fi
}
# Prompt for user confirmation (compatible with different shells)
prompt_confirmation() {
message="$1"
printf "%s (y/N): " "$message"
# Try to read user input, fallback if read fails
answer=""
if read answer </dev/tty 2>/dev/null; then
case "$answer" in
[Yy]|[Yy][Ee][Ss])
return 0
;;
*)
return 1
;;
esac
else
echo ""
echo "Cannot read user input. Defaulting to No."
return 1
fi
}
download_artifact() {
if [ -n "${OS_TYPE}" ] && [ -n "${ARCH_TYPE}" ]; then
# Use the latest stable released version.
@@ -71,17 +119,104 @@ download_artifact() {
fi
echo "Downloading ${BIN}, OS: ${OS_TYPE}, Arch: ${ARCH_TYPE}, Version: ${VERSION}"
PACKAGE_NAME="${BIN}-${OS_TYPE}-${ARCH_TYPE}-${VERSION}.tar.gz"
PKG_NAME="${BIN}-${OS_TYPE}-${ARCH_TYPE}-${VERSION}"
PACKAGE_NAME="${PKG_NAME}.tar.gz"
SHA256_FILE="${PKG_NAME}.sha256sum"
if [ -n "${PACKAGE_NAME}" ]; then
wget "https://github.com/${GITHUB_ORG}/${GITHUB_REPO}/releases/download/${VERSION}/${PACKAGE_NAME}"
# Check if files already exist and prompt for override
if [ -f "${PACKAGE_NAME}" ]; then
echo "File ${PACKAGE_NAME} already exists."
if prompt_confirmation "Do you want to override it?"; then
echo "Overriding existing file..."
rm -f "${PACKAGE_NAME}"
else
echo "Skipping download. Using existing file."
fi
fi
if [ -f "${BIN}" ]; then
echo "Binary ${BIN} already exists."
if prompt_confirmation "Do you want to override it?"; then
echo "Will override existing binary..."
rm -f "${BIN}"
else
echo "Installation cancelled."
exit 0
fi
fi
# Download package if not exists
if [ ! -f "${PACKAGE_NAME}" ]; then
echo "Downloading ${PACKAGE_NAME}..."
# Use curl instead of wget for better compatibility
if command -v curl >/dev/null 2>&1; then
if ! curl -L -o "${PACKAGE_NAME}" "https://github.com/${GITHUB_ORG}/${GITHUB_REPO}/releases/download/${VERSION}/${PACKAGE_NAME}"; then
echo "Error: Failed to download ${PACKAGE_NAME}"
exit 1
fi
elif command -v wget >/dev/null 2>&1; then
if ! wget -O "${PACKAGE_NAME}" "https://github.com/${GITHUB_ORG}/${GITHUB_REPO}/releases/download/${VERSION}/${PACKAGE_NAME}"; then
echo "Error: Failed to download ${PACKAGE_NAME}"
exit 1
fi
else
echo "Error: Neither curl nor wget is available for downloading."
exit 1
fi
fi
# Download and verify SHA256 checksum
echo "Downloading SHA256 checksum..."
sha256_download_success=0
if command -v curl >/dev/null 2>&1; then
if curl -L -s -o "${SHA256_FILE}" "https://github.com/${GITHUB_ORG}/${GITHUB_REPO}/releases/download/${VERSION}/${SHA256_FILE}" 2>/dev/null; then
sha256_download_success=1
fi
elif command -v wget >/dev/null 2>&1; then
if wget -q -O "${SHA256_FILE}" "https://github.com/${GITHUB_ORG}/${GITHUB_REPO}/releases/download/${VERSION}/${SHA256_FILE}" 2>/dev/null; then
sha256_download_success=1
fi
fi
if [ $sha256_download_success -eq 1 ] && [ -f "${SHA256_FILE}" ]; then
expected_sha256=$(cat "${SHA256_FILE}" | cut -d' ' -f1)
if [ -n "$expected_sha256" ]; then
if ! verify_sha256 "${PACKAGE_NAME}" "${expected_sha256}"; then
echo "SHA256 verification failed. Removing downloaded file."
rm -f "${PACKAGE_NAME}" "${SHA256_FILE}"
exit 1
fi
else
echo "Warning: Could not parse SHA256 checksum from file."
fi
rm -f "${SHA256_FILE}"
else
echo "Warning: Could not download SHA256 checksum file. Skipping verification."
fi
# Extract the binary and clean the rest.
tar xvf "${PACKAGE_NAME}" && \
mv "${PACKAGE_NAME%.tar.gz}/${BIN}" "${PWD}" && \
rm -r "${PACKAGE_NAME}" && \
rm -r "${PACKAGE_NAME%.tar.gz}" && \
echo "Run './${BIN} --help' to get started"
echo "Extracting ${PACKAGE_NAME}..."
if ! tar xf "${PACKAGE_NAME}"; then
echo "Error: Failed to extract ${PACKAGE_NAME}"
exit 1
fi
# Find the binary in the extracted directory
extracted_dir="${PACKAGE_NAME%.tar.gz}"
if [ -f "${extracted_dir}/${BIN}" ]; then
mv "${extracted_dir}/${BIN}" "${PWD}/"
rm -f "${PACKAGE_NAME}"
rm -rf "${extracted_dir}"
chmod +x "${BIN}"
echo "Installation completed successfully!"
echo "Run './${BIN} --help' to get started"
else
echo "Error: Binary ${BIN} not found in extracted archive"
rm -f "${PACKAGE_NAME}"
rm -rf "${extracted_dir}"
exit 1
fi
fi
fi
}

View File

@@ -42,7 +42,16 @@ impl BloomFilterApplier {
) -> Result<Vec<Range<usize>>> {
let rows_per_segment = self.meta.rows_per_segment as usize;
let start_seg = search_range.start / rows_per_segment;
let end_seg = search_range.end.div_ceil(rows_per_segment);
let mut end_seg = search_range.end.div_ceil(rows_per_segment);
if end_seg == self.meta.segment_loc_indices.len() + 1 {
// In a previous version, there was a bug where if the last segment was all null,
// this segment would not be written into the index. This caused the slice
// `self.meta.segment_loc_indices[start_seg..end_seg]` to go out of bounds due to
// the missing segment. Since the `search` function does not search for nulls,
// we can simply ignore the last segment in this buggy scenario.
end_seg -= 1;
}
let locs = &self.meta.segment_loc_indices[start_seg..end_seg];

View File

@@ -64,6 +64,9 @@ pub struct BloomFilterCreator {
/// Storage for finalized Bloom filters.
finalized_bloom_filters: FinalizedBloomFilterStorage,
/// Row count that finalized so far.
finalized_row_count: usize,
/// Global memory usage of the bloom filter creator.
global_memory_usage: Arc<AtomicUsize>,
}
@@ -96,6 +99,7 @@ impl BloomFilterCreator {
global_memory_usage,
global_memory_usage_threshold,
),
finalized_row_count: 0,
}
}
@@ -136,6 +140,7 @@ impl BloomFilterCreator {
if self.accumulated_row_count % self.rows_per_segment == 0 {
self.finalize_segment().await?;
self.finalized_row_count = self.accumulated_row_count;
}
}
@@ -161,6 +166,7 @@ impl BloomFilterCreator {
if self.accumulated_row_count % self.rows_per_segment == 0 {
self.finalize_segment().await?;
self.finalized_row_count = self.accumulated_row_count;
}
Ok(())
@@ -168,7 +174,7 @@ impl BloomFilterCreator {
/// Finalizes any remaining segments and writes the bloom filters and metadata to the provided writer.
pub async fn finish(&mut self, mut writer: impl AsyncWrite + Unpin) -> Result<()> {
if !self.cur_seg_distinct_elems.is_empty() {
if self.accumulated_row_count > self.finalized_row_count {
self.finalize_segment().await?;
}
@@ -406,4 +412,35 @@ mod tests {
assert!(bf.contains(&b"f"));
}
}
#[tokio::test]
async fn test_final_seg_all_null() {
let mut writer = Cursor::new(Vec::new());
let mut creator = BloomFilterCreator::new(
2,
Arc::new(MockExternalTempFileProvider::new()),
Arc::new(AtomicUsize::new(0)),
None,
);
creator
.push_n_row_elems(4, vec![b"a".to_vec(), b"b".to_vec()])
.await
.unwrap();
creator.push_row_elems(Vec::new()).await.unwrap();
creator.finish(&mut writer).await.unwrap();
let bytes = writer.into_inner();
let total_size = bytes.len();
let meta_size_offset = total_size - 4;
let meta_size = u32::from_le_bytes((&bytes[meta_size_offset..]).try_into().unwrap());
let meta_bytes = &bytes[total_size - meta_size as usize - 4..total_size - 4];
let meta = BloomFilterMeta::decode(meta_bytes).unwrap();
assert_eq!(meta.rows_per_segment, 2);
assert_eq!(meta.segment_count, 3);
assert_eq!(meta.row_count, 5);
}
}