test: sqlness test for insert default (#873)

* test: sqlness test for insert default

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

* empty line

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

* add more sqls

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

* fix typos

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

* update test according to typo fix

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

Signed-off-by: Ruihang Xia <waynestxia@gmail.com>

.env.example

@@ -0,0 +1,4 @@
+# Settings for s3 test
+GT_S3_BUCKET=S3 bucket
+GT_S3_ACCESS_KEY_ID=S3 access key id
+GT_S3_ACCESS_KEY=S3 secret access key

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,86 @@
+---
+name: Bug report
+description: Is something not working? Help us fix it!
+labels: [ "bug" ]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Take some time to fill out this bug report. Thank you!
+
+  - type: dropdown
+    id: type
+    attributes:
+      label: What type of bug is this?
+      multiple: true
+      options:
+        - Configuration
+        - Crash
+        - Data corruption
+        - Incorrect result
+        - Locking issue
+        - Performance issue
+        - Unexpected error
+        - Other
+    validations:
+      required: true
+
+  - type: dropdown
+    id: subsystem
+    attributes:
+      label: What subsystems are affected?
+      description: You can pick multiple subsystems.
+      multiple: true
+      options:
+        - Standalone mode
+        - Frontend
+        - Datanode
+        - Meta
+        - Other
+    validations:
+      required: true
+
+  - type: textarea
+    id: what-happened
+    attributes:
+      label: What happened?
+      description: |
+        Tell us what happened and also what you would have expected to
+        happen instead.
+      placeholder: "Describe the bug"
+    validations:
+      required: true
+
+  - type: input
+    id: os
+    attributes:
+      label: What operating system did you use?
+      description: |
+        Please provide OS, version, and architecture. For example:
+        Windows 10 x64, Ubuntu 21.04 x64, Mac OS X 10.5 ARM, Rasperry
+        Pi i386, etc.
+      placeholder: "Ubuntu 21.04 x64"
+    validations:
+      required: true
+
+  - type: textarea
+    id: logs
+    attributes:
+      label: Relevant log output and stack trace
+      description: |
+        Please copy and paste any relevant log output or a stack
+        trace. This will be automatically formatted into code, so no
+        need for backticks.
+      render: bash
+
+  - type: textarea
+    id: reproduce
+    attributes:
+      label: How can we reproduce the bug?
+      description: |
+        Please walk us through and provide steps and details on how
+        to reproduce the issue. If possible, provide scripts that we
+        can run to trigger the bug.
+      render: bash
+    validations:
+      required: true

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,8 @@
+blank_issues_enabled: false
+contact_links:
+  - name: Greptime Community Slack
+    url: https://greptime.com/slack
+    about: Get free help from the Greptime community
+  - name: Greptime Community Discussion
+    url: https://github.com/greptimeTeam/greptimedb/discussions
+    about: Get free help from the Greptime community

.github/ISSUE_TEMPLATE/enhancement.yml

@@ -0,0 +1,39 @@
+---
+name: Enhancement
+description: Suggest an enhancement to existing functionality
+labels: [ "enhancement" ]
+body:
+  - type: dropdown
+    id: type
+    attributes:
+      label: What type of enhancement is this?
+      multiple: true
+      options:
+        - API improvement
+        - Configuration
+        - Performance
+        - Refactor
+        - Tech debt reduction
+        - User experience
+        - Other
+    validations:
+      required: true
+
+  - type: textarea
+    id: what
+    attributes:
+      label: What does the enhancement do?
+      description: |
+        Give a high-level overview of how you
+        suggest improving an existing feature or functionality.
+    validations:
+     required: true
+
+  - type: textarea
+    id: implementation
+    attributes:
+      label: Implementation challenges
+      description: |
+        Share any ideas of how to implement the enhancement.
+    validations:
+      required: false

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,42 @@
+---
+name: Feature request
+description: Suggest a new feature for GreptimeDB
+labels: [ "feature request" ]
+body:
+  - type: markdown
+    id: info
+    attributes:
+      value: |
+        Only use this template to suggest a new feature that doesn't already exist in GreptimeDB.
+        For enhancements to existing features, use the "Enhancement" issue template. For bugs,
+        use the bug report template.
+
+  - type: textarea
+    id: what
+    attributes:
+      label: What problem does the new feature solve?
+      description: |
+        Describe the problem and why it is important to solve. Did you consider alternative
+        solutions, perhaps outside the database? Why is it better to add the feature to
+        GreptimeDB?
+    validations:
+      required: true
+
+  - type: textarea
+    id: how
+    attributes:
+      label: What does the feature do?
+      description: |
+        Give a high-level overview of what the feature does and how it would work.
+    validations:
+     required: true
+
+  - type: textarea
+    id: implementation
+    attributes:
+      label: Implementation challenges
+      description: |
+        If you have ideas of how to implement the feature, and any particularly
+        challenging issues to overcome, then provide them here.
+    validations:
+      required: false

.github/pull_request_template.md

@@ -0,0 +1,19 @@
+I hereby agree to the terms of the [GreptimeDB CLA](https://gist.github.com/xtang/6378857777706e568c1949c7578592cc)
+
+## What's changed and what's your intention?
+
+_PLEASE DO NOT LEAVE THIS EMPTY !!!_
+
+Please explain IN DETAIL what the changes are in this PR and why they are needed:
+
+- Summarize your change (**mandatory**)
+- How does this PR work? Need a brief introduction for the changed logic (optional)
+- Describe clearly one logical change and avoid lazy messages (optional)
+- Describe any limitations of the current code (optional)
+
+## Checklist
+
+- [ ]  I have written the necessary rustdoc comments.
+- [ ]  I have added the necessary unit tests and integration tests.
+
+## Refer to a related PR or issue link (optional)

.github/workflows/coverage.yml

@@ -24,7 +24,7 @@ on:
 name: Code coverage
 
 env:
-  RUST_TOOLCHAIN: nightly-2022-07-14
+  RUST_TOOLCHAIN: nightly-2022-12-20
 
 jobs:
   coverage:
@@ -34,6 +34,11 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
+      - uses: KyleMayes/install-llvm-action@v1
+        with:
+          version: "14.0"
       - name: Install toolchain
         uses: dtolnay/rust-toolchain@master
         with:
@@ -41,10 +46,6 @@ jobs:
           components: llvm-tools-preview
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
-      - name: Cleanup disk
-        uses: curoky/cleanup-disk-action@v2.0
-        with:
-          retain: 'rust'
       - name: Install latest nextest release
         uses: taiki-e/install-action@nextest
       - name: Install cargo-llvm-cov
@@ -52,6 +53,7 @@ jobs:
       - name: Collect coverage data
         run: cargo llvm-cov nextest --workspace --lcov --output-path lcov.info
         env:
+          CARGO_BUILD_RUSTFLAGS: "-C link-arg=-fuse-ld=lld"
           RUST_BACKTRACE: 1
           CARGO_INCREMENTAL: 0
           GT_S3_BUCKET: ${{ secrets.S3_BUCKET }}

.github/workflows/develop.yml

@@ -23,9 +23,16 @@ on:
 name: CI
 
 env:
-  RUST_TOOLCHAIN: nightly-2022-07-14
+  RUST_TOOLCHAIN: nightly-2022-12-20
 
 jobs:
+  typos:
+    name: Spell Check with Typos
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - uses: crate-ci/typos@v1.0.4
+
   check:
     name: Check
     if: github.event.pull_request.draft == false
@@ -34,6 +41,8 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
       - uses: dtolnay/rust-toolchain@master
         with:
           toolchain: ${{ env.RUST_TOOLCHAIN }}
@@ -42,6 +51,23 @@ jobs:
       - name: Run cargo check
         run: cargo check --workspace --all-targets
 
+  toml:
+    name: Toml Check
+    if: github.event.pull_request.draft == false
+    runs-on: ubuntu-latest
+    timeout-minutes: 60
+    steps:
+      - uses: actions/checkout@v3
+      - uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - name: Rust Cache
+        uses: Swatinem/rust-cache@v2
+      - name: Install taplo
+        run: cargo install taplo-cli --version ^0.8 --locked
+      - name: Run taplo
+        run: taplo format --check --option "indent_string=    "
+
   # Use coverage to run test.
   # test:
   #   name: Test Suite
@@ -57,6 +83,8 @@ jobs:
   #         path: ./llvm
   #         key: llvm
   #     - uses: arduino/setup-protoc@v1
+  #       with:
+  #         repo-token: ${{ secrets.GITHUB_TOKEN }}
   #     - uses: KyleMayes/install-llvm-action@v1
   #       with:
   #         version: "14.0"
@@ -82,6 +110,24 @@ jobs:
   #         GT_S3_ACCESS_KEY: ${{ secrets.S3_ACCESS_KEY }}
   #         UNITTEST_LOG_DIR: "__unittest_logs"
 
+  sqlness:
+    name: Sqlness Test
+    if: github.event.pull_request.draft == false
+    runs-on: ubuntu-latest-8-cores
+    timeout-minutes: 60
+    steps:
+      - uses: actions/checkout@v3
+      - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
+      - uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - name: Rust Cache
+        uses: Swatinem/rust-cache@v2
+      - name: Run sqlness
+        run: cargo run --bin sqlness-runner
+
   fmt:
     name: Rustfmt
     if: github.event.pull_request.draft == false
@@ -90,6 +136,8 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
       - uses: dtolnay/rust-toolchain@master
         with:
           toolchain: ${{ env.RUST_TOOLCHAIN }}
@@ -107,6 +155,8 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
       - uses: dtolnay/rust-toolchain@master
         with:
           toolchain: ${{ env.RUST_TOOLCHAIN }}

.github/workflows/doc-issue.yml

@@ -0,0 +1,25 @@
+name: Create Issue in docs repo on doc related changes
+
+on:
+  issues:
+    types:
+      - labeled
+  pull_request_target:
+    types:
+      - labeled
+
+jobs:
+  doc_issue:
+    if: github.event.label.name == 'doc update required'
+    runs-on: ubuntu-latest
+    steps:
+      - name: create an issue in doc repo
+        uses: dacbd/create-issue-action@main
+        with:
+          owner: GreptimeTeam
+          repo: docs
+          token: ${{ secrets.DOCS_REPO_TOKEN }}
+          title: Update docs for ${{ github.event.issue.title || github.event.pull_request.title }}
+          body: |
+            A document change request is generated from
+            ${{ github.event.issue.html_url || github.event.pull_request.html_url }}

.github/workflows/release.yml

@@ -2,12 +2,23 @@ on:
   push:
     tags:
       - "v*.*.*"
+  schedule:
+    # At 00:00 on Monday.
+    - cron: '0 0 * * 1'
   workflow_dispatch:
 
 name: Release
 
 env:
-  RUST_TOOLCHAIN: nightly-2022-07-14
+  RUST_TOOLCHAIN: nightly-2022-12-20
+
+  # FIXME(zyy17): Would be better to use `gh release list -L 1 | cut -f 3` to get the latest release version tag, but for a long time, we will stay at 'v0.1.0-alpha-*'.
+  SCHEDULED_BUILD_VERSION_PREFIX: v0.1.0-alpha
+
+  # In the future, we can change SCHEDULED_PERIOD to nightly.
+  SCHEDULED_PERIOD: weekly
+
+  CARGO_PROFILE: weekly
 
 jobs:
   build:
@@ -73,13 +84,16 @@ jobs:
       - name: Output package versions
         run: protoc --version ; cargo version ; rustc --version ; gcc --version ; g++ --version
 
+      - name: Run tests
+        run: make unit-test integration-test sqlness-test
+
       - name: Run cargo build
-        run: cargo build ${{ matrix.opts }} --release --locked --target ${{ matrix.arch }}
+        run: cargo build ${{ matrix.opts }} --profile ${{ env.CARGO_PROFILE }} --locked --target ${{ matrix.arch }}
 
       - name: Calculate checksum and rename binary
         shell: bash
         run: |
-          cd target/${{ matrix.arch }}/release
+          cd target/${{ matrix.arch }}/${{ env.CARGO_PROFILE }}
           chmod +x greptime
           tar -zcvf ${{ matrix.file }}.tgz greptime
           echo $(shasum -a 256 ${{ matrix.file }}.tgz | cut -f1 -d' ') > ${{ matrix.file }}.sha256sum
@@ -88,13 +102,13 @@ jobs:
         uses: actions/upload-artifact@v3
         with:
           name: ${{ matrix.file }}
-          path: target/${{ matrix.arch }}/release/${{ matrix.file }}.tgz
+          path: target/${{ matrix.arch }}/${{ env.CARGO_PROFILE }}/${{ matrix.file }}.tgz
 
       - name: Upload checksum of artifacts
         uses: actions/upload-artifact@v3
         with:
           name: ${{ matrix.file }}.sha256sum
-          path: target/${{ matrix.arch }}/release/${{ matrix.file }}.sha256sum
+          path: target/${{ matrix.arch }}/${{ env.CARGO_PROFILE }}/${{ matrix.file }}.sha256sum
   release:
     name: Release artifacts
     needs: [build]
@@ -106,8 +120,32 @@ jobs:
       - name: Download artifacts
         uses: actions/download-artifact@v3
 
+      - name: Configure scheduled build version # the version would be ${SCHEDULED_BUILD_VERSION_PREFIX}-YYYYMMDD-${SCHEDULED_PERIOD}, like v0.1.0-alpha-20221119-weekly.
+        shell: bash
+        if: github.event_name == 'schedule'
+        run: |
+          buildTime=`date "+%Y%m%d"`
+          SCHEDULED_BUILD_VERSION=${{ env.SCHEDULED_BUILD_VERSION_PREFIX }}-$buildTime-${{ env.SCHEDULED_PERIOD }}
+          echo "SCHEDULED_BUILD_VERSION=${SCHEDULED_BUILD_VERSION}" >> $GITHUB_ENV
+
+      - name: Create scheduled build git tag
+        if: github.event_name == 'schedule'
+        run: |
+          git tag ${{ env.SCHEDULED_BUILD_VERSION }}
+
+      - name: Publish scheduled release # configure the different release title and tags.
+        uses: softprops/action-gh-release@v1
+        if: github.event_name == 'schedule'
+        with:
+          name: "Release ${{ env.SCHEDULED_BUILD_VERSION }}"
+          tag_name: ${{ env.SCHEDULED_BUILD_VERSION }}
+          generate_release_notes: true
+          files: |
+            **/greptime-*
+
       - name: Publish release
         uses: softprops/action-gh-release@v1
+        if: github.event_name != 'schedule'
         with:
           name: "Release ${{ github.ref_name }}"
           files: |
@@ -145,12 +183,12 @@ jobs:
           tar xvf greptime-linux-arm64.tgz
           rm greptime-linux-arm64.tgz
 
-      - name: Login to GitHub Container Registry
+      - name: Login to UCloud Container Registry
         uses: docker/login-action@v2
         with:
-          registry: ghcr.io
-          username: ${{ github.actor }}
-          password: ${{ secrets.GITHUB_TOKEN }}
+          registry: uhub.service.ucloud.cn
+          username: ${{ secrets.UCLOUD_USERNAME }}
+          password: ${{ secrets.UCLOUD_PASSWORD }}
 
       - name: Login to Dockerhub
         uses: docker/login-action@v2
@@ -158,11 +196,20 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
 
+      - name: Configure scheduled build image tag # the tag would be ${SCHEDULED_BUILD_VERSION_PREFIX}-YYYYMMDD-${SCHEDULED_PERIOD}
+        shell: bash
+        if: github.event_name == 'schedule'
+        run: |
+          buildTime=`date "+%Y%m%d"`
+          SCHEDULED_BUILD_VERSION=${{ env.SCHEDULED_BUILD_VERSION_PREFIX }}-$buildTime-${{ env.SCHEDULED_PERIOD }}
+          echo "IMAGE_TAG=${SCHEDULED_BUILD_VERSION:1}" >> $GITHUB_ENV
+
       - name: Configure tag # If the release tag is v0.1.0, then the image version tag will be 0.1.0.
         shell: bash
+        if: github.event_name != 'schedule'
         run: |
           VERSION=${{ github.ref_name }}
-          echo "VERSION=${VERSION:1}" >> $GITHUB_ENV
+          echo "IMAGE_TAG=${VERSION:1}" >> $GITHUB_ENV
 
       - name: Set up QEMU
         uses: docker/setup-qemu-action@v2
@@ -179,5 +226,6 @@ jobs:
           platforms: linux/amd64,linux/arm64
           tags: |
             greptime/greptimedb:latest
-            greptime/greptimedb:${{ env.VERSION }}
-            ghcr.io/greptimeteam/greptimedb:${{ env.VERSION }}
+            greptime/greptimedb:${{ env.IMAGE_TAG }}
+            uhub.service.ucloud.cn/greptime/greptimedb:latest
+            uhub.service.ucloud.cn/greptime/greptimedb:${{ env.IMAGE_TAG }}

.gitignore

@@ -18,6 +18,7 @@ debug/
 
 # JetBrains IDE config directory
 .idea/
+*.iml
 
 # VSCode IDE config directory
 .vscode/
@@ -31,3 +32,6 @@ logs/
 
 # Benchmark dataset
 benchmarks/data
+
+# dotenv
+.env

.pre-commit-config.yaml

@@ -5,11 +5,11 @@ repos:
     -    id: conventional-pre-commit
          stages: [commit-msg]
 
--   repo: https://github.com/DevinR528/cargo-sort
-    rev: e6a795bc6b2c0958f9ef52af4863bbd7cc17238f
-    hooks:
-    -    id: cargo-sort
-         args: ["--workspace", "--print"]
+# -   repo: https://github.com/DevinR528/cargo-sort
+#     rev: e6a795bc6b2c0958f9ef52af4863bbd7cc17238f
+#     hooks:
+#     -    id: cargo-sort
+#          args: ["--workspace"]
 
 -   repo: https://github.com/doublify/pre-commit-rust
     rev: v1.0

Cargo.toml

@@ -11,11 +11,11 @@ members = [
     "src/common/function",
     "src/common/function-macro",
     "src/common/grpc",
+    "src/common/grpc-expr",
     "src/common/query",
     "src/common/recordbatch",
     "src/common/runtime",
     "src/common/substrait",
-    "src/common/insert",
     "src/common/telemetry",
     "src/common/time",
     "src/datanode",
@@ -24,16 +24,56 @@ members = [
     "src/log-store",
     "src/meta-client",
     "src/meta-srv",
+    "src/mito",
     "src/object-store",
+    "src/promql",
     "src/query",
     "src/script",
     "src/servers",
+    "src/session",
     "src/sql",
     "src/storage",
     "src/store-api",
     "src/table",
-    "src/table-engine",
+    "tests-integration",
+    "tests/runner",
 ]
 
+[workspace.package]
+version = "0.1.0"
+edition = "2021"
+license = "Apache-2.0"
+
+[workspace.dependencies]
+arrow = "29.0"
+arrow-flight = "29.0"
+arrow-schema = { version = "29.0", features = ["serde"] }
+async-stream = "0.3"
+async-trait = "0.1"
+# TODO(LFC): Use released Datafusion when it officially dpendent on Arrow 29.0
+datafusion = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+datafusion-common = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+datafusion-expr = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+datafusion-optimizer = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+datafusion-physical-expr = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+datafusion-sql = { git = "https://github.com/apache/arrow-datafusion.git", rev = "4917235a398ae20145c87d20984e6367dc1a0c1e" }
+futures = "0.3"
+futures-util = "0.3"
+parquet = "29.0"
+paste = "1.0"
+prost = "0.11"
+serde = { version = "1.0", features = ["derive"] }
+snafu = { version = "0.7", features = ["backtraces"] }
+sqlparser = "0.28"
+tokio = { version = "1", features = ["full"] }
+tonic = "0.8"
+
 [profile.release]
 debug = true
+
+[profile.weekly]
+inherits = "release"
+strip = true
+lto = true
+debug = false
+incremental = false

Makefile

@@ -0,0 +1,67 @@
+IMAGE_REGISTRY ?= greptimedb
+IMAGE_TAG ?= latest
+
+##@ Build
+
+.PHONY: build
+build: ## Build debug version greptime.
+	cargo build
+
+.PHONY: release
+release:  ## Build release version greptime.
+	cargo build --release
+
+.PHONY: clean
+clean: ## Clean the project.
+	cargo clean
+
+.PHONY: fmt
+fmt: ## Format all the Rust code.
+	cargo fmt --all
+
+.PHONY: docker-image
+docker-image: ## Build docker image.
+	docker build --network host -f docker/Dockerfile -t ${IMAGE_REGISTRY}:${IMAGE_TAG} .
+
+##@ Test
+
+.PHONY: unit-test
+unit-test: ## Run unit test.
+	cargo test --workspace
+
+.PHONY: integration-test
+integration-test: ## Run integation test.
+	cargo test integration
+
+.PHONY: sqlness-test
+sqlness-test: ## Run sqlness test.
+	cargo run --bin sqlness-runner
+
+.PHONY: check
+check: ## Cargo check all the targets.
+	cargo check --workspace --all-targets
+
+.PHONY: clippy
+clippy: ## Check clippy rules.
+	cargo clippy --workspace --all-targets -- -D warnings -D clippy::print_stdout -D clippy::print_stderr
+
+.PHONY: fmt-check
+fmt-check: ## Check code format.
+	cargo fmt --all -- --check
+
+##@ General
+
+# The help target prints out all targets with their descriptions organized
+# beneath their categories. The categories are represented by '##@' and the
+# target descriptions by '##'. The awk commands is responsible for reading the
+# entire set of makefiles included in this invocation, looking for lines of the
+# file as xyz: ## something, and then pretty-format the target and help. Then,
+# if there's a line with ##@ something, that gets pretty-printed as a category.
+# More info on the usage of ANSI control characters for terminal formatting:
+# https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_parameters
+# More info on the awk command:
+# https://linuxcommand.org/lc3_adv_awk.php
+
+.PHONY: help
+help: ## Display help messages.
+	@awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n  make \033[36m<target>\033[0m\n"} /^[a-zA-Z_0-9-]+:.*?##/ { printf "  \033[36m%-20s\033[0m %s\n", $$1, $$2 } /^##@/ { printf "\n\033[1m%s\033[0m\n", substr($$0, 5) } ' $(MAKEFILE_LIST)

README.md

@@ -1,7 +1,12 @@
 <p align="center">
-    <img src="/docs/logo-text-padding.png" alt="GreptimeDB Logo" width="400px"></img>
+  <picture>
+    <source media="(prefers-color-scheme: light)" srcset="/docs/logo-text-padding.png">
+    <source media="(prefers-color-scheme: dark)" srcset="/docs/logo-text-padding-dark.png">
+    <img alt="GreptimeDB Logo" src="/docs/logo-text-padding.png" width="400px">
+  </picture>
 </p>
 
+
 <h3 align="center">
     The next-generation hybrid timeseries/analytics processing database in the cloud
 </h3>
@@ -53,8 +58,9 @@ To compile GreptimeDB from source, you'll need:
   install correct Rust version for you.
 - Protobuf: `protoc` is required for compiling `.proto` files. `protobuf` is
   available from major package manager on macos and linux distributions. You can
-  find an installation instructions
-  [here](https://grpc.io/docs/protoc-installation/).
+  find an installation instructions [here](https://grpc.io/docs/protoc-installation/).
+  **Note that `protoc` version needs to be >= 3.15** because we have used the `optional`
+  keyword. You can check it with `protoc --version`.
 
 #### Build with Docker
 
@@ -114,7 +120,10 @@ about Kubernetes deployment, check our [docs](https://docs.greptime.com/).
 4. Query the data:
 
    ```SQL
-   mysql> SELECT * FROM monitor;
+   SELECT * FROM monitor;
+   ```
+
+   ```TEXT
    +-------+---------------------+------+--------+
    | host  | ts                  | cpu  | memory |
    +-------+---------------------+------+--------+
@@ -129,12 +138,23 @@ You can always cleanup test database by removing `/tmp/greptimedb`.
 
 ## Resources
 
+### Installation
+
 - [Pre-built Binaries](https://github.com/GreptimeTeam/greptimedb/releases):
   downloadable pre-built binaries for Linux and MacOS
 - [Docker Images](https://hub.docker.com/r/greptime/greptimedb): pre-built
   Docker images
 - [`gtctl`](https://github.com/GreptimeTeam/gtctl): the command-line tool for
   Kubernetes deployment
+
+### Documentation
+
+- GreptimeDB [User Guide](https://docs.greptime.com/user-guide/concepts.html)
+- GreptimeDB [Developer
+  Guide](https://docs.greptime.com/developer-guide/overview.html)
+
+### SDK
+
 - [GreptimeDB Java
   Client](https://github.com/GreptimeTeam/greptimedb-client-java)
 
@@ -145,6 +165,8 @@ break things. Benchmark on development branch may not represent its potential
 performance. We release pre-built binaries constantly for functional
 evaluation. Do not use it in production at the moment.
 
+For future plans, check out [GreptimeDB roadmap](https://github.com/GreptimeTeam/greptimedb/issues/669).
+
 ## Community
 
 Our core team is thrilled too see you participate in any ways you like. When you are stuck, try to
@@ -158,16 +180,10 @@ community, please check out:
 
 In addition, you may:
 
-- View our official [Blog](https://greptime.com/blog)
+- View our official [Blog](https://greptime.com/blogs/index)
 - Connect us with [Linkedin](https://www.linkedin.com/company/greptime/)
 - Follow us on [Twitter](https://twitter.com/greptime)
 
-## Documentation
-
-- GreptimeDB [User Guide](https://docs.greptime.com/user-guide/concepts.html)
-- GreptimeDB [Developer
-  Guide](https://docs.greptime.com/developer-guide/overview.html)
-
 ## License
 
 GreptimeDB uses the [Apache 2.0 license][1] to strike a balance between
@@ -184,3 +200,4 @@ Please refer to [contribution guidelines](CONTRIBUTING.md) for more information.
 - GreptimeDB's query engine is powered by [Apache Arrow DataFusion](https://github.com/apache/arrow-datafusion).
 - [OpenDAL](https://github.com/datafuselabs/opendal) from [Datafuse Labs](https://github.com/datafuselabs) gives GreptimeDB a very general and elegant data access abstraction layer.
 - GreptimeDB’s meta service is based on [etcd](https://etcd.io/).
+- GreptimeDB uses [RustPython](https://github.com/RustPython/RustPython) for experimental embedded python scripting.

benchmarks/Cargo.toml

@@ -1,14 +1,14 @@
 [package]
 name = "benchmarks"
-version = "0.1.0"
-edition = "2021"
-license = "Apache-2.0"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 
 [dependencies]
-arrow = "10"
+arrow.workspace = true
 clap = { version = "4.0", features = ["derive"] }
 client = { path = "../src/client" }
 indicatif = "0.17.1"
 itertools = "0.10.5"
-parquet = { version = "*" }
-tokio = { version = "1.21", features = ["full"] }
+parquet.workspace = true
+tokio.workspace = true

benchmarks/src/bin/nyc-taxi.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,27 +15,21 @@
 //! Use the taxi trip records from New York City dataset to bench. You can download the dataset from
 //! [here](https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page).
 
-#![feature(once_cell)]
 #![allow(clippy::print_stdout)]
 
 use std::collections::HashMap;
 use std::path::{Path, PathBuf};
-use std::sync::Arc;
 use std::time::Instant;
 
 use arrow::array::{ArrayRef, PrimitiveArray, StringArray, TimestampNanosecondArray};
 use arrow::datatypes::{DataType, Float64Type, Int64Type};
 use arrow::record_batch::RecordBatch;
 use clap::Parser;
-use client::admin::Admin;
-use client::api::v1::codec::InsertBatch;
 use client::api::v1::column::Values;
-use client::api::v1::{insert_expr, Column, ColumnDataType, ColumnDef, CreateExpr, InsertExpr};
-use client::{Client, Database, Select};
+use client::api::v1::{Column, ColumnDataType, ColumnDef, CreateTableExpr, InsertRequest, TableId};
+use client::{Client, Database};
 use indicatif::{MultiProgress, ProgressBar, ProgressStyle};
-use parquet::arrow::{ArrowReader, ParquetFileArrowReader};
-use parquet::file::reader::FileReader;
-use parquet::file::serialized_reader::SerializedFileReader;
+use parquet::arrow::arrow_reader::ParquetRecordBatchReaderBuilder;
 use tokio::task::JoinSet;
 
 const DATABASE_NAME: &str = "greptime";
@@ -87,45 +81,43 @@ async fn write_data(
     pb_style: ProgressStyle,
 ) -> u128 {
     let file = std::fs::File::open(&path).unwrap();
-    let file_reader = Arc::new(SerializedFileReader::new(file).unwrap());
-    let row_num = file_reader.metadata().file_metadata().num_rows();
-    let record_batch_reader = ParquetFileArrowReader::new(file_reader)
-        .get_record_reader(batch_size)
+    let record_batch_reader_builder = ParquetRecordBatchReaderBuilder::try_new(file).unwrap();
+    let row_num = record_batch_reader_builder
+        .metadata()
+        .file_metadata()
+        .num_rows();
+    let record_batch_reader = record_batch_reader_builder
+        .with_batch_size(batch_size)
+        .build()
         .unwrap();
     let progress_bar = mpb.add(ProgressBar::new(row_num as _));
     progress_bar.set_style(pb_style);
-    progress_bar.set_message(format!("{:?}", path));
+    progress_bar.set_message(format!("{path:?}"));
 
     let mut total_rpc_elapsed_ms = 0;
 
     for record_batch in record_batch_reader {
         let record_batch = record_batch.unwrap();
-        let row_count = record_batch.num_rows();
-        let insert_batch = convert_record_batch(record_batch).into();
-        let insert_expr = InsertExpr {
+        let (columns, row_count) = convert_record_batch(record_batch);
+        let request = InsertRequest {
             schema_name: "public".to_string(),
             table_name: TABLE_NAME.to_string(),
-            expr: Some(insert_expr::Expr::Values(insert_expr::Values {
-                values: vec![insert_batch],
-            })),
-            options: HashMap::default(),
             region_number: 0,
+            columns,
+            row_count,
         };
         let now = Instant::now();
-        db.insert(insert_expr).await.unwrap();
+        db.insert(request).await.unwrap();
         let elapsed = now.elapsed();
         total_rpc_elapsed_ms += elapsed.as_millis();
         progress_bar.inc(row_count as _);
     }
 
-    progress_bar.finish_with_message(format!(
-        "file {:?} done in {}ms",
-        path, total_rpc_elapsed_ms
-    ));
+    progress_bar.finish_with_message(format!("file {path:?} done in {total_rpc_elapsed_ms}ms",));
     total_rpc_elapsed_ms
 }
 
-fn convert_record_batch(record_batch: RecordBatch) -> InsertBatch {
+fn convert_record_batch(record_batch: RecordBatch) -> (Vec<Column>, u32) {
     let schema = record_batch.schema();
     let fields = schema.fields();
     let row_count = record_batch.num_rows();
@@ -143,10 +135,7 @@ fn convert_record_batch(record_batch: RecordBatch) -> InsertBatch {
         columns.push(column);
     }
 
-    InsertBatch {
-        columns,
-        row_count: row_count as _,
-    }
+    (columns, row_count as _)
 }
 
 fn build_values(column: &ArrayRef) -> Values {
@@ -217,133 +206,134 @@ fn build_values(column: &ArrayRef) -> Values {
         | DataType::FixedSizeList(_, _)
         | DataType::LargeList(_)
         | DataType::Struct(_)
-        | DataType::Union(_, _)
+        | DataType::Union(_, _, _)
         | DataType::Dictionary(_, _)
-        | DataType::Decimal(_, _)
+        | DataType::Decimal128(_, _)
+        | DataType::Decimal256(_, _)
         | DataType::Map(_, _) => todo!(),
     }
 }
 
-fn create_table_expr() -> CreateExpr {
-    CreateExpr {
-        catalog_name: Some(CATALOG_NAME.to_string()),
-        schema_name: Some(SCHEMA_NAME.to_string()),
+fn create_table_expr() -> CreateTableExpr {
+    CreateTableExpr {
+        catalog_name: CATALOG_NAME.to_string(),
+        schema_name: SCHEMA_NAME.to_string(),
         table_name: TABLE_NAME.to_string(),
-        desc: None,
+        desc: "".to_string(),
         column_defs: vec![
             ColumnDef {
                 name: "VendorID".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "tpep_pickup_datetime".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "tpep_dropoff_datetime".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "passenger_count".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "trip_distance".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "RatecodeID".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "store_and_fwd_flag".to_string(),
                 datatype: ColumnDataType::String as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "PULocationID".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "DOLocationID".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "payment_type".to_string(),
                 datatype: ColumnDataType::Int64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "fare_amount".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "extra".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "mta_tax".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "tip_amount".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "tolls_amount".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "improvement_surcharge".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "total_amount".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "congestion_surcharge".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "airport_fee".to_string(),
                 datatype: ColumnDataType::Float64 as i32,
                 is_nullable: true,
-                default_constraint: None,
+                default_constraint: vec![],
             },
         ],
         time_index: "tpep_pickup_datetime".to_string(),
@@ -351,7 +341,7 @@ fn create_table_expr() -> CreateExpr {
         create_if_not_exists: false,
         table_options: Default::default(),
         region_ids: vec![0],
-        table_id: Some(0),
+        table_id: Some(TableId { id: 0 }),
     }
 }
 
@@ -360,25 +350,23 @@ fn query_set() -> HashMap<String, String> {
 
     ret.insert(
         "count_all".to_string(),
-        format!("SELECT COUNT(*) FROM {};", TABLE_NAME),
+        format!("SELECT COUNT(*) FROM {TABLE_NAME};"),
     );
 
     ret.insert(
         "fare_amt_by_passenger".to_string(),
-        format!("SELECT passenger_count, MIN(fare_amount), MAX(fare_amount), SUM(fare_amount) FROM {} GROUP BY passenger_count",TABLE_NAME)
+        format!("SELECT passenger_count, MIN(fare_amount), MAX(fare_amount), SUM(fare_amount) FROM {TABLE_NAME} GROUP BY passenger_count")
     );
 
     ret
 }
 
-async fn do_write(args: &Args, client: &Client) {
-    let admin = Admin::new("admin", client.clone());
-
+async fn do_write(args: &Args, db: &Database) {
     let mut file_list = get_file_list(args.path.clone().expect("Specify data path in argument"));
     let mut write_jobs = JoinSet::new();
 
-    let create_table_result = admin.create(create_table_expr()).await;
-    println!("Create table result: {:?}", create_table_result);
+    let create_table_result = db.create(create_table_expr()).await;
+    println!("Create table result: {create_table_result:?}");
 
     let progress_bar_style = ProgressStyle::with_template(
         "[{elapsed_precise}] {bar:60.cyan/blue} {pos:>7}/{len:7} {msg}",
@@ -392,7 +380,7 @@ async fn do_write(args: &Args, client: &Client) {
     let batch_size = args.batch_size;
     for _ in 0..args.thread_num {
         if let Some(path) = file_list.pop() {
-            let db = Database::new(DATABASE_NAME, client.clone());
+            let db = db.clone();
             let mpb = multi_progress_bar.clone();
             let pb_style = progress_bar_style.clone();
             write_jobs.spawn(async move { write_data(batch_size, &db, path, mpb, pb_style).await });
@@ -401,7 +389,7 @@ async fn do_write(args: &Args, client: &Client) {
     while write_jobs.join_next().await.is_some() {
         file_progress.inc(1);
         if let Some(path) = file_list.pop() {
-            let db = Database::new(DATABASE_NAME, client.clone());
+            let db = db.clone();
             let mpb = multi_progress_bar.clone();
             let pb_style = progress_bar_style.clone();
             write_jobs.spawn(async move { write_data(batch_size, &db, path, mpb, pb_style).await });
@@ -411,10 +399,10 @@ async fn do_write(args: &Args, client: &Client) {
 
 async fn do_query(num_iter: usize, db: &Database) {
     for (query_name, query) in query_set() {
-        println!("Running query: {}", query);
+        println!("Running query: {query}");
         for i in 0..num_iter {
             let now = Instant::now();
-            let _res = db.select(Select::Sql(query.clone())).await.unwrap();
+            let _res = db.sql(&query).await.unwrap();
             let elapsed = now.elapsed();
             println!(
                 "query {}, iteration {}: {}ms",
@@ -436,13 +424,13 @@ fn main() {
         .unwrap()
         .block_on(async {
             let client = Client::with_urls(vec![&args.endpoint]);
+            let db = Database::new(DATABASE_NAME, client);
 
             if !args.skip_write {
-                do_write(&args, &client).await;
+                do_write(&args, &db).await;
             }
 
             if !args.skip_read {
-                let db = Database::new(DATABASE_NAME, client.clone());
                 do_query(args.iter_num, &db).await;
             }
         })

codecov.yml

@@ -7,3 +7,4 @@ coverage:
     patch: off
 ignore:
   - "**/error*.rs" # ignore all error.rs files
+  - "tests/runner/*.rs" # ignore integration test runner

config/datanode.example.toml

@@ -1,17 +1,25 @@
 node_id = 42
 mode = 'distributed'
-rpc_addr = '0.0.0.0:3001'
-wal_dir = '/tmp/greptimedb/wal'
+rpc_addr = '127.0.0.1:3001'
 rpc_runtime_size = 8
-mysql_addr = '0.0.0.0:3306'
+mysql_addr = '127.0.0.1:4406'
 mysql_runtime_size = 4
+enable_memory_catalog = false
+
+[wal]
+dir = "/tmp/greptimedb/wal"
+file_size = '1GB'
+purge_interval = '10m'
+purge_threshold = '50GB'
+read_batch_size = 128
+sync_write = false
 
 [storage]
 type = 'File'
 data_dir = '/tmp/greptimedb/data/'
 
 [meta_client_opts]
-metasrv_addr = '1.1.1.1:3002'
+metasrv_addrs = ['127.0.0.1:3002']
 timeout_millis = 3000
 connect_timeout_millis = 5000
 tcp_nodelay = false

config/frontend.example.toml

@@ -1,9 +1,12 @@
 mode = 'distributed'
 datanode_rpc_addr = '127.0.0.1:3001'
-http_addr = '0.0.0.0:4000'
+
+[http_options]
+addr = '127.0.0.1:4000'
+timeout = "30s"
 
 [meta_client_opts]
-metasrv_addr = '1.1.1.1:3002'
+metasrv_addrs = ['127.0.0.1:3002']
 timeout_millis = 3000
 connect_timeout_millis = 5000
 tcp_nodelay = false

config/metasrv.example.toml

@@ -1,4 +1,4 @@
 bind_addr = '127.0.0.1:3002'
-server_addr = '0.0.0.0:3002'
-store_addr = '127.0.0.1:2380'
-datanode_lease_secs = 30
+server_addr = '127.0.0.1:3002'
+store_addr = '127.0.0.1:2379'
+datanode_lease_secs = 15

config/standalone.example.toml

@@ -1,27 +1,37 @@
 node_id = 0
 mode = 'standalone'
-http_addr = '0.0.0.0:4000'
-datanode_mysql_addr = '0.0.0.0:3306'
-datanode_mysql_runtime_size = 4
-wal_dir = '/tmp/greptimedb/wal/'
+enable_memory_catalog = false
+
+[http_options]
+addr = '127.0.0.1:4000'
+timeout = "30s"
+
+[wal]
+dir = "/tmp/greptimedb/wal"
+file_size = '1GB'
+purge_interval = '10m'
+purge_threshold = '50GB'
+read_batch_size = 128
+sync_write = false
+
 
 [storage]
 type = 'File'
 data_dir = '/tmp/greptimedb/data/'
 
 [grpc_options]
-addr = '0.0.0.0:4001'
+addr = '127.0.0.1:4001'
 runtime_size = 8
 
 [mysql_options]
-addr = '0.0.0.0:4002'
+addr = '127.0.0.1:4002'
 runtime_size = 2
 
 [influxdb_options]
 enable = true
 
 [opentsdb_options]
-addr = '0.0.0.0:4242'
+addr = '127.0.0.1:4242'
 enable = true
 runtime_size = 2
 
@@ -29,6 +39,6 @@ runtime_size = 2
 enable = true
 
 [postgres_options]
-addr = '0.0.0.0:4003'
+addr = '127.0.0.1:4003'
 runtime_size = 2
 check_pwd = false

docker/Dockerfile

@@ -24,6 +24,8 @@ RUN cargo build --release
 # TODO(zyy17): Maybe should use the more secure container image.
 FROM ubuntu:22.04 as base
 
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get -y install ca-certificates
+
 WORKDIR /greptime
 COPY --from=builder /greptimedb/target/release/greptime /greptime/bin/
 ENV PATH /greptime/bin/:$PATH

docker/ci/Dockerfile

@@ -1,5 +1,7 @@
 FROM ubuntu:22.04
 
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get -y install ca-certificates
+
 ARG TARGETARCH
 
 ADD $TARGETARCH/greptime /greptime/bin/

docs/how-to/how-to-write-aggregate-function.md

@@ -55,7 +55,7 @@ The DataFusion basically execute aggregate like this:
 2. Call `update_batch` on each accumulator with partitioned data, to let you update your aggregate calculation.
 3. Call `state` to get each accumulator's internal state, the medial calculation result.
 4. Call `merge_batch` to merge all accumulator's internal state to one.
-5. Execute `evalute` on the chosen one to get the final calculation result.
+5. Execute `evaluate` on the chosen one to get the final calculation result.
 
 Once you know the meaning of each method, you can easily write your accumulator. You can refer to `Median` accumulator or `SUM` accumulator defined in  file `my_sum_udaf_example.rs` for more details.
 
@@ -63,7 +63,7 @@ Once you know the meaning of each method, you can easily write your accumulator.
 
 You can call `register_aggregate_function` method in query engine to register your aggregate function. To do that, you have to new an instance of struct `AggregateFunctionMeta`. The struct has three fields, first is the name of your aggregate function's name. The function name is case-sensitive due to DataFusion's restriction. We strongly recommend using lowercase for your name. If you have to use uppercase name, wrap your aggregate function with quotation marks. For example, if you define an aggregate function named "my_aggr", you can use "`SELECT MY_AGGR(x)`"; if you define "my_AGGR", you have to use "`SELECT "my_AGGR"(x)`".
 
-The second field is arg_counts ,the count of the arguments. Like accumulator `percentile`, caculating the p_number of the column. We need to input the value of column and the value of p to cacalate, and so the count of the arguments is two.
+The second field is arg_counts ,the count of the arguments. Like accumulator `percentile`, calculating the p_number of the column. We need to input the value of column and the value of p to cacalate, and so the count of the arguments is two.
 
 The third field is a function about how to create your accumulator creator that you defined in step 1 above. Create creator, that's a bit intertwined, but it is how we make DataFusion use a newly created aggregate function each time it executes a SQL, preventing the stored input types from affecting each other. The key detail can be starting looking at our `DfContextProviderAdapter` struct's `get_aggregate_meta` method.
 

docs/rfcs/2022-12-20-promql-in-rust/rfc.md

@@ -0,0 +1,175 @@
+---
+Feature Name: "promql-in-rust"
+Tracking Issue: https://github.com/GreptimeTeam/greptimedb/issues/596
+Date: 2022-12-20
+Author: "Ruihang Xia <waynestxia@gmail.com>"
+---
+
+Rewrite PromQL in Rust
+----------------------
+
+# Summary
+A Rust native implementation of PromQL, for GreptimeDB.
+
+# Motivation
+Prometheus and its query language PromQL prevails in the cloud-native observability area, which is an important scenario for time series database like GreptimeDB. We already have support for its remote read and write protocols. Users can now integrate GreptimeDB as the storage backend to existing Prometheus deployment, but cannot run PromQL query directly on GreptimeDB like SQL.
+
+This RFC proposes to add support for PromQL. Because it was created in Go, we can't use the existing code easily. For interoperability, performance and extendability, porting its logic to Rust is a good choice.
+
+# Details
+
+## Overview
+One of the goals is to make use of our existing basic operators, execution model and runtime to reduce the work. So the entire proposal is built on top of Apache Arrow DataFusion. The rewrote PromQL logic is manifested as `Expr` or `Execution Plan` in DataFusion. And both the intermediate data structure and the result is in the format of `Arrow`'s `RecordBatch`.
+
+The following sections are organized in a top-down manner. Starts with evaluation procedure. Then introduces the building blocks of our new PromQL operation. Follows by an explanation of data model. And end with an example logic plan.
+
+*This RFC is heavily related to Prometheus and PromQL. It won't repeat some basic concepts of them.*
+
+## Evaluation
+
+The original implementation is like an interpreter of parsed PromQL AST. It has two characteristics: (1) Operations are evaluated in place after they are parsed to AST. And some key parameters are separated from the AST because they do not present in the query, but come from other places like another field in the HTTP payload. (2) calculation is performed per timestamp. You can see this pattern many times:
+```go
+for ts := ev.startTimestamp; ts <= ev.endTimestamp; ts += ev.interval {}
+```
+
+These bring out two differences in the proposed implementation. First, to make it more general and clear, the evaluation procedure is reorganized into serval phases (and is the same as DataFusion's). And second, data are evaluated by time series (corresponding to "columnar calculation", if think timestamp as row number).
+
+```
+                                      Logic
+   Query             AST              Plan
+ ─────────► Parser ───────► Logical ────────► Physical ────┐
+                            Planner           Planner      │
+                                                           │
+ ◄───────────────────────────── Executor  ◄────────────────┘
+        Evaluation Result                     Execution
+                                                Plan
+```
+
+- Parser 
+
+    Provided by [`promql-parser`](https://github.com/GreptimeTeam/promql-parser) crate. Same as the original implementation.
+
+- Logical Planner
+
+    Generates a logical plan with all the needed parameters. It should accept something like `EvalStmt` in Go's implementation, which contains query time range, evaluation interval and lookback range.
+
+    Another important thing done here is assembling the logic plan, with all the operations baked into logically. Like what's the filter and time range to read, how the data then flows through a selector into a binary operation, etc. Or what's the output schema of every single step. The generated logic plan is deterministic without variables, and can be `EXPLAIN`ed clearly.
+
+- Physical Planner
+
+    This step converts a logic plan into evaluatable execution plan. There are not many special things like the previous step. Except when a query is going to be executed distributedly. In this case, a logic plan will be divided into serval parts and sent to serval nodes. One physical planner only sees its own part.
+
+- Executor
+
+    As its name shows, this step calculates data to result. And all new calculation logic, the implementation of PromQL in rust, is placed here. And the rewrote functions are using `RecordBatch` and `Array` from `Arrow` as the intermediate data structure.
+
+    Each "batch" contains only data from single time series. This is from the underlying storage implementation. Though it's not a requirement of this RFC, having this property can simplify some functions.
+    
+    Another thing to mention is the rewrote functions don't aware of timestamp or value columns, they are defined only based on the input data types. For example, `increase()` function in PromQL calculates the unbiased delta of data, its implementation here only does this single thing. Let's compare the signature of two implementations:
+
+    - Go
+        ```go
+        func funcIncrease(vals []parser.Value, args parser.Expressions) Vector {}
+        ```
+    - Rust
+        ```rust
+        fn prom_increase(input: Array) -> Array {}
+        ```
+
+    Some unimportant parameters are omitted. The original Go version only writes the logic for `Point`'s value, either float or histogram. But the proposed rewritten one accepts a generic `Array` as input, which can be any type that suits, from `i8` to `u64` to `TimestampNanosecond`.
+
+## Plan and Expression
+
+They are structures to express logic from PromQL. The proposed implementation is built on top of DataFusion, thus our plan and expression are in form of `ExtensionPlan` and `ScalarUDF`. The only difference between them in this context is the return type: plan returns a record batch while expression returns a single column.
+
+This RFC proposes to add four new plans, they are fundamental building blocks that mainly handle data selection logic in PromQL, for the following calculation expressions. 
+
+- `SeriesNormalize`
+
+    Sort data inside one series on the timestamp column, and bias "offset" if has. This plan usually comes after `TableScan` (or `TableScan` and `Filter`) plan.
+
+- `VectorManipulator` and `MatrixManipulator`
+
+    Corresponding to `InstantSelector` and `RangeSelector`. We don't calculate timestamp by timestamp, thus use "vector" instead of "instant", this image shows the difference. And "matrix" is another name for "range vector", for not confused with our "vector". The following section will detail how they are implemented using Arrow.
+
+    ![instant_and_vector](instant-and-vector.png)
+
+    Due to "interval" parameter in PromQL, data after "selector" (or "manipulator" here) are usually shorter than input. And we have to modify the entire record batch to shorten both timestamp, value and tag columns. So they are formed as plan.
+
+- `PromAggregator`
+
+    The carrier of aggregator expressions. This should not be very different from the DataFusion built-in `Aggregate` plan, except PromQL can use "group without" to do reverse selection.
+
+PromQL has around 70 expressions and functions. But luckily we can reuse lots of them from DataFusion. Like unary expression, binary expression and aggregator. We only need to implement those PromQL-specific expressions, like `rate` or `percentile`. The following table lists some typical functions in PromQL, and their signature in the proposed implementation. Other function should be the same.
+
+| Name               	| In Param(s)                                          	| Out Param(s) 	| Explain            	|
+|--------------------	|------------------------------------------------------	|--------------	|--------------------	|
+| instant_delta      	| Matrix T                                             	| Array T      	| idelta in PromQL   	|
+| increase           	| Matrix T                                             	| Array T      	| increase in PromQL 	|
+| extrapolate_factor 	| - Matrix T<br>- Array Timestamp<br>- Array Timestamp 	| Array T      	| *                   	|
+
+*: *`extrapolate_factor` is one of the "dark sides" in PromQL. In short it's a translation of this [paragraph](https://github.com/prometheus/prometheus/blob/0372e259baf014bbade3134fd79bcdfd8cbdef2c/promql/functions.go#L134-L159)*
+
+To reuse those common calculation logic, we can break them into serval expressions, and assemble in the logic planning phase. Like `rate()` in PromQL can be represented as `increase / extrapolate_factor`.
+
+## Data Model
+
+This part explains how data is represented. Following the data model in GreptimeDB, all the data are stored as table, with tag columns, timestamp column and value column. Table to record batch is very straightforward. So an instant vector can be thought of as a row (though as said before, we don't use instant vectors) in the table. Given four basic types in PromQL: scalar, string, instant vector and range vector, only the last "range vector" need some tricks to adapt our columnar calculation.
+
+Range vector is some sort of matrix, it's consisted of small one-dimension vectors, with each being an input of range function. And, applying range function to a range vector can be thought of kind of convolution.
+
+![range-vector-with-matrix](range-vector-with-matrix.png)
+
+(Left is an illustration of range vector. Notice the Y-axis has no meaning, it's just put different pieces separately. The right side is an imagined "matrix" as range function. Multiplying the left side to it can get a one-dimension "matrix" with four elements. That's the evaluation result of a range vector.)
+
+To adapt this range vector to record batch, it should be represented by a column. This RFC proposes to use `DictionaryArray` from Arrow to represent range vector, or `Matrix`. This is "misusing" `DictionaryArray` to ship some additional information about an array. Because the range vector is sliding over one series, we only need to know the `offset` and `length` of each slides to reconstruct the matrix from an array:
+
+![matrix-from-array](matrix-from-array.png)
+
+The length is not fixed, it depends on the input's timestamp. An PoC implementation of `Matrix` and `increase()` can be found in [this repo](https://github.com/waynexia/corroding-prometheus).
+
+## Example
+
+The logic plan of PromQL query
+```promql
+# start: 2022-12-20T10:00:00
+# end: 2022-12-21T10:00:00
+# interval: 1m
+# lookback: 30s
+sum (rate(request_duration[5m])) by (idc)
+```
+looks like
+
+<!-- title: 'PromAggregator: \naggr = sum, column = idc'
+operator: prom
+inputs:
+- title: 'Matrix Manipulator: \ninterval = 1m, range = 5m, expr = div(increase(value), extrapolate_factor(timestamp))'
+  operator: prom
+  inputs:
+  - title: 'Series Normalize: \noffset = 0'
+    operator: prom
+    inputs:
+    - title: 'Filter: \ntimetamp > 2022-12-20T10:00:00 && timestamp < 2022-12-21T10:00:00'
+      operator: filter
+      inputs:
+      - title: 'Table Scan: \ntable = request_duration, timetamp > 2022-12-20T10:00:00 && timestamp < 2022-12-21T10:00:00'
+        operator: scan -->
+
+![example](example.png)
+
+# Drawbacks
+
+Human-being is always error-prone. It's harder to endeavor to rewrite from the ground and requires more attention to ensure correctness, than translate line-by-line. And, since the evaluator's architecture are different, it might be painful to catch up with PromQL's breaking update (if any) in the future.
+
+Misusing Arrow's DictionaryVector as Matrix is another point. This hack needs some `unsafe` function call to bypass Arrow's check. And though Arrow's API is stable, this is still an undocumented behavior.
+
+# Alternatives
+
+There are a few alternatives we've considered:
+- Wrap the existing PromQL's implementation via FFI, and import it to GreptimeDB.
+- Translate its evaluator engine line-by-line, rather than rewrite one.
+- Integrate the Prometheus server into GreptimeDB via RPC, making it a detached execution engine for PromQL.
+
+The first and second options are making a separate execution engine in GreptimeDB, they may alleviate the pain during rewriting, but will have negative impacts to afterward evolve like resource management. And introduce another deploy component in the last option will bring a complex deploy architecture.
+
+And all of them are more or less redundant in data transportation that affects performance and resources. The proposed built-in executing procedure is also easy to integrate and expose to the existing SQL interface GreptimeDB currently provides. Some concepts in PromQL like sliding windows (range vector in PromQL) are very convenient and ergonomic in analyzing series data. This makes it not only a PromQL evaluator, but also an enhancement to our query system.

rust-toolchain

@@ -1 +1 @@
-nightly-2022-07-14
+nightly-2022-12-20

src/api/Cargo.toml

@@ -1,18 +1,18 @@
 [package]
 name = "api"
-version = "0.1.0"
-edition = "2021"
-license = "Apache-2.0"
-# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 
 [dependencies]
+arrow-flight.workspace = true
 common-base = { path = "../common/base" }
-common-time = { path = "../common/time" }
 common-error = { path = "../common/error" }
+common-time = { path = "../common/time" }
 datatypes = { path = "../datatypes" }
-prost = "0.11"
+prost.workspace = true
 snafu = { version = "0.7", features = ["backtraces"] }
-tonic = "0.8"
+tonic.workspace = true
 
 [build-dependencies]
 tonic-build = "0.8"

src/api/build.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,18 +12,11 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use std::path::PathBuf;
-
 fn main() {
-    let default_out_dir = PathBuf::from(std::env::var("OUT_DIR").unwrap());
     tonic_build::configure()
-        .file_descriptor_set_path(default_out_dir.join("greptime_fd.bin"))
         .compile(
             &[
-                "greptime/v1/insert.proto",
-                "greptime/v1/select.proto",
-                "greptime/v1/physical_plan.proto",
-                "greptime/v1/greptime.proto",
+                "greptime/v1/database.proto",
                 "greptime/v1/meta/common.proto",
                 "greptime/v1/meta/heartbeat.proto",
                 "greptime/v1/meta/route.proto",

src/api/greptime/v1/admin.proto

@@ -1,69 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1;
-
-import "greptime/v1/column.proto";
-import "greptime/v1/common.proto";
-
-message AdminRequest {
-  string name = 1;
-  repeated AdminExpr exprs = 2;
-}
-
-message AdminResponse {
-  repeated AdminResult results = 1;
-}
-
-message AdminExpr {
-  ExprHeader header = 1;
-  oneof expr {
-    CreateExpr create = 2;
-    AlterExpr alter = 3;
-    CreateDatabaseExpr create_database = 4;
-  }
-}
-
-message AdminResult {
-  ResultHeader header = 1;
-  oneof result {
-    MutateResult mutate = 2;
-  }
-}
-
-// TODO(hl): rename to CreateTableExpr
-message CreateExpr {
-  optional string catalog_name = 1;
-  optional string schema_name = 2;
-  string table_name = 3;
-  optional string desc = 4;
-  repeated ColumnDef column_defs = 5;
-  string time_index = 6;
-  repeated string primary_keys = 7;
-  bool create_if_not_exists = 8;
-  map<string, string> table_options = 9;
-  optional uint32 table_id = 10;
-  repeated uint32 region_ids = 11;
-}
-
-message AlterExpr {
-  optional string catalog_name = 1;
-  optional string schema_name = 2;
-  string table_name = 3;
-  oneof kind {
-    AddColumns add_columns = 4;
-  }
-}
-
-message AddColumns {
-  repeated AddColumn add_columns = 1;
-}
-
-message AddColumn {
-  ColumnDef column_def = 1;
-  bool is_key = 2;
-}
-
-message CreateDatabaseExpr {
-  //TODO(hl): maybe rename to schema_name?
-  string database_name = 1;
-}

src/api/greptime/v1/column.proto

@@ -32,7 +32,10 @@ message Column {
 
     repeated int32 date_values = 14;
     repeated int64 datetime_values = 15;
-    repeated int64 ts_millis_values = 16;
+    repeated int64 ts_second_values = 16;
+    repeated int64 ts_millisecond_values = 17;
+    repeated int64 ts_microsecond_values = 18;
+    repeated int64 ts_nanosecond_values = 19;
   }
   // The array of non-null values in this column.
   //
@@ -56,7 +59,7 @@ message ColumnDef {
   string name = 1;
   ColumnDataType datatype = 2;
   bool is_nullable = 3;
-  optional bytes default_constraint = 4;
+  bytes default_constraint = 4;
 }
 
 enum ColumnDataType {
@@ -75,5 +78,8 @@ enum ColumnDataType {
   STRING = 12;
   DATE = 13;
   DATETIME = 14;
-  TIMESTAMP = 15;
+  TIMESTAMP_SECOND = 15;
+  TIMESTAMP_MILLISECOND = 16;
+  TIMESTAMP_MICROSECOND = 17;
+  TIMESTAMP_NANOSECOND = 18;
 }

src/api/greptime/v1/common.proto

@@ -1,22 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1;
-
-message RequestHeader {
-  string tenant = 1;
-}
-
-message ExprHeader {
-  uint32 version = 1;
-}
-
-message ResultHeader {
-  uint32 version = 1;
-  uint32 code = 2;
-  string err_msg = 3;
-}
-
-message MutateResult {
-  uint32 success = 1;
-  uint32 failure = 2;
-}

src/api/greptime/v1/database.proto

@@ -2,80 +2,40 @@ syntax = "proto3";
 
 package greptime.v1;
 
-import "greptime/v1/common.proto";
+import "greptime/v1/ddl.proto";
+import "greptime/v1/column.proto";
 
-message DatabaseRequest {
-  string name = 1;
-  repeated ObjectExpr exprs = 2;
-}
-
-message DatabaseResponse {
-  repeated ObjectResult results = 1;
-}
-
-message ObjectExpr {
-  ExprHeader header = 1;
-  oneof expr {
-    InsertExpr insert = 2;
-    SelectExpr select = 3;
-    UpdateExpr update = 4;
-    DeleteExpr delete = 5;
+message GreptimeRequest {
+  oneof request {
+    InsertRequest insert = 1;
+    QueryRequest query = 2;
+    DdlRequest ddl = 3;
   }
 }
 
-// TODO(fys): Only support sql now, and will support promql etc in the future
-message SelectExpr {
-  oneof expr {
+message QueryRequest {
+  oneof query {
     string sql = 1;
     bytes logical_plan = 2;
-    PhysicalPlan physical_plan = 15;
   }
 }
 
-message PhysicalPlan {
-  bytes original_ql = 1;
-  bytes plan = 2;
-}
-
-message InsertExpr {
+message InsertRequest {
   string schema_name = 1;
   string table_name = 2;
 
-  message Values {
-    repeated bytes values = 3;
-  }
+  // Data is represented here.
+  repeated Column columns = 3;
 
-  oneof expr {
-    Values values = 4;
+  // The row_count of all columns, which include null and non-null values.
+  //
+  // Note: the row_count of all columns in a InsertRequest must be same.
+  uint32 row_count = 4;
 
-    // TODO(LFC): Remove field "sql" in InsertExpr.
-    // When Frontend instance received an insertion SQL (`insert into ...`), it's anticipated to parse the SQL and
-    // assemble the values to insert to feed Datanode. In other words, inserting data through Datanode instance's GRPC
-    // interface shouldn't use SQL directly.
-    // Then why the "sql" field exists here? It's because the Frontend needs table schema to create the values to insert,
-    // which is currently not able to find anywhere. (Maybe the table schema is suppose to be fetched from Meta?)
-    // The "sql" field is meant to be removed in the future.
-    string sql = 5;
-  }
-
-  /// The region number of current insert request.
-  uint32 region_number = 6;
-  map<string, bytes> options = 7;
+  // The region number of current insert request.
+  uint32 region_number = 5;
 }
 
-// TODO(jiachun)
-message UpdateExpr {}
-// TODO(jiachun)
-message DeleteExpr {}
-
-message ObjectResult {
-  ResultHeader header = 1;
-  oneof result {
-    SelectResult select = 2;
-    MutateResult mutate = 3;
-  }
-}
-
-message SelectResult {
-  bytes raw_data = 1;
+message FlightDataExt {
+  uint32 affected_rows = 1;
 }

src/api/greptime/v1/ddl.proto

@@ -0,0 +1,79 @@
+syntax = "proto3";
+
+package greptime.v1;
+
+import "greptime/v1/column.proto";
+
+// "Data Definition Language" requests, that create, modify or delete the database structures but not the data.
+// `DdlRequest` could carry more information than plain SQL, for example, the "table_id" in `CreateTableExpr`.
+// So create a new DDL expr if you need it.
+message DdlRequest {
+  oneof expr {
+    CreateDatabaseExpr create_database = 1;
+    CreateTableExpr create_table = 2;
+    AlterExpr alter = 3;
+    DropTableExpr drop_table = 4;
+  }
+}
+
+message CreateTableExpr {
+  string catalog_name = 1;
+  string schema_name = 2;
+  string table_name = 3;
+  string desc = 4;
+  repeated ColumnDef column_defs = 5;
+  string time_index = 6;
+  repeated string primary_keys = 7;
+  bool create_if_not_exists = 8;
+  map<string, string> table_options = 9;
+  TableId table_id = 10;
+  repeated uint32 region_ids = 11;
+}
+
+message AlterExpr {
+  string catalog_name = 1;
+  string schema_name = 2;
+  string table_name = 3;
+  oneof kind {
+    AddColumns add_columns = 4;
+    DropColumns drop_columns = 5;
+    RenameTable rename_table = 6;
+  }
+}
+
+message DropTableExpr {
+  string catalog_name = 1;
+  string schema_name = 2;
+  string table_name = 3;
+}
+
+message CreateDatabaseExpr {
+  //TODO(hl): maybe rename to schema_name?
+  string database_name = 1;
+  bool create_if_not_exists = 2;
+}
+
+message AddColumns {
+  repeated AddColumn add_columns = 1;
+}
+
+message DropColumns {
+  repeated DropColumn drop_columns = 1;
+}
+
+message RenameTable {
+  string new_table_name = 1;
+}
+
+message AddColumn {
+  ColumnDef column_def = 1;
+  bool is_key = 2;
+}
+
+message DropColumn {
+  string name = 1;
+}
+
+message TableId {
+  uint32 id = 1;
+}

src/api/greptime/v1/greptime.proto

@@ -1,22 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1;
-
-import "greptime/v1/admin.proto";
-import "greptime/v1/common.proto";
-import "greptime/v1/database.proto";
-
-service Greptime {
-  rpc Batch(BatchRequest) returns (BatchResponse) {}
-}
-
-message BatchRequest {
-  RequestHeader header = 1;
-  repeated AdminRequest admins = 2;
-  repeated DatabaseRequest databases = 3;
-}
-
-message BatchResponse {
-  repeated AdminResponse admins = 1;
-  repeated DatabaseResponse databases = 2;
-}

src/api/greptime/v1/insert.proto

@@ -1,14 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1.codec;
-
-import "greptime/v1/column.proto";
-
-message InsertBatch {
-  repeated Column columns = 1;
-  uint32 row_count = 2;
-}
-
-message RegionNumber {
-  uint32 id = 1;
-}

src/api/greptime/v1/meta/heartbeat.proto

@@ -26,7 +26,7 @@ message HeartbeatRequest {
   TimeInterval report_interval = 4;
   // Node stat
   NodeStat node_stat = 5;
-  // Region stats in this node
+  // Region stats on this node
   repeated RegionStat region_stats = 6;
   // Follower nodes and stats, empty on follower nodes
   repeated ReplicaStat replica_stats = 7;
@@ -34,19 +34,19 @@ message HeartbeatRequest {
 
 message NodeStat {
   // The read capacity units during this period
-  uint64 rcus = 1;
+  int64 rcus = 1;
   // The write capacity units during this period
-  uint64 wcus = 2;
-  // Table number in this node
-  uint64 table_num = 3;
-  // Regon number in this node
-  uint64 region_num = 4;
+  int64 wcus = 2;
+  // How many tables on this node
+  int64 table_num = 3;
+  // How many regions on this node
+  int64 region_num = 4;
 
   double cpu_usage = 5;
   double load = 6;
-  // Read disk I/O in the node
+  // Read disk IO on this node
   double read_io_rate = 7;
-  // Write disk I/O in the node
+  // Write disk IO on this node
   double write_io_rate = 8;
 
   // Others
@@ -57,13 +57,13 @@ message RegionStat {
   uint64 region_id = 1;
   TableName table_name = 2;
   // The read capacity units during this period
-  uint64 rcus = 3;
+  int64 rcus = 3;
   // The write capacity units during this period
-  uint64 wcus = 4;
-  // Approximate region size
-  uint64 approximate_size = 5;
-  // Approximate number of rows
-  uint64 approximate_rows = 6;
+  int64 wcus = 4;
+  // Approximate bytes of this region
+  int64 approximate_bytes = 5;
+  // Approximate number of rows in this region
+  int64 approximate_rows = 6;
 
   // Others
   map<string, string> attrs = 100;

src/api/greptime/v1/meta/route.proto

@@ -5,6 +5,8 @@ package greptime.v1.meta;
 import "greptime/v1/meta/common.proto";
 
 service Router {
+  rpc Create(CreateRequest) returns (RouteResponse) {}
+
   // Fetch routing information for tables. The smallest unit is the complete
   // routing information(all regions) of a table.
   // 
@@ -26,7 +28,14 @@ service Router {
   //
   rpc Route(RouteRequest) returns (RouteResponse) {}
 
-  rpc Create(CreateRequest) returns (RouteResponse) {}
+  rpc Delete(DeleteRequest) returns (RouteResponse) {}
+}
+
+message CreateRequest {
+  RequestHeader header = 1;
+
+  TableName table_name = 2;
+  repeated Partition partitions = 3;
 }
 
 message RouteRequest {
@@ -35,6 +44,12 @@ message RouteRequest {
   repeated TableName table_names = 2;
 }
 
+message DeleteRequest {
+  RequestHeader header = 1;
+
+  TableName table_name = 2; 
+}
+
 message RouteResponse {
   ResponseHeader header = 1;
 
@@ -42,13 +57,6 @@ message RouteResponse {
   repeated TableRoute table_routes = 3;
 }
 
-message CreateRequest {
-  RequestHeader header = 1;
-
-  TableName table_name = 2;
-  repeated Partition partitions = 3;
-}
-
 message TableRoute {
   Table table = 1;
   repeated RegionRoute region_routes = 2;

src/api/greptime/v1/meta/store.proto

@@ -20,6 +20,9 @@ service Store {
 
   // DeleteRange deletes the given range from the key-value store.
   rpc DeleteRange(DeleteRangeRequest) returns (DeleteRangeResponse);
+
+  // MoveValue atomically renames the key to the given updated key.
+  rpc MoveValue(MoveValueRequest) returns (MoveValueResponse);
 }
 
 message RangeRequest {
@@ -136,3 +139,21 @@ message DeleteRangeResponse {
   // returned.
   repeated KeyValue prev_kvs = 3;
 }
+
+message MoveValueRequest {
+  RequestHeader header = 1;
+
+  // If from_key dose not exist, return the value of to_key (if it exists).
+  // If from_key exists, move the value of from_key to to_key (i.e. rename),
+  // and return the value.
+  bytes from_key = 2;
+  bytes to_key = 3;
+}
+
+message MoveValueResponse {
+  ResponseHeader header = 1;
+
+  // If from_key dose not exist, return the value of to_key (if it exists).
+  // If from_key exists, return the value of from_key.
+  KeyValue kv = 2;
+}

src/api/greptime/v1/physical_plan.proto

@@ -1,33 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1.codec;
-
-message PhysicalPlanNode {
-  oneof PhysicalPlanType {
-    ProjectionExecNode projection = 1;
-    MockInputExecNode mock = 99;
-    // TODO(fys): impl other physical plan node
-  }
-}
-
-message ProjectionExecNode {
-  PhysicalPlanNode input = 1;
-  repeated PhysicalExprNode expr = 2;
-  repeated string expr_name = 3;
-}
-
-message PhysicalExprNode {
-  oneof ExprType {
-    PhysicalColumn column = 1;
-    // TODO(fys): impl other physical expr node
-  }
-}
-
-message PhysicalColumn {
-  string name = 1;
-  uint64 index = 2;
-}
-
-message MockInputExecNode {
-  string name = 1;
-}

src/api/greptime/v1/select.proto

@@ -1,10 +0,0 @@
-syntax = "proto3";
-
-package greptime.v1.codec;
-
-import "greptime/v1/column.proto";
-
-message SelectResult {
-  repeated Column columns = 1;
-  uint32 row_count = 2;
-}

src/api/src/error.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -33,6 +33,28 @@ pub enum Error {
         from: ConcreteDataType,
         backtrace: Backtrace,
     },
+
+    #[snafu(display(
+        "Failed to convert column default constraint, column: {}, source: {}",
+        column,
+        source
+    ))]
+    ConvertColumnDefaultConstraint {
+        column: String,
+        #[snafu(backtrace)]
+        source: datatypes::error::Error,
+    },
+
+    #[snafu(display(
+        "Invalid column default constraint, column: {}, source: {}",
+        column,
+        source
+    ))]
+    InvalidColumnDefaultConstraint {
+        column: String,
+        #[snafu(backtrace)]
+        source: datatypes::error::Error,
+    },
 }
 
 impl ErrorExt for Error {
@@ -40,6 +62,8 @@ impl ErrorExt for Error {
         match self {
             Error::UnknownColumnDataType { .. } => StatusCode::InvalidArguments,
             Error::IntoColumnDataType { .. } => StatusCode::Unexpected,
+            Error::ConvertColumnDefaultConstraint { source, .. }
+            | Error::InvalidColumnDefaultConstraint { source, .. } => source.status_code(),
         }
     }
     fn backtrace_opt(&self) -> Option<&Backtrace> {

src/api/src/helper.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,6 +15,7 @@
 use common_base::BitVec;
 use common_time::timestamp::TimeUnit;
 use datatypes::prelude::ConcreteDataType;
+use datatypes::types::TimestampType;
 use datatypes::value::Value;
 use datatypes::vectors::VectorRef;
 use snafu::prelude::*;
@@ -56,7 +57,16 @@ impl From<ColumnDataTypeWrapper> for ConcreteDataType {
             ColumnDataType::String => ConcreteDataType::string_datatype(),
             ColumnDataType::Date => ConcreteDataType::date_datatype(),
             ColumnDataType::Datetime => ConcreteDataType::datetime_datatype(),
-            ColumnDataType::Timestamp => ConcreteDataType::timestamp_millis_datatype(),
+            ColumnDataType::TimestampSecond => ConcreteDataType::timestamp_second_datatype(),
+            ColumnDataType::TimestampMillisecond => {
+                ConcreteDataType::timestamp_millisecond_datatype()
+            }
+            ColumnDataType::TimestampMicrosecond => {
+                ConcreteDataType::timestamp_microsecond_datatype()
+            }
+            ColumnDataType::TimestampNanosecond => {
+                ConcreteDataType::timestamp_nanosecond_datatype()
+            }
         }
     }
 }
@@ -81,7 +91,12 @@ impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
             ConcreteDataType::String(_) => ColumnDataType::String,
             ConcreteDataType::Date(_) => ColumnDataType::Date,
             ConcreteDataType::DateTime(_) => ColumnDataType::Datetime,
-            ConcreteDataType::Timestamp(_) => ColumnDataType::Timestamp,
+            ConcreteDataType::Timestamp(unit) => match unit {
+                TimestampType::Second(_) => ColumnDataType::TimestampSecond,
+                TimestampType::Millisecond(_) => ColumnDataType::TimestampMillisecond,
+                TimestampType::Microsecond(_) => ColumnDataType::TimestampMicrosecond,
+                TimestampType::Nanosecond(_) => ColumnDataType::TimestampNanosecond,
+            },
             ConcreteDataType::Null(_) | ConcreteDataType::List(_) => {
                 return error::IntoColumnDataTypeSnafu { from: datatype }.fail()
             }
@@ -153,8 +168,20 @@ impl Values {
                 datetime_values: Vec::with_capacity(capacity),
                 ..Default::default()
             },
-            ColumnDataType::Timestamp => Values {
-                ts_millis_values: Vec::with_capacity(capacity),
+            ColumnDataType::TimestampSecond => Values {
+                ts_second_values: Vec::with_capacity(capacity),
+                ..Default::default()
+            },
+            ColumnDataType::TimestampMillisecond => Values {
+                ts_millisecond_values: Vec::with_capacity(capacity),
+                ..Default::default()
+            },
+            ColumnDataType::TimestampMicrosecond => Values {
+                ts_microsecond_values: Vec::with_capacity(capacity),
+                ..Default::default()
+            },
+            ColumnDataType::TimestampNanosecond => Values {
+                ts_nanosecond_values: Vec::with_capacity(capacity),
                 ..Default::default()
             },
         }
@@ -187,9 +214,12 @@ impl Column {
             Value::Binary(val) => values.binary_values.push(val.to_vec()),
             Value::Date(val) => values.date_values.push(val.val()),
             Value::DateTime(val) => values.datetime_values.push(val.val()),
-            Value::Timestamp(val) => values
-                .ts_millis_values
-                .push(val.convert_to(TimeUnit::Millisecond)),
+            Value::Timestamp(val) => match val.unit() {
+                TimeUnit::Second => values.ts_second_values.push(val.value()),
+                TimeUnit::Millisecond => values.ts_millisecond_values.push(val.value()),
+                TimeUnit::Microsecond => values.ts_microsecond_values.push(val.value()),
+                TimeUnit::Nanosecond => values.ts_nanosecond_values.push(val.value()),
+            },
             Value::List(_) => unreachable!(),
         });
         self.null_mask = null_mask.into_vec();
@@ -200,7 +230,10 @@ impl Column {
 mod tests {
     use std::sync::Arc;
 
-    use datatypes::vectors::BooleanVector;
+    use datatypes::vectors::{
+        BooleanVector, TimestampMicrosecondVector, TimestampMillisecondVector,
+        TimestampNanosecondVector, TimestampSecondVector,
+    };
 
     use super::*;
 
@@ -258,8 +291,8 @@ mod tests {
         let values = values.datetime_values;
         assert_eq!(2, values.capacity());
 
-        let values = Values::with_capacity(ColumnDataType::Timestamp, 2);
-        let values = values.ts_millis_values;
+        let values = Values::with_capacity(ColumnDataType::TimestampMillisecond, 2);
+        let values = values.ts_millisecond_values;
         assert_eq!(2, values.capacity());
     }
 
@@ -326,8 +359,8 @@ mod tests {
             ColumnDataTypeWrapper(ColumnDataType::Datetime).into()
         );
         assert_eq!(
-            ConcreteDataType::timestamp_millis_datatype(),
-            ColumnDataTypeWrapper(ColumnDataType::Timestamp).into()
+            ConcreteDataType::timestamp_millisecond_datatype(),
+            ColumnDataTypeWrapper(ColumnDataType::TimestampMillisecond).into()
         );
     }
 
@@ -394,8 +427,8 @@ mod tests {
             ConcreteDataType::datetime_datatype().try_into().unwrap()
         );
         assert_eq!(
-            ColumnDataTypeWrapper(ColumnDataType::Timestamp),
-            ConcreteDataType::timestamp_millis_datatype()
+            ColumnDataTypeWrapper(ColumnDataType::TimestampMillisecond),
+            ConcreteDataType::timestamp_millisecond_datatype()
                 .try_into()
                 .unwrap()
         );
@@ -412,7 +445,48 @@ mod tests {
         assert!(result.is_err());
         assert_eq!(
             result.unwrap_err().to_string(),
-            "Failed to create column datatype from List(ListType { inner: Boolean(BooleanType) })"
+            "Failed to create column datatype from List(ListType { item_type: Boolean(BooleanType) })"
+        );
+    }
+
+    #[test]
+    fn test_column_put_timestamp_values() {
+        let mut column = Column {
+            column_name: "test".to_string(),
+            semantic_type: 0,
+            values: Some(Values {
+                ..Default::default()
+            }),
+            null_mask: vec![],
+            datatype: 0,
+        };
+
+        let vector = Arc::new(TimestampNanosecondVector::from_vec(vec![1, 2, 3]));
+        column.push_vals(3, vector);
+        assert_eq!(
+            vec![1, 2, 3],
+            column.values.as_ref().unwrap().ts_nanosecond_values
+        );
+
+        let vector = Arc::new(TimestampMillisecondVector::from_vec(vec![4, 5, 6]));
+        column.push_vals(3, vector);
+        assert_eq!(
+            vec![4, 5, 6],
+            column.values.as_ref().unwrap().ts_millisecond_values
+        );
+
+        let vector = Arc::new(TimestampMicrosecondVector::from_vec(vec![7, 8, 9]));
+        column.push_vals(3, vector);
+        assert_eq!(
+            vec![7, 8, 9],
+            column.values.as_ref().unwrap().ts_microsecond_values
+        );
+
+        let vector = Arc::new(TimestampSecondVector::from_vec(vec![10, 11, 12]));
+        column.push_vals(3, vector);
+        assert_eq!(
+            vec![10, 11, 12],
+            column.values.as_ref().unwrap().ts_second_values
         );
     }
 

src/api/src/lib.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,7 +15,6 @@
 pub mod error;
 pub mod helper;
 pub mod prometheus;
-pub mod result;
 pub mod serde;
 pub mod v1;
 

src/api/src/prometheus.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/api/src/result.rs

@@ -1,203 +0,0 @@
-// Copyright 2022 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.
-
-use common_error::prelude::ErrorExt;
-
-use crate::v1::codec::SelectResult;
-use crate::v1::{
-    admin_result, object_result, AdminResult, MutateResult, ObjectResult, ResultHeader,
-    SelectResult as SelectResultRaw,
-};
-
-pub const PROTOCOL_VERSION: u32 = 1;
-
-pub type Success = u32;
-pub type Failure = u32;
-
-#[derive(Default)]
-pub struct ObjectResultBuilder {
-    version: u32,
-    code: u32,
-    err_msg: Option<String>,
-    result: Option<Body>,
-}
-
-pub enum Body {
-    Mutate((Success, Failure)),
-    Select(SelectResult),
-}
-
-impl ObjectResultBuilder {
-    pub fn new() -> Self {
-        Self {
-            version: PROTOCOL_VERSION,
-            ..Default::default()
-        }
-    }
-
-    #[allow(dead_code)]
-    pub fn version(mut self, version: u32) -> Self {
-        self.version = version;
-        self
-    }
-
-    pub fn status_code(mut self, code: u32) -> Self {
-        self.code = code;
-        self
-    }
-
-    pub fn err_msg(mut self, err_msg: String) -> Self {
-        self.err_msg = Some(err_msg);
-        self
-    }
-
-    pub fn mutate_result(mut self, success: u32, failure: u32) -> Self {
-        self.result = Some(Body::Mutate((success, failure)));
-        self
-    }
-
-    pub fn select_result(mut self, select_result: SelectResult) -> Self {
-        self.result = Some(Body::Select(select_result));
-        self
-    }
-
-    pub fn build(self) -> ObjectResult {
-        let header = Some(ResultHeader {
-            version: self.version,
-            code: self.code,
-            err_msg: self.err_msg.unwrap_or_default(),
-        });
-
-        let result = match self.result {
-            Some(Body::Mutate((success, failure))) => {
-                Some(object_result::Result::Mutate(MutateResult {
-                    success,
-                    failure,
-                }))
-            }
-            Some(Body::Select(select)) => Some(object_result::Result::Select(SelectResultRaw {
-                raw_data: select.into(),
-            })),
-            None => None,
-        };
-
-        ObjectResult { header, result }
-    }
-}
-
-pub fn build_err_result(err: &impl ErrorExt) -> ObjectResult {
-    ObjectResultBuilder::new()
-        .status_code(err.status_code() as u32)
-        .err_msg(err.to_string())
-        .build()
-}
-
-#[derive(Debug)]
-pub struct AdminResultBuilder {
-    version: u32,
-    code: u32,
-    err_msg: Option<String>,
-    mutate: Option<(Success, Failure)>,
-}
-
-impl AdminResultBuilder {
-    pub fn status_code(mut self, code: u32) -> Self {
-        self.code = code;
-        self
-    }
-
-    pub fn err_msg(mut self, err_msg: String) -> Self {
-        self.err_msg = Some(err_msg);
-        self
-    }
-
-    pub fn mutate_result(mut self, success: u32, failure: u32) -> Self {
-        self.mutate = Some((success, failure));
-        self
-    }
-
-    pub fn build(self) -> AdminResult {
-        let header = Some(ResultHeader {
-            version: self.version,
-            code: self.code,
-            err_msg: self.err_msg.unwrap_or_default(),
-        });
-
-        let result = if let Some((success, failure)) = self.mutate {
-            Some(admin_result::Result::Mutate(MutateResult {
-                success,
-                failure,
-            }))
-        } else {
-            None
-        };
-
-        AdminResult { header, result }
-    }
-}
-
-impl Default for AdminResultBuilder {
-    fn default() -> Self {
-        Self {
-            version: PROTOCOL_VERSION,
-            code: 0,
-            err_msg: None,
-            mutate: None,
-        }
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use common_error::status_code::StatusCode;
-
-    use super::*;
-    use crate::error::UnknownColumnDataTypeSnafu;
-    use crate::v1::{object_result, MutateResult};
-
-    #[test]
-    fn test_object_result_builder() {
-        let obj_result = ObjectResultBuilder::new()
-            .version(101)
-            .status_code(500)
-            .err_msg("Failed to read this file!".to_string())
-            .mutate_result(100, 20)
-            .build();
-        let header = obj_result.header.unwrap();
-        assert_eq!(101, header.version);
-        assert_eq!(500, header.code);
-        assert_eq!("Failed to read this file!", header.err_msg);
-
-        let result = obj_result.result.unwrap();
-        assert_eq!(
-            object_result::Result::Mutate(MutateResult {
-                success: 100,
-                failure: 20,
-            }),
-            result
-        );
-    }
-
-    #[test]
-    fn test_build_err_result() {
-        let err = UnknownColumnDataTypeSnafu { datatype: 1 }.build();
-        let err_result = build_err_result(&err);
-        let header = err_result.header.unwrap();
-        let result = err_result.result;
-
-        assert_eq!(PROTOCOL_VERSION, header.version);
-        assert_eq!(StatusCode::InvalidArguments as u32, header.code);
-        assert!(result.is_none());
-    }
-}

src/api/src/serde.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,7 +15,6 @@
 pub use prost::DecodeError;
 use prost::Message;
 
-use crate::v1::codec::{InsertBatch, PhysicalPlanNode, RegionNumber, SelectResult};
 use crate::v1::meta::TableRouteValue;
 
 macro_rules! impl_convert_with_bytes {
@@ -36,158 +35,4 @@ macro_rules! impl_convert_with_bytes {
     };
 }
 
-impl_convert_with_bytes!(InsertBatch);
-impl_convert_with_bytes!(SelectResult);
-impl_convert_with_bytes!(PhysicalPlanNode);
-impl_convert_with_bytes!(RegionNumber);
 impl_convert_with_bytes!(TableRouteValue);
-
-#[cfg(test)]
-mod tests {
-    use std::ops::Deref;
-
-    use crate::v1::codec::*;
-    use crate::v1::{column, Column};
-
-    const SEMANTIC_TAG: i32 = 0;
-
-    #[test]
-    fn test_convert_insert_batch() {
-        let insert_batch = mock_insert_batch();
-
-        let bytes: Vec<u8> = insert_batch.into();
-        let insert: InsertBatch = bytes.deref().try_into().unwrap();
-
-        assert_eq!(8, insert.row_count);
-        assert_eq!(1, insert.columns.len());
-
-        let column = &insert.columns[0];
-        assert_eq!("foo", column.column_name);
-        assert_eq!(SEMANTIC_TAG, column.semantic_type);
-        assert_eq!(vec![1], column.null_mask);
-        assert_eq!(
-            vec![2, 3, 4, 5, 6, 7, 8],
-            column.values.as_ref().unwrap().i32_values
-        );
-    }
-
-    #[should_panic]
-    #[test]
-    fn test_convert_insert_batch_wrong() {
-        let insert_batch = mock_insert_batch();
-
-        let mut bytes: Vec<u8> = insert_batch.into();
-
-        // modify some bytes
-        bytes[0] = 0b1;
-        bytes[1] = 0b1;
-
-        let insert: InsertBatch = bytes.deref().try_into().unwrap();
-
-        assert_eq!(8, insert.row_count);
-        assert_eq!(1, insert.columns.len());
-
-        let column = &insert.columns[0];
-        assert_eq!("foo", column.column_name);
-        assert_eq!(SEMANTIC_TAG, column.semantic_type);
-        assert_eq!(vec![1], column.null_mask);
-        assert_eq!(
-            vec![2, 3, 4, 5, 6, 7, 8],
-            column.values.as_ref().unwrap().i32_values
-        );
-    }
-
-    #[test]
-    fn test_convert_select_result() {
-        let select_result = mock_select_result();
-
-        let bytes: Vec<u8> = select_result.into();
-        let result: SelectResult = bytes.deref().try_into().unwrap();
-
-        assert_eq!(8, result.row_count);
-        assert_eq!(1, result.columns.len());
-
-        let column = &result.columns[0];
-        assert_eq!("foo", column.column_name);
-        assert_eq!(SEMANTIC_TAG, column.semantic_type);
-        assert_eq!(vec![1], column.null_mask);
-        assert_eq!(
-            vec![2, 3, 4, 5, 6, 7, 8],
-            column.values.as_ref().unwrap().i32_values
-        );
-    }
-
-    #[should_panic]
-    #[test]
-    fn test_convert_select_result_wrong() {
-        let select_result = mock_select_result();
-
-        let mut bytes: Vec<u8> = select_result.into();
-
-        // modify some bytes
-        bytes[0] = 0b1;
-        bytes[1] = 0b1;
-
-        let result: SelectResult = bytes.deref().try_into().unwrap();
-
-        assert_eq!(8, result.row_count);
-        assert_eq!(1, result.columns.len());
-
-        let column = &result.columns[0];
-        assert_eq!("foo", column.column_name);
-        assert_eq!(SEMANTIC_TAG, column.semantic_type);
-        assert_eq!(vec![1], column.null_mask);
-        assert_eq!(
-            vec![2, 3, 4, 5, 6, 7, 8],
-            column.values.as_ref().unwrap().i32_values
-        );
-    }
-
-    #[test]
-    fn test_convert_region_id() {
-        let region_id = RegionNumber { id: 12 };
-
-        let bytes: Vec<u8> = region_id.into();
-        let region_id: RegionNumber = bytes.deref().try_into().unwrap();
-
-        assert_eq!(12, region_id.id);
-    }
-
-    fn mock_insert_batch() -> InsertBatch {
-        let values = column::Values {
-            i32_values: vec![2, 3, 4, 5, 6, 7, 8],
-            ..Default::default()
-        };
-        let null_mask = vec![1];
-        let column = Column {
-            column_name: "foo".to_string(),
-            semantic_type: SEMANTIC_TAG,
-            values: Some(values),
-            null_mask,
-            ..Default::default()
-        };
-        InsertBatch {
-            columns: vec![column],
-            row_count: 8,
-        }
-    }
-
-    fn mock_select_result() -> SelectResult {
-        let values = column::Values {
-            i32_values: vec![2, 3, 4, 5, 6, 7, 8],
-            ..Default::default()
-        };
-        let null_mask = vec![1];
-        let column = Column {
-            column_name: "foo".to_string(),
-            semantic_type: SEMANTIC_TAG,
-            values: Some(values),
-            null_mask,
-            ..Default::default()
-        };
-        SelectResult {
-            columns: vec![column],
-            row_count: 8,
-        }
-    }
-}

src/api/src/v1.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,10 +15,5 @@
 #![allow(clippy::derive_partial_eq_without_eq)]
 tonic::include_proto!("greptime.v1");
 
-pub const GREPTIME_FD_SET: &[u8] = tonic::include_file_descriptor_set!("greptime_fd");
-
-pub mod codec {
-    tonic::include_proto!("greptime.v1.codec");
-}
-
+mod column_def;
 pub mod meta;

src/api/src/v1/column_def.rs

@@ -0,0 +1,39 @@
+// 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.
+
+use datatypes::schema::{ColumnDefaultConstraint, ColumnSchema};
+use snafu::ResultExt;
+
+use crate::error::{self, Result};
+use crate::helper::ColumnDataTypeWrapper;
+use crate::v1::ColumnDef;
+
+impl ColumnDef {
+    pub fn try_as_column_schema(&self) -> Result<ColumnSchema> {
+        let data_type = ColumnDataTypeWrapper::try_new(self.datatype)?;
+
+        let constraint = if self.default_constraint.is_empty() {
+            None
+        } else {
+            Some(
+                ColumnDefaultConstraint::try_from(self.default_constraint.as_slice())
+                    .context(error::ConvertColumnDefaultConstraintSnafu { column: &self.name })?,
+            )
+        };
+
+        ColumnSchema::new(&self.name, data_type.into(), self.is_nullable)
+            .with_default_constraint(constraint)
+            .context(error::InvalidColumnDefaultConstraintSnafu { column: &self.name })
+    }
+}

src/api/src/v1/meta.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -145,10 +145,12 @@ gen_set_header!(HeartbeatRequest);
 gen_set_header!(RouteRequest);
 gen_set_header!(CreateRequest);
 gen_set_header!(RangeRequest);
+gen_set_header!(DeleteRequest);
 gen_set_header!(PutRequest);
 gen_set_header!(BatchPutRequest);
 gen_set_header!(CompareAndPutRequest);
 gen_set_header!(DeleteRangeRequest);
+gen_set_header!(MoveValueRequest);
 
 #[cfg(test)]
 mod tests {

src/catalog/Cargo.toml

@@ -1,14 +1,13 @@
 [package]
 name = "catalog"
-version = "0.1.0"
-edition = "2021"
-license = "Apache-2.0"
-# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 
 [dependencies]
 api = { path = "../api" }
 arc-swap = "1.0"
-async-stream = "0.3"
+async-stream.workspace = true
 async-trait = "0.1"
 backoff = { version = "0.4", features = ["tokio"] }
 common-catalog = { path = "../common/catalog" }
@@ -19,29 +18,25 @@ common-recordbatch = { path = "../common/recordbatch" }
 common-runtime = { path = "../common/runtime" }
 common-telemetry = { path = "../common/telemetry" }
 common-time = { path = "../common/time" }
-datafusion = { git = "https://github.com/apache/arrow-datafusion.git", branch = "arrow2", features = [
-    "simd",
-] }
+datafusion.workspace = true
 datatypes = { path = "../datatypes" }
 futures = "0.3"
-futures-util = "0.3"
+futures-util.workspace = true
 lazy_static = "1.4"
 meta-client = { path = "../meta-client" }
-opendal = "0.17"
 regex = "1.6"
 serde = "1.0"
 serde_json = "1.0"
 snafu = { version = "0.7", features = ["backtraces"] }
 storage = { path = "../storage" }
 table = { path = "../table" }
-tokio = { version = "1.18", features = ["full"] }
+tokio.workspace = true
 
 [dev-dependencies]
 chrono = "0.4"
 log-store = { path = "../log-store" }
+mito = { path = "../mito", features = ["test"] }
 object-store = { path = "../object-store" }
-opendal = "0.17"
 storage = { path = "../storage" }
-table-engine = { path = "../table-engine" }
 tempdir = "0.3"
-tokio = { version = "1.0", features = ["full"] }
+tokio.workspace = true

src/catalog/src/error.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -17,7 +17,7 @@ use std::any::Any;
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::prelude::{Snafu, StatusCode};
 use datafusion::error::DataFusionError;
-use datatypes::arrow;
+use datatypes::prelude::ConcreteDataType;
 use datatypes::schema::RawSchema;
 use snafu::{Backtrace, ErrorCompat};
 
@@ -51,14 +51,12 @@ pub enum Error {
     SystemCatalog { msg: String, backtrace: Backtrace },
 
     #[snafu(display(
-        "System catalog table type mismatch, expected: binary, found: {:?} source: {}",
+        "System catalog table type mismatch, expected: binary, found: {:?}",
         data_type,
-        source
     ))]
     SystemCatalogTypeMismatch {
-        data_type: arrow::datatypes::DataType,
-        #[snafu(backtrace)]
-        source: datatypes::error::Error,
+        data_type: ConcreteDataType,
+        backtrace: Backtrace,
     },
 
     #[snafu(display("Invalid system catalog entry type: {:?}", entry_type))]
@@ -88,13 +86,14 @@ pub enum Error {
         backtrace: Backtrace,
     },
 
-    #[snafu(display("Cannot find schema, schema info: {}", schema_info))]
+    #[snafu(display("Cannot find schema {} in catalog {}", schema, catalog))]
     SchemaNotFound {
-        schema_info: String,
+        catalog: String,
+        schema: String,
         backtrace: Backtrace,
     },
 
-    #[snafu(display("Table {} already exists", table))]
+    #[snafu(display("Table `{}` already exists", table))]
     TableExists { table: String, backtrace: Backtrace },
 
     #[snafu(display("Schema {} already exists", schema))]
@@ -109,6 +108,12 @@ pub enum Error {
         source: BoxedError,
     },
 
+    #[snafu(display("Operation {} not implemented yet", operation))]
+    Unimplemented {
+        operation: String,
+        backtrace: Backtrace,
+    },
+
     #[snafu(display("Failed to open table, table info: {}, source: {}", table_info, source))]
     OpenTable {
         table_info: String,
@@ -159,6 +164,9 @@ pub enum Error {
         source: datatypes::error::Error,
     },
 
+    #[snafu(display("Failure during SchemaProvider operation, source: {}", source))]
+    SchemaProviderOperation { source: BoxedError },
+
     #[snafu(display("Failed to execute system catalog table scan, source: {}", source))]
     SystemCatalogTableScanExec {
         #[snafu(backtrace)]
@@ -185,8 +193,8 @@ pub enum Error {
         source: meta_client::error::Error,
     },
 
-    #[snafu(display("Invalid table schema in catalog, source: {:?}", source))]
-    InvalidSchemaInCatalog {
+    #[snafu(display("Invalid table info in catalog, source: {}", source))]
+    InvalidTableInfoInCatalog {
         #[snafu(backtrace)]
         source: datatypes::error::Error,
     },
@@ -216,11 +224,13 @@ impl ErrorExt for Error {
             | Error::ValueDeserialize { .. }
             | Error::Io { .. } => StatusCode::StorageUnavailable,
 
+            Error::RegisterTable { .. } | Error::SystemCatalogTypeMismatch { .. } => {
+                StatusCode::Internal
+            }
+
             Error::ReadSystemCatalog { source, .. } => source.status_code(),
-            Error::SystemCatalogTypeMismatch { source, .. } => source.status_code(),
             Error::InvalidCatalogValue { source, .. } => source.status_code(),
 
-            Error::RegisterTable { .. } => StatusCode::Internal,
             Error::TableExists { .. } => StatusCode::TableAlreadyExists,
             Error::SchemaExists { .. } => StatusCode::InvalidArguments,
 
@@ -233,8 +243,12 @@ impl ErrorExt for Error {
             Error::SystemCatalogTableScan { source } => source.status_code(),
             Error::SystemCatalogTableScanExec { source } => source.status_code(),
             Error::InvalidTableSchema { source, .. } => source.status_code(),
-            Error::InvalidSchemaInCatalog { .. } => StatusCode::Unexpected,
-            Error::Internal { source, .. } => source.status_code(),
+            Error::InvalidTableInfoInCatalog { .. } => StatusCode::Unexpected,
+            Error::Internal { source, .. } | Error::SchemaProviderOperation { source } => {
+                source.status_code()
+            }
+
+            Error::Unimplemented { .. } => StatusCode::Unsupported,
         }
     }
 
@@ -255,8 +269,6 @@ impl From<Error> for DataFusionError {
 
 #[cfg(test)]
 mod tests {
-    use common_error::mock::MockError;
-    use datatypes::arrow::datatypes::DataType;
     use snafu::GenerateImplicitData;
 
     use super::*;
@@ -277,22 +289,6 @@ mod tests {
             InvalidKeySnafu { key: None }.build().status_code()
         );
 
-        assert_eq!(
-            StatusCode::StorageUnavailable,
-            Error::OpenSystemCatalog {
-                source: table::error::Error::new(MockError::new(StatusCode::StorageUnavailable))
-            }
-            .status_code()
-        );
-
-        assert_eq!(
-            StatusCode::StorageUnavailable,
-            Error::CreateSystemCatalog {
-                source: table::error::Error::new(MockError::new(StatusCode::StorageUnavailable))
-            }
-            .status_code()
-        );
-
         assert_eq!(
             StatusCode::StorageUnavailable,
             Error::SystemCatalog {
@@ -305,11 +301,8 @@ mod tests {
         assert_eq!(
             StatusCode::Internal,
             Error::SystemCatalogTypeMismatch {
-                data_type: DataType::Boolean,
-                source: datatypes::error::Error::UnsupportedArrowType {
-                    arrow_type: DataType::Boolean,
-                    backtrace: Backtrace::generate()
-                }
+                data_type: ConcreteDataType::binary_datatype(),
+                backtrace: Backtrace::generate(),
             }
             .status_code()
         );

src/catalog/src/helper.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,54 +15,56 @@
 use std::collections::HashMap;
 use std::fmt::{Display, Formatter};
 
+use common_catalog::error::{
+    DeserializeCatalogEntryValueSnafu, Error, InvalidCatalogSnafu, SerializeCatalogEntryValueSnafu,
+};
 use lazy_static::lazy_static;
 use regex::Regex;
 use serde::{Deserialize, Serialize, Serializer};
 use snafu::{ensure, OptionExt, ResultExt};
-use table::metadata::{RawTableMeta, TableId, TableVersion};
+use table::metadata::{RawTableInfo, TableId, TableVersion};
 
-use crate::consts::{
-    CATALOG_KEY_PREFIX, SCHEMA_KEY_PREFIX, TABLE_GLOBAL_KEY_PREFIX, TABLE_REGIONAL_KEY_PREFIX,
-};
-use crate::error::{
-    DeserializeCatalogEntryValueSnafu, Error, InvalidCatalogSnafu, SerializeCatalogEntryValueSnafu,
-};
+const CATALOG_KEY_PREFIX: &str = "__c";
+const SCHEMA_KEY_PREFIX: &str = "__s";
+const TABLE_GLOBAL_KEY_PREFIX: &str = "__tg";
+const TABLE_REGIONAL_KEY_PREFIX: &str = "__tr";
+
+const ALPHANUMERICS_NAME_PATTERN: &str = "[a-zA-Z_][a-zA-Z0-9_]*";
 
 lazy_static! {
-    static ref CATALOG_KEY_PATTERN: Regex =
-        Regex::new(&format!("^{}-([a-zA-Z_]+)$", CATALOG_KEY_PREFIX)).unwrap();
+    static ref CATALOG_KEY_PATTERN: Regex = Regex::new(&format!(
+        "^{CATALOG_KEY_PREFIX}-({ALPHANUMERICS_NAME_PATTERN})$"
+    ))
+    .unwrap();
 }
 
 lazy_static! {
     static ref SCHEMA_KEY_PATTERN: Regex = Regex::new(&format!(
-        "^{}-([a-zA-Z_]+)-([a-zA-Z_]+)$",
-        SCHEMA_KEY_PREFIX
+        "^{SCHEMA_KEY_PREFIX}-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})$"
     ))
     .unwrap();
 }
 
 lazy_static! {
     static ref TABLE_GLOBAL_KEY_PATTERN: Regex = Regex::new(&format!(
-        "^{}-([a-zA-Z_]+)-([a-zA-Z_]+)-([a-zA-Z0-9_]+)$",
-        TABLE_GLOBAL_KEY_PREFIX
+        "^{TABLE_GLOBAL_KEY_PREFIX}-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})$"
     ))
     .unwrap();
 }
 
 lazy_static! {
     static ref TABLE_REGIONAL_KEY_PATTERN: Regex = Regex::new(&format!(
-        "^{}-([a-zA-Z_]+)-([a-zA-Z_]+)-([a-zA-Z0-9_]+)-([0-9]+)$",
-        TABLE_REGIONAL_KEY_PREFIX
+        "^{TABLE_REGIONAL_KEY_PREFIX}-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})-([0-9]+)$"
     ))
     .unwrap();
 }
 
 pub fn build_catalog_prefix() -> String {
-    format!("{}-", CATALOG_KEY_PREFIX)
+    format!("{CATALOG_KEY_PREFIX}-")
 }
 
 pub fn build_schema_prefix(catalog_name: impl AsRef<str>) -> String {
-    format!("{}-{}-", SCHEMA_KEY_PREFIX, catalog_name.as_ref())
+    format!("{SCHEMA_KEY_PREFIX}-{}-", catalog_name.as_ref())
 }
 
 pub fn build_table_global_prefix(
@@ -70,8 +72,7 @@ pub fn build_table_global_prefix(
     schema_name: impl AsRef<str>,
 ) -> String {
     format!(
-        "{}-{}-{}-",
-        TABLE_GLOBAL_KEY_PREFIX,
+        "{TABLE_GLOBAL_KEY_PREFIX}-{}-{}-",
         catalog_name.as_ref(),
         schema_name.as_ref()
     )
@@ -90,6 +91,7 @@ pub fn build_table_regional_prefix(
 }
 
 /// Table global info has only one key across all datanodes so it does not have `node_id` field.
+#[derive(Clone)]
 pub struct TableGlobalKey {
     pub catalog_name: String,
     pub schema_name: String,
@@ -126,17 +128,19 @@ impl TableGlobalKey {
 
 /// Table global info contains necessary info for a datanode to create table regions, including
 /// table id, table meta(schema...), region id allocation across datanodes.
-#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
 pub struct TableGlobalValue {
-    /// Table id is the same across all datanodes.
-    pub id: TableId,
     /// Id of datanode that created the global table info kv. only for debugging.
     pub node_id: u64,
-    // TODO(LFC): Maybe remove it?
     /// Allocation of region ids across all datanodes.
     pub regions_id_map: HashMap<u64, Vec<u32>>,
-    // TODO(LFC): Too much for assembling the table schema that DistTable needs, find another way.
-    pub meta: RawTableMeta,
+    pub table_info: RawTableInfo,
+}
+
+impl TableGlobalValue {
+    pub fn table_id(&self) -> TableId {
+        self.table_info.ident.table_id
+    }
 }
 
 /// Table regional info that varies between datanode, so it contains a `node_id` field.
@@ -258,6 +262,10 @@ macro_rules! define_catalog_value {
                             .context(DeserializeCatalogEntryValueSnafu { raw: s.as_ref() })
                     }
 
+                    pub fn from_bytes(bytes: impl AsRef<[u8]>) -> Result<Self, Error> {
+                         Self::parse(&String::from_utf8_lossy(bytes.as_ref()))
+                    }
+
                     pub fn as_bytes(&self) -> Result<Vec<u8>, Error> {
                         Ok(serde_json::to_string(self)
                             .context(SerializeCatalogEntryValueSnafu)?
@@ -279,6 +287,7 @@ define_catalog_value!(
 mod tests {
     use datatypes::prelude::ConcreteDataType;
     use datatypes::schema::{ColumnSchema, RawSchema, Schema};
+    use table::metadata::{RawTableMeta, TableIdent, TableType};
 
     use super::*;
 
@@ -339,14 +348,26 @@ mod tests {
             region_numbers: vec![1],
         };
 
+        let table_info = RawTableInfo {
+            ident: TableIdent {
+                table_id: 42,
+                version: 1,
+            },
+            name: "table_1".to_string(),
+            desc: Some("blah".to_string()),
+            catalog_name: "catalog_1".to_string(),
+            schema_name: "schema_1".to_string(),
+            meta,
+            table_type: TableType::Base,
+        };
+
         let value = TableGlobalValue {
-            id: 42,
             node_id: 0,
             regions_id_map: HashMap::from([(0, vec![1, 2, 3])]),
-            meta,
+            table_info,
         };
         let serialized = serde_json::to_string(&value).unwrap();
-        let deserialized = TableGlobalValue::parse(&serialized).unwrap();
+        let deserialized = TableGlobalValue::parse(serialized).unwrap();
         assert_eq!(value, deserialized);
     }
 }

src/catalog/src/lib.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -15,10 +15,11 @@
 #![feature(assert_matches)]
 
 use std::any::Any;
+use std::fmt::{Debug, Formatter};
 use std::sync::Arc;
 
 use common_telemetry::info;
-use snafu::ResultExt;
+use snafu::{OptionExt, ResultExt};
 use table::engine::{EngineContext, TableEngineRef};
 use table::metadata::TableId;
 use table::requests::CreateTableRequest;
@@ -28,6 +29,7 @@ use crate::error::{CreateTableSnafu, Result};
 pub use crate::schema::{SchemaProvider, SchemaProviderRef};
 
 pub mod error;
+pub mod helper;
 pub mod local;
 pub mod remote;
 pub mod schema;
@@ -83,12 +85,20 @@ pub trait CatalogManager: CatalogList {
     /// Starts a catalog manager.
     async fn start(&self) -> Result<()>;
 
-    /// Registers a table given given catalog/schema to catalog manager,
-    /// returns table registered.
-    async fn register_table(&self, request: RegisterTableRequest) -> Result<usize>;
+    /// Registers a table within given catalog/schema to catalog manager,
+    /// returns whether the table registered.
+    async fn register_table(&self, request: RegisterTableRequest) -> Result<bool>;
 
-    /// Register a schema with catalog name and schema name.
-    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<usize>;
+    /// Deregisters a table within given catalog/schema to catalog manager,
+    /// returns whether the table deregistered.
+    async fn deregister_table(&self, request: DeregisterTableRequest) -> Result<bool>;
+
+    /// Register a schema with catalog name and schema name. Retuens whether the
+    /// schema registered.
+    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<bool>;
+
+    /// Rename a table to [RenameTableRequest::new_table_name], returns whether the table is renamed.
+    async fn rename_table(&self, request: RenameTableRequest) -> Result<bool>;
 
     /// Register a system table, should be called before starting the manager.
     async fn register_system_table(&self, request: RegisterSystemTableRequest)
@@ -123,6 +133,34 @@ pub struct RegisterTableRequest {
     pub table: TableRef,
 }
 
+impl Debug for RegisterTableRequest {
+    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
+        f.debug_struct("RegisterTableRequest")
+            .field("catalog", &self.catalog)
+            .field("schema", &self.schema)
+            .field("table_name", &self.table_name)
+            .field("table_id", &self.table_id)
+            .field("table", &self.table.table_info())
+            .finish()
+    }
+}
+
+#[derive(Debug, Clone)]
+pub struct RenameTableRequest {
+    pub catalog: String,
+    pub schema: String,
+    pub table_name: String,
+    pub new_table_name: String,
+    pub table_id: TableId,
+}
+
+#[derive(Clone)]
+pub struct DeregisterTableRequest {
+    pub catalog: String,
+    pub schema: String,
+    pub table_name: String,
+}
+
 #[derive(Debug, Clone)]
 pub struct RegisterSchemaRequest {
     pub catalog: String,
@@ -131,7 +169,7 @@ pub struct RegisterSchemaRequest {
 
 /// Formats table fully-qualified name
 pub fn format_full_table_name(catalog: &str, schema: &str, table: &str) -> String {
-    format!("{}.{}.{}", catalog, schema, table)
+    format!("{catalog}.{schema}.{table}")
 }
 
 pub trait CatalogProviderFactory {
@@ -161,8 +199,7 @@ pub(crate) async fn handle_system_table_request<'a, M: CatalogManager>(
                 .await
                 .with_context(|_| CreateTableSnafu {
                     table_info: format!(
-                        "{}.{}.{}, id: {}",
-                        catalog_name, schema_name, table_name, table_id,
+                        "{catalog_name}.{schema_name}.{table_name}, id: {table_id}",
                     ),
                 })?;
             manager
@@ -174,7 +211,7 @@ pub(crate) async fn handle_system_table_request<'a, M: CatalogManager>(
                     table: table.clone(),
                 })
                 .await?;
-            info!("Created and registered system table: {}", table_name);
+            info!("Created and registered system table: {table_name}");
             table
         };
         if let Some(hook) = req.open_hook {
@@ -183,3 +220,38 @@ pub(crate) async fn handle_system_table_request<'a, M: CatalogManager>(
     }
     Ok(())
 }
+
+/// The number of regions in the datanode node.
+pub fn region_number(catalog_manager: &CatalogManagerRef) -> Result<u64> {
+    let mut region_number: u64 = 0;
+
+    for catalog_name in catalog_manager.catalog_names()? {
+        let catalog =
+            catalog_manager
+                .catalog(&catalog_name)?
+                .context(error::CatalogNotFoundSnafu {
+                    catalog_name: &catalog_name,
+                })?;
+
+        for schema_name in catalog.schema_names()? {
+            let schema = catalog
+                .schema(&schema_name)?
+                .context(error::SchemaNotFoundSnafu {
+                    catalog: &catalog_name,
+                    schema: &schema_name,
+                })?;
+
+            for table_name in schema.table_names()? {
+                let table = schema
+                    .table(&table_name)?
+                    .context(error::TableNotFoundSnafu {
+                        table_info: &table_name,
+                    })?;
+
+                let region_numbers = &table.table_info().meta.region_numbers;
+                region_number += region_numbers.len() as u64;
+            }
+        }
+    }
+    Ok(region_number)
+}

src/catalog/src/local.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/catalog/src/local/manager.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -21,7 +21,7 @@ use common_catalog::consts::{
     SYSTEM_CATALOG_NAME, SYSTEM_CATALOG_TABLE_NAME,
 };
 use common_recordbatch::{RecordBatch, SendableRecordBatchStream};
-use common_telemetry::info;
+use common_telemetry::{error, info};
 use datatypes::prelude::ScalarVector;
 use datatypes::vectors::{BinaryVector, UInt8Vector};
 use futures_util::lock::Mutex;
@@ -36,7 +36,7 @@ use table::TableRef;
 use crate::error::{
     CatalogNotFoundSnafu, IllegalManagerStateSnafu, OpenTableSnafu, ReadSystemCatalogSnafu, Result,
     SchemaExistsSnafu, SchemaNotFoundSnafu, SystemCatalogSnafu, SystemCatalogTypeMismatchSnafu,
-    TableExistsSnafu, TableNotFoundSnafu,
+    TableExistsSnafu, TableNotFoundSnafu, UnimplementedSnafu,
 };
 use crate::local::memory::{MemoryCatalogManager, MemoryCatalogProvider, MemorySchemaProvider};
 use crate::system::{
@@ -46,8 +46,9 @@ use crate::system::{
 use crate::tables::SystemCatalog;
 use crate::{
     format_full_table_name, handle_system_table_request, CatalogList, CatalogManager,
-    CatalogProvider, CatalogProviderRef, RegisterSchemaRequest, RegisterSystemTableRequest,
-    RegisterTableRequest, SchemaProvider, SchemaProviderRef,
+    CatalogProvider, CatalogProviderRef, DeregisterTableRequest, RegisterSchemaRequest,
+    RegisterSystemTableRequest, RegisterTableRequest, RenameTableRequest, SchemaProvider,
+    SchemaProviderRef,
 };
 
 /// A `CatalogManager` consists of a system catalog and a bunch of user catalogs.
@@ -57,6 +58,7 @@ pub struct LocalCatalogManager {
     engine: TableEngineRef,
     next_table_id: AtomicU32,
     init_lock: Mutex<bool>,
+    register_lock: Mutex<()>,
     system_table_requests: Mutex<Vec<RegisterSystemTableRequest>>,
 }
 
@@ -76,6 +78,7 @@ impl LocalCatalogManager {
             engine,
             next_table_id: AtomicU32::new(MIN_USER_TABLE_ID),
             init_lock: Mutex::new(false),
+            register_lock: Mutex::new(()),
             system_table_requests: Mutex::new(Vec::default()),
         })
     }
@@ -143,27 +146,34 @@ impl LocalCatalogManager {
     /// Convert `RecordBatch` to a vector of `Entry`.
     fn record_batch_to_entry(rb: RecordBatch) -> Result<Vec<Entry>> {
         ensure!(
-            rb.df_recordbatch.columns().len() >= 6,
+            rb.num_columns() >= 6,
             SystemCatalogSnafu {
-                msg: format!("Length mismatch: {}", rb.df_recordbatch.columns().len())
+                msg: format!("Length mismatch: {}", rb.num_columns())
             }
         );
 
-        let entry_type = UInt8Vector::try_from_arrow_array(&rb.df_recordbatch.columns()[0])
-            .with_context(|_| SystemCatalogTypeMismatchSnafu {
-                data_type: rb.df_recordbatch.columns()[ENTRY_TYPE_INDEX]
-                    .data_type()
-                    .clone(),
+        let entry_type = rb
+            .column(ENTRY_TYPE_INDEX)
+            .as_any()
+            .downcast_ref::<UInt8Vector>()
+            .with_context(|| SystemCatalogTypeMismatchSnafu {
+                data_type: rb.column(ENTRY_TYPE_INDEX).data_type(),
             })?;
 
-        let key = BinaryVector::try_from_arrow_array(&rb.df_recordbatch.columns()[1])
-            .with_context(|_| SystemCatalogTypeMismatchSnafu {
-                data_type: rb.df_recordbatch.columns()[KEY_INDEX].data_type().clone(),
+        let key = rb
+            .column(KEY_INDEX)
+            .as_any()
+            .downcast_ref::<BinaryVector>()
+            .with_context(|| SystemCatalogTypeMismatchSnafu {
+                data_type: rb.column(KEY_INDEX).data_type(),
             })?;
 
-        let value = BinaryVector::try_from_arrow_array(&rb.df_recordbatch.columns()[3])
-            .with_context(|_| SystemCatalogTypeMismatchSnafu {
-                data_type: rb.df_recordbatch.columns()[VALUE_INDEX].data_type().clone(),
+        let value = rb
+            .column(VALUE_INDEX)
+            .as_any()
+            .downcast_ref::<BinaryVector>()
+            .with_context(|| SystemCatalogTypeMismatchSnafu {
+                data_type: rb.column(VALUE_INDEX).data_type(),
             })?;
 
         let mut res = Vec::with_capacity(rb.num_rows());
@@ -232,7 +242,8 @@ impl LocalCatalogManager {
         let schema = catalog
             .schema(&t.schema_name)?
             .context(SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", &t.catalog_name, &t.schema_name),
+                catalog: &t.catalog_name,
+                schema: &t.schema_name,
             })?;
 
         let context = EngineContext {};
@@ -241,6 +252,7 @@ impl LocalCatalogManager {
             schema_name: t.schema_name.clone(),
             table_name: t.table_name.clone(),
             table_id: t.table_id,
+            region_numbers: vec![0],
         };
 
         let option = self
@@ -308,7 +320,7 @@ impl CatalogManager for LocalCatalogManager {
         self.init().await
     }
 
-    async fn register_table(&self, request: RegisterTableRequest) -> Result<usize> {
+    async fn register_table(&self, request: RegisterTableRequest) -> Result<bool> {
         let started = self.init_lock.lock().await;
 
         ensure!(
@@ -328,30 +340,93 @@ impl CatalogManager for LocalCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                catalog: catalog_name,
+                schema: schema_name,
             })?;
 
-        if schema.table_exist(&request.table_name)? {
-            return TableExistsSnafu {
-                table: format_full_table_name(catalog_name, schema_name, &request.table_name),
+        {
+            let _lock = self.register_lock.lock().await;
+            if let Some(existing) = schema.table(&request.table_name)? {
+                if existing.table_info().ident.table_id != request.table_id {
+                    error!(
+                        "Unexpected table register request: {:?}, existing: {:?}",
+                        request,
+                        existing.table_info()
+                    );
+                    return TableExistsSnafu {
+                        table: format_full_table_name(
+                            catalog_name,
+                            schema_name,
+                            &request.table_name,
+                        ),
+                    }
+                    .fail();
+                }
+                // Try to register table with same table id, just ignore.
+                Ok(false)
+            } else {
+                // table does not exist
+                self.system
+                    .register_table(
+                        catalog_name.clone(),
+                        schema_name.clone(),
+                        request.table_name.clone(),
+                        request.table_id,
+                    )
+                    .await?;
+                schema.register_table(request.table_name, request.table)?;
+                Ok(true)
             }
-            .fail();
         }
+    }
 
+    async fn rename_table(&self, request: RenameTableRequest) -> Result<bool> {
+        let started = self.init_lock.lock().await;
+
+        ensure!(
+            *started,
+            IllegalManagerStateSnafu {
+                msg: "Catalog manager not started",
+            }
+        );
+
+        let catalog_name = &request.catalog;
+        let schema_name = &request.schema;
+
+        let catalog = self
+            .catalogs
+            .catalog(catalog_name)?
+            .context(CatalogNotFoundSnafu { catalog_name })?;
+
+        let schema = catalog
+            .schema(schema_name)?
+            .with_context(|| SchemaNotFoundSnafu {
+                catalog: catalog_name,
+                schema: schema_name,
+            })?;
+
+        // rename table in system catalog
         self.system
             .register_table(
                 catalog_name.clone(),
                 schema_name.clone(),
-                request.table_name.clone(),
+                request.new_table_name.clone(),
                 request.table_id,
             )
             .await?;
-
-        schema.register_table(request.table_name, request.table)?;
-        Ok(1)
+        Ok(schema
+            .rename_table(&request.table_name, request.new_table_name)
+            .is_ok())
     }
 
-    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<usize> {
+    async fn deregister_table(&self, _request: DeregisterTableRequest) -> Result<bool> {
+        UnimplementedSnafu {
+            operation: "deregister table",
+        }
+        .fail()
+    }
+
+    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<bool> {
         let started = self.init_lock.lock().await;
         ensure!(
             *started,
@@ -366,17 +441,21 @@ impl CatalogManager for LocalCatalogManager {
             .catalogs
             .catalog(catalog_name)?
             .context(CatalogNotFoundSnafu { catalog_name })?;
-        if catalog.schema(schema_name)?.is_some() {
-            return SchemaExistsSnafu {
-                schema: schema_name,
-            }
-            .fail();
+
+        {
+            let _lock = self.register_lock.lock().await;
+            ensure!(
+                catalog.schema(schema_name)?.is_none(),
+                SchemaExistsSnafu {
+                    schema: schema_name,
+                }
+            );
+            self.system
+                .register_schema(request.catalog, schema_name.clone())
+                .await?;
+            catalog.register_schema(request.schema, Arc::new(MemorySchemaProvider::new()))?;
+            Ok(true)
         }
-        self.system
-            .register_schema(request.catalog, schema_name.clone())
-            .await?;
-        catalog.register_schema(request.schema, Arc::new(MemorySchemaProvider::new()))?;
-        Ok(1)
     }
 
     async fn register_system_table(&self, request: RegisterSystemTableRequest) -> Result<()> {
@@ -415,7 +494,8 @@ impl CatalogManager for LocalCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                catalog: catalog_name,
+                schema: schema_name,
             })?;
         schema.table(table_name)
     }

src/catalog/src/local/memory.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -19,16 +19,20 @@ use std::sync::atomic::{AtomicU32, Ordering};
 use std::sync::{Arc, RwLock};
 
 use common_catalog::consts::MIN_USER_TABLE_ID;
+use common_telemetry::error;
 use snafu::OptionExt;
 use table::metadata::TableId;
 use table::table::TableIdProvider;
 use table::TableRef;
 
-use crate::error::{CatalogNotFoundSnafu, Result, SchemaNotFoundSnafu, TableExistsSnafu};
+use crate::error::{
+    CatalogNotFoundSnafu, Result, SchemaNotFoundSnafu, TableExistsSnafu, TableNotFoundSnafu,
+};
 use crate::schema::SchemaProvider;
 use crate::{
-    CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef, RegisterSchemaRequest,
-    RegisterSystemTableRequest, RegisterTableRequest, SchemaProviderRef,
+    CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef, DeregisterTableRequest,
+    RegisterSchemaRequest, RegisterSystemTableRequest, RegisterTableRequest, RenameTableRequest,
+    SchemaProviderRef,
 };
 
 /// Simple in-memory list of catalogs
@@ -69,7 +73,7 @@ impl CatalogManager for MemoryCatalogManager {
         Ok(())
     }
 
-    async fn register_table(&self, request: RegisterTableRequest) -> Result<usize> {
+    async fn register_table(&self, request: RegisterTableRequest) -> Result<bool> {
         let catalogs = self.catalogs.write().unwrap();
         let catalog = catalogs
             .get(&request.catalog)
@@ -80,14 +84,53 @@ impl CatalogManager for MemoryCatalogManager {
         let schema = catalog
             .schema(&request.schema)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", &request.catalog, &request.schema),
+                catalog: &request.catalog,
+                schema: &request.schema,
             })?;
         schema
             .register_table(request.table_name, request.table)
-            .map(|v| if v.is_some() { 0 } else { 1 })
+            .map(|v| v.is_none())
     }
 
-    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<usize> {
+    async fn rename_table(&self, request: RenameTableRequest) -> Result<bool> {
+        let catalogs = self.catalogs.write().unwrap();
+        let catalog = catalogs
+            .get(&request.catalog)
+            .context(CatalogNotFoundSnafu {
+                catalog_name: &request.catalog,
+            })?
+            .clone();
+        let schema = catalog
+            .schema(&request.schema)?
+            .with_context(|| SchemaNotFoundSnafu {
+                catalog: &request.catalog,
+                schema: &request.schema,
+            })?;
+        Ok(schema
+            .rename_table(&request.table_name, request.new_table_name)
+            .is_ok())
+    }
+
+    async fn deregister_table(&self, request: DeregisterTableRequest) -> Result<bool> {
+        let catalogs = self.catalogs.write().unwrap();
+        let catalog = catalogs
+            .get(&request.catalog)
+            .context(CatalogNotFoundSnafu {
+                catalog_name: &request.catalog,
+            })?
+            .clone();
+        let schema = catalog
+            .schema(&request.schema)?
+            .with_context(|| SchemaNotFoundSnafu {
+                catalog: &request.catalog,
+                schema: &request.schema,
+            })?;
+        schema
+            .deregister_table(&request.table_name)
+            .map(|v| v.is_some())
+    }
+
+    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<bool> {
         let catalogs = self.catalogs.write().unwrap();
         let catalog = catalogs
             .get(&request.catalog)
@@ -95,11 +138,12 @@ impl CatalogManager for MemoryCatalogManager {
                 catalog_name: &request.catalog,
             })?;
         catalog.register_schema(request.schema, Arc::new(MemorySchemaProvider::new()))?;
-        Ok(1)
+        Ok(true)
     }
 
     async fn register_system_table(&self, _request: RegisterSystemTableRequest) -> Result<()> {
-        unimplemented!()
+        // TODO(ruihang): support register system table request
+        Ok(())
     }
 
     fn schema(&self, catalog: &str, schema: &str) -> Result<Option<SchemaProviderRef>> {
@@ -251,11 +295,35 @@ impl SchemaProvider for MemorySchemaProvider {
     }
 
     fn register_table(&self, name: String, table: TableRef) -> Result<Option<TableRef>> {
-        if self.table_exist(name.as_str())? {
-            return TableExistsSnafu { table: name }.fail()?;
-        }
         let mut tables = self.tables.write().unwrap();
-        Ok(tables.insert(name, table))
+        if let Some(existing) = tables.get(name.as_str()) {
+            // if table with the same name but different table id exists, then it's a fatal bug
+            if existing.table_info().ident.table_id != table.table_info().ident.table_id {
+                error!(
+                    "Unexpected table register: {:?}, existing: {:?}",
+                    table.table_info(),
+                    existing.table_info()
+                );
+                return TableExistsSnafu { table: name }.fail()?;
+            }
+            Ok(Some(existing.clone()))
+        } else {
+            Ok(tables.insert(name, table))
+        }
+    }
+
+    fn rename_table(&self, name: &str, new_name: String) -> Result<TableRef> {
+        let mut tables = self.tables.write().unwrap();
+        if tables.get(name).is_some() {
+            let table = tables.remove(name).unwrap();
+            tables.insert(new_name, table.clone());
+            Ok(table)
+        } else {
+            TableNotFoundSnafu {
+                table_info: name.to_string(),
+            }
+            .fail()?
+        }
     }
 
     fn deregister_table(&self, name: &str) -> Result<Option<TableRef>> {
@@ -315,13 +383,92 @@ mod tests {
             .unwrap()
             .is_none());
         assert!(provider.table_exist(table_name).unwrap());
-        let other_table = NumbersTable::default();
+        let other_table = NumbersTable::new(12);
         let result = provider.register_table(table_name.to_string(), Arc::new(other_table));
         let err = result.err().unwrap();
         assert!(err.backtrace_opt().is_some());
         assert_eq!(StatusCode::TableAlreadyExists, err.status_code());
     }
 
+    #[tokio::test]
+    async fn test_mem_provider_rename_table() {
+        let provider = MemorySchemaProvider::new();
+        let table_name = "num";
+        assert!(!provider.table_exist(table_name).unwrap());
+        let test_table: TableRef = Arc::new(NumbersTable::default());
+        // register test table
+        assert!(provider
+            .register_table(table_name.to_string(), test_table.clone())
+            .unwrap()
+            .is_none());
+        assert!(provider.table_exist(table_name).unwrap());
+
+        // rename test table
+        let new_table_name = "numbers";
+        provider
+            .rename_table(table_name, new_table_name.to_string())
+            .unwrap();
+
+        // test old table name not exist
+        assert!(!provider.table_exist(table_name).unwrap());
+        assert!(provider.deregister_table(table_name).unwrap().is_none());
+
+        // test new table name exists
+        assert!(provider.table_exist(new_table_name).unwrap());
+        let registered_table = provider.table(new_table_name).unwrap().unwrap();
+        assert_eq!(
+            registered_table.table_info().ident.table_id,
+            test_table.table_info().ident.table_id
+        );
+
+        let other_table = Arc::new(NumbersTable::new(2));
+        let result = provider.register_table(new_table_name.to_string(), other_table);
+        let err = result.err().unwrap();
+        assert_eq!(StatusCode::TableAlreadyExists, err.status_code());
+    }
+
+    #[tokio::test]
+    async fn test_catalog_rename_table() {
+        let catalog = MemoryCatalogManager::default();
+        let schema = catalog
+            .schema(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME)
+            .unwrap()
+            .unwrap();
+
+        // register table
+        let table_name = "num";
+        let table_id = 2333;
+        let table: TableRef = Arc::new(NumbersTable::new(table_id));
+        let register_table_req = RegisterTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: table_name.to_string(),
+            table_id,
+            table,
+        };
+        assert!(catalog.register_table(register_table_req).await.unwrap());
+        assert!(schema.table_exist(table_name).unwrap());
+
+        // rename table
+        let new_table_name = "numbers";
+        let rename_table_req = RenameTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: table_name.to_string(),
+            new_table_name: new_table_name.to_string(),
+            table_id,
+        };
+        assert!(catalog.rename_table(rename_table_req).await.unwrap());
+        assert!(!schema.table_exist(table_name).unwrap());
+        assert!(schema.table_exist(new_table_name).unwrap());
+
+        let registered_table = catalog
+            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, new_table_name)
+            .unwrap()
+            .unwrap();
+        assert_eq!(registered_table.table_info().ident.table_id, table_id);
+    }
+
     #[test]
     pub fn test_register_if_absent() {
         let list = MemoryCatalogManager::default();
@@ -340,4 +487,34 @@ mod tests {
             .downcast_ref::<MemoryCatalogManager>()
             .unwrap();
     }
+
+    #[tokio::test]
+    pub async fn test_catalog_deregister_table() {
+        let catalog = MemoryCatalogManager::default();
+        let schema = catalog
+            .schema(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME)
+            .unwrap()
+            .unwrap();
+
+        let register_table_req = RegisterTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: "numbers".to_string(),
+            table_id: 2333,
+            table: Arc::new(NumbersTable::default()),
+        };
+        catalog.register_table(register_table_req).await.unwrap();
+        assert!(schema.table_exist("numbers").unwrap());
+
+        let deregister_table_req = DeregisterTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: "numbers".to_string(),
+        };
+        catalog
+            .deregister_table(deregister_table_req)
+            .await
+            .unwrap();
+        assert!(!schema.table_exist("numbers").unwrap());
+    }
 }

src/catalog/src/remote.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/catalog/src/remote/client.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/catalog/src/remote/manager.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -20,10 +20,6 @@ use std::sync::Arc;
 use arc_swap::ArcSwap;
 use async_stream::stream;
 use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, MIN_USER_TABLE_ID};
-use common_catalog::{
-    build_catalog_prefix, build_schema_prefix, build_table_global_prefix, CatalogKey, CatalogValue,
-    SchemaKey, SchemaValue, TableGlobalKey, TableGlobalValue, TableRegionalKey, TableRegionalValue,
-};
 use common_telemetry::{debug, info};
 use futures::Stream;
 use futures_util::StreamExt;
@@ -37,13 +33,17 @@ use tokio::sync::Mutex;
 
 use crate::error::{
     CatalogNotFoundSnafu, CreateTableSnafu, InvalidCatalogValueSnafu, InvalidTableSchemaSnafu,
-    OpenTableSnafu, Result, SchemaNotFoundSnafu, TableExistsSnafu,
+    OpenTableSnafu, Result, SchemaNotFoundSnafu, TableExistsSnafu, UnimplementedSnafu,
+};
+use crate::helper::{
+    build_catalog_prefix, build_schema_prefix, build_table_global_prefix, CatalogKey, CatalogValue,
+    SchemaKey, SchemaValue, TableGlobalKey, TableGlobalValue, TableRegionalKey, TableRegionalValue,
 };
 use crate::remote::{Kv, KvBackendRef};
 use crate::{
     handle_system_table_request, CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef,
-    RegisterSchemaRequest, RegisterSystemTableRequest, RegisterTableRequest, SchemaProvider,
-    SchemaProviderRef,
+    DeregisterTableRequest, RegisterSchemaRequest, RegisterSystemTableRequest,
+    RegisterTableRequest, RenameTableRequest, SchemaProvider, SchemaProviderRef,
 };
 
 /// Catalog manager based on metasrv.
@@ -154,8 +154,8 @@ impl RemoteCatalogManager {
                 }
                 let table_key = TableGlobalKey::parse(&String::from_utf8_lossy(&k))
                     .context(InvalidCatalogValueSnafu)?;
-                let table_value = TableGlobalValue::parse(&String::from_utf8_lossy(&v))
-                    .context(InvalidCatalogValueSnafu)?;
+                let table_value =
+                    TableGlobalValue::from_bytes(&v).context(InvalidCatalogValueSnafu)?;
 
                 info!(
                     "Found catalog table entry, key: {}, value: {:?}",
@@ -250,10 +250,7 @@ impl RemoteCatalogManager {
             let table_ref = self.open_or_create_table(&table_key, &table_value).await?;
             schema.register_table(table_key.table_name.to_string(), table_ref)?;
             info!("Registered table {}", &table_key.table_name);
-            if table_value.id > max_table_id {
-                info!("Max table id: {} -> {}", max_table_id, table_value.id);
-                max_table_id = table_value.id;
-            }
+            max_table_id = max_table_id.max(table_value.table_id());
             table_num += 1;
         }
         info!(
@@ -311,25 +308,30 @@ impl RemoteCatalogManager {
             ..
         } = table_key;
 
+        let table_id = table_value.table_id();
+
         let TableGlobalValue {
-            id,
-            meta,
+            table_info,
             regions_id_map,
             ..
         } = table_value;
 
+        // unwrap safety: checked in yielding this table when `iter_remote_tables`
+        let region_numbers = regions_id_map.get(&self.node_id).unwrap();
+
         let request = OpenTableRequest {
             catalog_name: catalog_name.clone(),
             schema_name: schema_name.clone(),
             table_name: table_name.clone(),
-            table_id: *id,
+            table_id,
+            region_numbers: region_numbers.clone(),
         };
         match self
             .engine
             .open_table(&context, request)
             .await
             .with_context(|_| OpenTableSnafu {
-                table_info: format!("{}.{}.{}, id:{}", catalog_name, schema_name, table_name, id,),
+                table_info: format!("{catalog_name}.{schema_name}.{table_name}, id:{table_id}"),
             })? {
             Some(table) => {
                 info!(
@@ -344,22 +346,23 @@ impl RemoteCatalogManager {
                     catalog_name, schema_name, table_name
                 );
 
+                let meta = &table_info.meta;
                 let schema = meta
                     .schema
                     .clone()
                     .try_into()
                     .context(InvalidTableSchemaSnafu {
-                        table_info: format!("{}.{}.{}", catalog_name, schema_name, table_name,),
+                        table_info: format!("{catalog_name}.{schema_name}.{table_name}"),
                         schema: meta.schema.clone(),
                     })?;
                 let req = CreateTableRequest {
-                    id: *id,
+                    id: table_id,
                     catalog_name: catalog_name.clone(),
                     schema_name: schema_name.clone(),
                     table_name: table_name.clone(),
                     desc: None,
                     schema: Arc::new(schema),
-                    region_numbers: regions_id_map.get(&self.node_id).unwrap().clone(), // this unwrap is safe because region_id_map is checked in `iter_remote_tables`
+                    region_numbers: region_numbers.clone(),
                     primary_key_indices: meta.primary_key_indices.clone(),
                     create_if_not_exists: true,
                     table_options: meta.options.clone(),
@@ -371,7 +374,7 @@ impl RemoteCatalogManager {
                     .context(CreateTableSnafu {
                         table_info: format!(
                             "{}.{}.{}, id:{}",
-                            &catalog_name, &schema_name, &table_name, id
+                            &catalog_name, &schema_name, &table_name, table_id
                         ),
                     })
             }
@@ -405,7 +408,7 @@ impl CatalogManager for RemoteCatalogManager {
         Ok(())
     }
 
-    async fn register_table(&self, request: RegisterTableRequest) -> Result<usize> {
+    async fn register_table(&self, request: RegisterTableRequest) -> Result<bool> {
         let catalog_name = request.catalog;
         let schema_name = request.schema;
         let catalog_provider = self.catalog(&catalog_name)?.context(CatalogNotFoundSnafu {
@@ -415,7 +418,8 @@ impl CatalogManager for RemoteCatalogManager {
             catalog_provider
                 .schema(&schema_name)?
                 .with_context(|| SchemaNotFoundSnafu {
-                    schema_info: format!("{}.{}", &catalog_name, &schema_name),
+                    catalog: &catalog_name,
+                    schema: &schema_name,
                 })?;
         if schema_provider.table_exist(&request.table_name)? {
             return TableExistsSnafu {
@@ -424,10 +428,17 @@ impl CatalogManager for RemoteCatalogManager {
             .fail();
         }
         schema_provider.register_table(request.table_name, request.table)?;
-        Ok(1)
+        Ok(true)
     }
 
-    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<usize> {
+    async fn deregister_table(&self, _request: DeregisterTableRequest) -> Result<bool> {
+        UnimplementedSnafu {
+            operation: "deregister table",
+        }
+        .fail()
+    }
+
+    async fn register_schema(&self, request: RegisterSchemaRequest) -> Result<bool> {
         let catalog_name = request.catalog;
         let schema_name = request.schema;
         let catalog_provider = self.catalog(&catalog_name)?.context(CatalogNotFoundSnafu {
@@ -435,7 +446,14 @@ impl CatalogManager for RemoteCatalogManager {
         })?;
         let schema_provider = self.new_schema_provider(&catalog_name, &schema_name);
         catalog_provider.register_schema(schema_name, schema_provider)?;
-        Ok(1)
+        Ok(true)
+    }
+
+    async fn rename_table(&self, _request: RenameTableRequest) -> Result<bool> {
+        UnimplementedSnafu {
+            operation: "rename table",
+        }
+        .fail()
     }
 
     async fn register_system_table(&self, request: RegisterSystemTableRequest) -> Result<()> {
@@ -464,7 +482,8 @@ impl CatalogManager for RemoteCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                catalog: catalog_name,
+                schema: schema_name,
             })?;
         schema.table(table_name)
     }
@@ -727,6 +746,13 @@ impl SchemaProvider for RemoteSchemaProvider {
         prev
     }
 
+    fn rename_table(&self, _name: &str, _new_name: String) -> Result<TableRef> {
+        UnimplementedSnafu {
+            operation: "rename table",
+        }
+        .fail()
+    }
+
     fn deregister_table(&self, name: &str) -> Result<Option<TableRef>> {
         let table_name = name.to_string();
         let table_key = self.build_regional_table_key(&table_name).to_string();

src/catalog/src/schema.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -35,6 +35,10 @@ pub trait SchemaProvider: Sync + Send {
     /// If a table of the same name existed before, it returns "Table already exists" error.
     fn register_table(&self, name: String, table: TableRef) -> Result<Option<TableRef>>;
 
+    /// If supported by the implementation, renames an existing table from this schema and returns it.
+    /// If no table of that name exists, returns "Table not found" error.
+    fn rename_table(&self, name: &str, new_name: String) -> Result<TableRef>;
+
     /// If supported by the implementation, removes an existing table from this schema and returns it.
     /// If no table of that name exists, returns Ok(None).
     fn deregister_table(&self, name: &str) -> Result<Option<TableRef>>;

src/catalog/src/system.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -21,14 +21,13 @@ use common_catalog::consts::{
     SYSTEM_CATALOG_TABLE_ID, SYSTEM_CATALOG_TABLE_NAME,
 };
 use common_query::logical_plan::Expr;
-use common_query::physical_plan::{PhysicalPlanRef, RuntimeEnv};
+use common_query::physical_plan::{PhysicalPlanRef, SessionContext};
 use common_recordbatch::SendableRecordBatchStream;
 use common_telemetry::debug;
-use common_time::timestamp::Timestamp;
 use common_time::util;
 use datatypes::prelude::{ConcreteDataType, ScalarVector};
 use datatypes::schema::{ColumnSchema, Schema, SchemaBuilder, SchemaRef};
-use datatypes::vectors::{BinaryVector, TimestampVector, UInt8Vector};
+use datatypes::vectors::{BinaryVector, TimestampMillisecondVector, UInt8Vector};
 use serde::{Deserialize, Serialize};
 use snafu::{ensure, OptionExt, ResultExt};
 use table::engine::{EngineContext, TableEngineRef};
@@ -43,7 +42,6 @@ use crate::error::{
 
 pub const ENTRY_TYPE_INDEX: usize = 0;
 pub const KEY_INDEX: usize = 1;
-pub const TIMESTAMP_INDEX: usize = 2;
 pub const VALUE_INDEX: usize = 3;
 
 pub struct SystemCatalogTable {
@@ -63,7 +61,7 @@ impl Table for SystemCatalogTable {
 
     async fn scan(
         &self,
-        _projection: &Option<Vec<usize>>,
+        _projection: Option<&Vec<usize>>,
         _filters: &[Expr],
         _limit: Option<usize>,
     ) -> table::Result<PhysicalPlanRef> {
@@ -87,6 +85,7 @@ impl SystemCatalogTable {
             schema_name: INFORMATION_SCHEMA_NAME.to_string(),
             table_name: SYSTEM_CATALOG_TABLE_NAME.to_string(),
             table_id: SYSTEM_CATALOG_TABLE_ID,
+            region_numbers: vec![0],
         };
         let schema = Arc::new(build_system_catalog_schema());
         let ctx = EngineContext::default();
@@ -110,7 +109,7 @@ impl SystemCatalogTable {
                 desc: Some("System catalog table".to_string()),
                 schema: schema.clone(),
                 region_numbers: vec![0],
-                primary_key_indices: vec![ENTRY_TYPE_INDEX, KEY_INDEX, TIMESTAMP_INDEX],
+                primary_key_indices: vec![ENTRY_TYPE_INDEX, KEY_INDEX],
                 create_if_not_exists: true,
                 table_options: HashMap::new(),
             };
@@ -127,14 +126,14 @@ impl SystemCatalogTable {
     /// Create a stream of all entries inside system catalog table
     pub async fn records(&self) -> Result<SendableRecordBatchStream> {
         let full_projection = None;
+        let ctx = SessionContext::new();
         let scan = self
             .table
-            .scan(&full_projection, &[], None)
+            .scan(full_projection, &[], None)
             .await
             .context(error::SystemCatalogTableScanSnafu)?;
         let stream = scan
-            .execute(0, Arc::new(RuntimeEnv::default()))
-            .await
+            .execute(0, ctx.task_ctx())
             .context(error::SystemCatalogTableScanExecSnafu)?;
         Ok(stream)
     }
@@ -162,7 +161,7 @@ fn build_system_catalog_schema() -> Schema {
         ),
         ColumnSchema::new(
             "timestamp".to_string(),
-            ConcreteDataType::timestamp_millis_datatype(),
+            ConcreteDataType::timestamp_millisecond_datatype(),
             false,
         )
         .with_time_index(true),
@@ -173,12 +172,12 @@ fn build_system_catalog_schema() -> Schema {
         ),
         ColumnSchema::new(
             "gmt_created".to_string(),
-            ConcreteDataType::timestamp_millis_datatype(),
+            ConcreteDataType::timestamp_millisecond_datatype(),
             false,
         ),
         ColumnSchema::new(
             "gmt_modified".to_string(),
-            ConcreteDataType::timestamp_millis_datatype(),
+            ConcreteDataType::timestamp_millisecond_datatype(),
             false,
         ),
     ];
@@ -187,18 +186,30 @@ fn build_system_catalog_schema() -> Schema {
     SchemaBuilder::try_from(cols).unwrap().build().unwrap()
 }
 
-pub fn build_table_insert_request(full_table_name: String, table_id: TableId) -> InsertRequest {
+/// Formats key string for table entry in system catalog
+#[inline]
+pub fn format_table_entry_key(catalog: &str, schema: &str, table_id: TableId) -> String {
+    format!("{catalog}.{schema}.{table_id}")
+}
+
+pub fn build_table_insert_request(
+    catalog: String,
+    schema: String,
+    table_name: String,
+    table_id: TableId,
+) -> InsertRequest {
+    let entry_key = format_table_entry_key(&catalog, &schema, table_id);
     build_insert_request(
         EntryType::Table,
-        full_table_name.as_bytes(),
-        serde_json::to_string(&TableEntryValue { table_id })
+        entry_key.as_bytes(),
+        serde_json::to_string(&TableEntryValue { table_name })
             .unwrap()
             .as_bytes(),
     )
 }
 
 pub fn build_schema_insert_request(catalog_name: String, schema_name: String) -> InsertRequest {
-    let full_schema_name = format!("{}.{}", catalog_name, schema_name);
+    let full_schema_name = format!("{catalog_name}.{schema_name}");
     build_insert_request(
         EntryType::Schema,
         full_schema_name.as_bytes(),
@@ -223,7 +234,7 @@ pub fn build_insert_request(entry_type: EntryType, key: &[u8], value: &[u8]) ->
     // Timestamp in key part is intentionally left to 0
     columns_values.insert(
         "timestamp".to_string(),
-        Arc::new(TimestampVector::from_slice(&[Timestamp::from_millis(0)])) as _,
+        Arc::new(TimestampMillisecondVector::from_slice(&[0])) as _,
     );
 
     columns_values.insert(
@@ -231,18 +242,15 @@ pub fn build_insert_request(entry_type: EntryType, key: &[u8], value: &[u8]) ->
         Arc::new(BinaryVector::from_slice(&[value])) as _,
     );
 
+    let now = util::current_time_millis();
     columns_values.insert(
         "gmt_created".to_string(),
-        Arc::new(TimestampVector::from_slice(&[Timestamp::from_millis(
-            util::current_time_millis(),
-        )])) as _,
+        Arc::new(TimestampMillisecondVector::from_slice(&[now])) as _,
     );
 
     columns_values.insert(
         "gmt_modified".to_string(),
-        Arc::new(TimestampVector::from_slice(&[Timestamp::from_millis(
-            util::current_time_millis(),
-        )])) as _,
+        Arc::new(TimestampMillisecondVector::from_slice(&[now])) as _,
     );
 
     InsertRequest {
@@ -289,8 +297,8 @@ pub fn decode_system_catalog(
         }
 
         EntryType::Table => {
-            // As for table entry, the key is a string with format: `<catalog_name>.<schema_name>.<table_name>`
-            // and the value is a JSON string with format: `{"table_id": <table_id>}`
+            // As for table entry, the key is a string with format: `<catalog_name>.<schema_name>.<table_id>`
+            // and the value is a JSON string with format: `{"table_name": <table_name>}`
             let table_parts = key.split('.').collect::<Vec<_>>();
             ensure!(
                 table_parts.len() >= 3,
@@ -302,11 +310,12 @@ pub fn decode_system_catalog(
             debug!("Table meta value: {}", String::from_utf8_lossy(value));
             let table_meta: TableEntryValue =
                 serde_json::from_slice(value).context(ValueDeserializeSnafu)?;
+            let table_id = table_parts[2].parse::<TableId>().unwrap();
             Ok(Entry::Table(TableEntry {
                 catalog_name: table_parts[0].to_string(),
                 schema_name: table_parts[1].to_string(),
-                table_name: table_parts[2].to_string(),
-                table_id: table_meta.table_id,
+                table_name: table_meta.table_name,
+                table_id,
             }))
         }
     }
@@ -366,25 +375,25 @@ pub struct TableEntry {
 
 #[derive(Debug, Serialize, Deserialize, PartialEq, Eq)]
 pub struct TableEntryValue {
-    pub table_id: TableId,
+    pub table_name: String,
 }
 
 #[cfg(test)]
 mod tests {
-    use log_store::fs::noop::NoopLogStore;
+    use log_store::NoopLogStore;
+    use mito::config::EngineConfig;
+    use mito::engine::MitoEngine;
     use object_store::ObjectStore;
     use storage::config::EngineConfig as StorageEngineConfig;
     use storage::EngineImpl;
     use table::metadata::TableType;
     use table::metadata::TableType::Base;
-    use table_engine::config::EngineConfig;
-    use table_engine::engine::MitoEngine;
     use tempdir::TempDir;
 
     use super::*;
 
     #[test]
-    pub fn test_decode_catalog_enrty() {
+    pub fn test_decode_catalog_entry() {
         let entry = decode_system_catalog(
             Some(EntryType::Catalog as u8),
             Some("some_catalog".as_bytes()),
@@ -394,7 +403,7 @@ mod tests {
         if let Entry::Catalog(e) = entry {
             assert_eq!("some_catalog", e.catalog_name);
         } else {
-            panic!("Unexpected type: {:?}", entry);
+            panic!("Unexpected type: {entry:?}");
         }
     }
 
@@ -411,7 +420,7 @@ mod tests {
             assert_eq!("some_catalog", e.catalog_name);
             assert_eq!("some_schema", e.schema_name);
         } else {
-            panic!("Unexpected type: {:?}", entry);
+            panic!("Unexpected type: {entry:?}");
         }
     }
 
@@ -419,8 +428,8 @@ mod tests {
     pub fn test_decode_table() {
         let entry = decode_system_catalog(
             Some(EntryType::Table as u8),
-            Some("some_catalog.some_schema.some_table".as_bytes()),
-            Some("{\"table_id\":42}".as_bytes()),
+            Some("some_catalog.some_schema.42".as_bytes()),
+            Some("{\"table_name\":\"some_table\"}".as_bytes()),
         )
         .unwrap();
 
@@ -430,7 +439,7 @@ mod tests {
             assert_eq!("some_table", e.table_name);
             assert_eq!(42, e.table_id);
         } else {
-            panic!("Unexpected type: {:?}", entry);
+            panic!("Unexpected type: {entry:?}");
         }
     }
 
@@ -439,7 +448,7 @@ mod tests {
     pub fn test_decode_mismatch() {
         decode_system_catalog(
             Some(EntryType::Table as u8),
-            Some("some_catalog.some_schema.some_table".as_bytes()),
+            Some("some_catalog.some_schema.42".as_bytes()),
             None,
         )
         .unwrap();
@@ -456,7 +465,7 @@ mod tests {
     pub async fn prepare_table_engine() -> (TempDir, TableEngineRef) {
         let dir = TempDir::new("system-table-test").unwrap();
         let store_dir = dir.path().to_string_lossy();
-        let accessor = opendal::services::fs::Builder::default()
+        let accessor = object_store::backend::fs::Builder::default()
             .root(&store_dir)
             .build()
             .unwrap();

src/catalog/src/tables.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -26,9 +26,9 @@ use common_query::logical_plan::Expr;
 use common_query::physical_plan::PhysicalPlanRef;
 use common_recordbatch::error::Result as RecordBatchResult;
 use common_recordbatch::{RecordBatch, RecordBatchStream};
-use datatypes::prelude::{ConcreteDataType, VectorBuilder};
+use datatypes::prelude::{ConcreteDataType, DataType};
 use datatypes::schema::{ColumnSchema, Schema, SchemaRef};
-use datatypes::value::Value;
+use datatypes::value::ValueRef;
 use datatypes::vectors::VectorRef;
 use futures::Stream;
 use snafu::ResultExt;
@@ -40,9 +40,7 @@ use table::{Table, TableRef};
 
 use crate::error::{Error, InsertCatalogRecordSnafu};
 use crate::system::{build_schema_insert_request, build_table_insert_request, SystemCatalogTable};
-use crate::{
-    format_full_table_name, CatalogListRef, CatalogProvider, SchemaProvider, SchemaProviderRef,
-};
+use crate::{CatalogListRef, CatalogProvider, SchemaProvider, SchemaProviderRef};
 
 /// Tables holds all tables created by user.
 pub struct Tables {
@@ -77,7 +75,7 @@ impl Table for Tables {
 
     async fn scan(
         &self,
-        _projection: &Option<Vec<usize>>,
+        _projection: Option<&Vec<usize>>,
         _filters: &[Expr],
         _limit: Option<usize>,
     ) -> table::error::Result<PhysicalPlanRef> {
@@ -149,26 +147,33 @@ fn tables_to_record_batch(
     engine: &str,
 ) -> Vec<VectorRef> {
     let mut catalog_vec =
-        VectorBuilder::with_capacity(ConcreteDataType::string_datatype(), table_names.len());
+        ConcreteDataType::string_datatype().create_mutable_vector(table_names.len());
     let mut schema_vec =
-        VectorBuilder::with_capacity(ConcreteDataType::string_datatype(), table_names.len());
+        ConcreteDataType::string_datatype().create_mutable_vector(table_names.len());
     let mut table_name_vec =
-        VectorBuilder::with_capacity(ConcreteDataType::string_datatype(), table_names.len());
+        ConcreteDataType::string_datatype().create_mutable_vector(table_names.len());
     let mut engine_vec =
-        VectorBuilder::with_capacity(ConcreteDataType::string_datatype(), table_names.len());
+        ConcreteDataType::string_datatype().create_mutable_vector(table_names.len());
 
     for table_name in table_names {
-        catalog_vec.push(&Value::String(catalog_name.into()));
-        schema_vec.push(&Value::String(schema_name.into()));
-        table_name_vec.push(&Value::String(table_name.into()));
-        engine_vec.push(&Value::String(engine.into()));
+        // Safety: All these vectors are string type.
+        catalog_vec
+            .push_value_ref(ValueRef::String(catalog_name))
+            .unwrap();
+        schema_vec
+            .push_value_ref(ValueRef::String(schema_name))
+            .unwrap();
+        table_name_vec
+            .push_value_ref(ValueRef::String(&table_name))
+            .unwrap();
+        engine_vec.push_value_ref(ValueRef::String(engine)).unwrap();
     }
 
     vec![
-        catalog_vec.finish(),
-        schema_vec.finish(),
-        table_name_vec.finish(),
-        engine_vec.finish(),
+        catalog_vec.to_vector(),
+        schema_vec.to_vector(),
+        table_name_vec.to_vector(),
+        engine_vec.to_vector(),
     ]
 }
 
@@ -226,6 +231,10 @@ impl SchemaProvider for InformationSchema {
         panic!("System catalog & schema does not support register table")
     }
 
+    fn rename_table(&self, _name: &str, _new_name: String) -> crate::error::Result<TableRef> {
+        unimplemented!("System catalog & schema does not support rename table")
+    }
+
     fn deregister_table(&self, _name: &str) -> crate::error::Result<Option<TableRef>> {
         panic!("System catalog & schema does not support deregister table")
     }
@@ -262,8 +271,7 @@ impl SystemCatalog {
         table_name: String,
         table_id: TableId,
     ) -> crate::error::Result<usize> {
-        let full_table_name = format_full_table_name(&catalog, &schema, &table_name);
-        let request = build_table_insert_request(full_table_name, table_id);
+        let request = build_table_insert_request(catalog, schema, table_name, table_id);
         self.information_schema
             .system
             .insert(request)
@@ -340,9 +348,7 @@ fn build_schema_for_tables() -> Schema {
 #[cfg(test)]
 mod tests {
     use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
-    use common_query::physical_plan::RuntimeEnv;
-    use datatypes::arrow::array::Utf8Array;
-    use datatypes::arrow::datatypes::DataType;
+    use common_query::physical_plan::SessionContext;
     use futures_util::StreamExt;
     use table::table::numbers::NumbersTable;
 
@@ -365,58 +371,48 @@ mod tests {
             .unwrap();
 
         let tables = Tables::new(catalog_list, "test_engine".to_string());
-        let tables_stream = tables.scan(&None, &[], None).await.unwrap();
-        let mut tables_stream = tables_stream
-            .execute(0, Arc::new(RuntimeEnv::default()))
-            .await
-            .unwrap();
+        let tables_stream = tables.scan(None, &[], None).await.unwrap();
+        let session_ctx = SessionContext::new();
+        let mut tables_stream = tables_stream.execute(0, session_ctx.task_ctx()).unwrap();
 
         if let Some(t) = tables_stream.next().await {
-            let batch = t.unwrap().df_recordbatch;
+            let batch = t.unwrap();
             assert_eq!(1, batch.num_rows());
             assert_eq!(4, batch.num_columns());
-            assert_eq!(&DataType::Utf8, batch.column(0).data_type());
-            assert_eq!(&DataType::Utf8, batch.column(1).data_type());
-            assert_eq!(&DataType::Utf8, batch.column(2).data_type());
-            assert_eq!(&DataType::Utf8, batch.column(3).data_type());
+            assert_eq!(
+                ConcreteDataType::string_datatype(),
+                batch.column(0).data_type()
+            );
+            assert_eq!(
+                ConcreteDataType::string_datatype(),
+                batch.column(1).data_type()
+            );
+            assert_eq!(
+                ConcreteDataType::string_datatype(),
+                batch.column(2).data_type()
+            );
+            assert_eq!(
+                ConcreteDataType::string_datatype(),
+                batch.column(3).data_type()
+            );
             assert_eq!(
                 "greptime",
-                batch
-                    .column(0)
-                    .as_any()
-                    .downcast_ref::<Utf8Array<i32>>()
-                    .unwrap()
-                    .value(0)
+                batch.column(0).get_ref(0).as_string().unwrap().unwrap()
             );
 
             assert_eq!(
                 "public",
-                batch
-                    .column(1)
-                    .as_any()
-                    .downcast_ref::<Utf8Array<i32>>()
-                    .unwrap()
-                    .value(0)
+                batch.column(1).get_ref(0).as_string().unwrap().unwrap()
             );
 
             assert_eq!(
                 "test_table",
-                batch
-                    .column(2)
-                    .as_any()
-                    .downcast_ref::<Utf8Array<i32>>()
-                    .unwrap()
-                    .value(0)
+                batch.column(2).get_ref(0).as_string().unwrap().unwrap()
             );
 
             assert_eq!(
                 "test_engine",
-                batch
-                    .column(3)
-                    .as_any()
-                    .downcast_ref::<Utf8Array<i32>>()
-                    .unwrap()
-                    .value(0)
+                batch.column(3).get_ref(0).as_string().unwrap().unwrap()
             );
         } else {
             panic!("Record batch should not be empty!")

src/catalog/tests/local_catalog_tests.rs

@@ -0,0 +1,169 @@
+// 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.
+
+#[cfg(test)]
+mod tests {
+    use std::sync::Arc;
+
+    use catalog::local::LocalCatalogManager;
+    use catalog::{CatalogManager, RegisterTableRequest, RenameTableRequest};
+    use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
+    use common_telemetry::{error, info};
+    use mito::config::EngineConfig;
+    use table::table::numbers::NumbersTable;
+    use table::TableRef;
+    use tokio::sync::Mutex;
+
+    async fn create_local_catalog_manager() -> Result<LocalCatalogManager, catalog::error::Error> {
+        let (_dir, object_store) =
+            mito::table::test_util::new_test_object_store("setup_mock_engine_and_table").await;
+        let mock_engine = Arc::new(mito::table::test_util::MockMitoEngine::new(
+            EngineConfig::default(),
+            mito::table::test_util::MockEngine::default(),
+            object_store,
+        ));
+        let catalog_manager = LocalCatalogManager::try_new(mock_engine).await.unwrap();
+        catalog_manager.start().await?;
+        Ok(catalog_manager)
+    }
+
+    #[tokio::test]
+    async fn test_rename_table() {
+        common_telemetry::init_default_ut_logging();
+        let catalog_manager = create_local_catalog_manager().await.unwrap();
+        // register table
+        let table_name = "test_table";
+        let table_id = 42;
+        let table = Arc::new(NumbersTable::new(table_id));
+        let request = RegisterTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: table_name.to_string(),
+            table_id,
+            table: table.clone(),
+        };
+        assert!(catalog_manager.register_table(request).await.unwrap());
+
+        // rename table
+        let new_table_name = "table_t";
+        let rename_table_req = RenameTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: table_name.to_string(),
+            new_table_name: new_table_name.to_string(),
+            table_id,
+        };
+        assert!(catalog_manager
+            .rename_table(rename_table_req)
+            .await
+            .unwrap());
+
+        let registered_table = catalog_manager
+            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, new_table_name)
+            .unwrap()
+            .unwrap();
+        assert_eq!(registered_table.table_info().ident.table_id, table_id);
+    }
+
+    #[tokio::test]
+    async fn test_duplicate_register() {
+        let catalog_manager = create_local_catalog_manager().await.unwrap();
+        let request = RegisterTableRequest {
+            catalog: DEFAULT_CATALOG_NAME.to_string(),
+            schema: DEFAULT_SCHEMA_NAME.to_string(),
+            table_name: "test_table".to_string(),
+            table_id: 42,
+            table: Arc::new(NumbersTable::new(42)),
+        };
+        assert!(catalog_manager
+            .register_table(request.clone())
+            .await
+            .unwrap());
+
+        // register table with same table id will succeed with 0 as return val.
+        assert!(!catalog_manager.register_table(request).await.unwrap());
+
+        let err = catalog_manager
+            .register_table(RegisterTableRequest {
+                catalog: DEFAULT_CATALOG_NAME.to_string(),
+                schema: DEFAULT_SCHEMA_NAME.to_string(),
+                table_name: "test_table".to_string(),
+                table_id: 43,
+                table: Arc::new(NumbersTable::new(43)),
+            })
+            .await
+            .unwrap_err();
+        assert!(
+            err.to_string()
+                .contains("Table `greptime.public.test_table` already exists"),
+            "Actual error message: {err}",
+        );
+    }
+
+    #[test]
+    fn test_concurrent_register() {
+        common_telemetry::init_default_ut_logging();
+        let rt = Arc::new(tokio::runtime::Builder::new_multi_thread().build().unwrap());
+        let catalog_manager =
+            Arc::new(rt.block_on(async { create_local_catalog_manager().await.unwrap() }));
+
+        let succeed: Arc<Mutex<Option<TableRef>>> = Arc::new(Mutex::new(None));
+
+        let mut handles = Vec::with_capacity(8);
+        for i in 0..8 {
+            let catalog = catalog_manager.clone();
+            let succeed = succeed.clone();
+            let handle = rt.spawn(async move {
+                let table_id = 42 + i;
+                let table = Arc::new(NumbersTable::new(table_id));
+                let req = RegisterTableRequest {
+                    catalog: DEFAULT_CATALOG_NAME.to_string(),
+                    schema: DEFAULT_SCHEMA_NAME.to_string(),
+                    table_name: "test_table".to_string(),
+                    table_id,
+                    table: table.clone(),
+                };
+                match catalog.register_table(req).await {
+                    Ok(res) => {
+                        if res {
+                            let mut succeed = succeed.lock().await;
+                            info!("Successfully registered table: {}", table_id);
+                            *succeed = Some(table);
+                        }
+                    }
+                    Err(_) => {
+                        error!("Failed to register table {}", table_id);
+                    }
+                }
+            });
+            handles.push(handle);
+        }
+
+        rt.block_on(async move {
+            for handle in handles {
+                handle.await.unwrap();
+            }
+            let guard = succeed.lock().await;
+            let table = guard.as_ref().unwrap();
+            let table_registered = catalog_manager
+                .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, "test_table")
+                .unwrap()
+                .unwrap();
+            assert_eq!(
+                table_registered.table_info().ident.table_id,
+                table.table_info().ident.table_id
+            );
+        });
+    }
+}

src/catalog/tests/mock.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -189,10 +189,10 @@ impl TableEngine for MockTableEngine {
         unimplemented!()
     }
 
-    fn get_table<'a>(
+    fn get_table(
         &self,
         _ctx: &EngineContext,
-        table_ref: &'a TableReference,
+        table_ref: &TableReference,
     ) -> table::Result<Option<TableRef>> {
         futures::executor::block_on(async {
             Ok(self
@@ -204,7 +204,7 @@ impl TableEngine for MockTableEngine {
         })
     }
 
-    fn table_exists<'a>(&self, _ctx: &EngineContext, table_ref: &'a TableReference) -> bool {
+    fn table_exists(&self, _ctx: &EngineContext, table_ref: &TableReference) -> bool {
         futures::executor::block_on(async {
             self.tables
                 .read()
@@ -217,7 +217,7 @@ impl TableEngine for MockTableEngine {
         &self,
         _ctx: &EngineContext,
         _request: DropTableRequest,
-    ) -> table::Result<()> {
+    ) -> table::Result<bool> {
         unimplemented!()
     }
 }

src/catalog/tests/remote_catalog_tests.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -22,12 +22,12 @@ mod tests {
     use std::collections::HashSet;
     use std::sync::Arc;
 
+    use catalog::helper::{CatalogKey, CatalogValue, SchemaKey, SchemaValue};
     use catalog::remote::{
         KvBackend, KvBackendRef, RemoteCatalogManager, RemoteCatalogProvider, RemoteSchemaProvider,
     };
     use catalog::{CatalogList, CatalogManager, RegisterTableRequest};
     use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
-    use common_catalog::{CatalogKey, CatalogValue, SchemaKey, SchemaValue};
     use datatypes::schema::Schema;
     use futures_util::StreamExt;
     use table::engine::{EngineContext, TableEngineRef};
@@ -202,7 +202,7 @@ mod tests {
             table_id,
             table,
         };
-        assert_eq!(1, catalog_manager.register_table(reg_req).await.unwrap());
+        assert!(catalog_manager.register_table(reg_req).await.unwrap());
         assert_eq!(
             HashSet::from([table_name, "numbers".to_string()]),
             default_schema
@@ -287,7 +287,7 @@ mod tests {
             .register_schema(schema_name.clone(), schema.clone())
             .expect("Register schema should not fail");
         assert!(prev.is_none());
-        assert_eq!(1, catalog_manager.register_table(reg_req).await.unwrap());
+        assert!(catalog_manager.register_table(reg_req).await.unwrap());
 
         assert_eq!(
             HashSet::from([schema_name.clone()]),

src/client/Cargo.toml

@@ -1,34 +1,34 @@
 [package]
 name = "client"
-version = "0.1.0"
-edition = "2021"
-license = "Apache-2.0"
-# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 
 [dependencies]
 api = { path = "../api" }
-async-stream = "0.3"
+arrow-flight.workspace = true
+async-stream.workspace = true
 common-base = { path = "../common/base" }
 common-error = { path = "../common/error" }
 common-grpc = { path = "../common/grpc" }
+common-grpc-expr = { path = "../common/grpc-expr" }
 common-query = { path = "../common/query" }
 common-recordbatch = { path = "../common/recordbatch" }
-common-insert = { path = "../common/insert" }
 common-time = { path = "../common/time" }
-datafusion = { git = "https://github.com/apache/arrow-datafusion.git", branch = "arrow2", features = [
-    "simd",
-] }
+datafusion.workspace = true
 datatypes = { path = "../datatypes" }
 enum_dispatch = "0.3"
+futures-util.workspace = true
 parking_lot = "0.12"
+prost.workspace = true
 rand = "0.8"
-snafu = { version = "0.7", features = ["backtraces"] }
-tonic = "0.8"
+snafu.workspace = true
+tonic.workspace = true
 
 [dev-dependencies]
 datanode = { path = "../datanode" }
 substrait = { path = "../common/substrait" }
-tokio = { version = "1.0", features = ["full"] }
+tokio.workspace = true
 tracing = "0.1"
 tracing-subscriber = { version = "0.3", features = ["env-filter"] }
 

src/client/examples/insert.rs

@@ -1,109 +0,0 @@
-// Copyright 2022 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.
-
-use std::collections::HashMap;
-
-use api::v1::codec::InsertBatch;
-use api::v1::*;
-use client::{Client, Database};
-fn main() {
-    tracing::subscriber::set_global_default(tracing_subscriber::FmtSubscriber::builder().finish())
-        .unwrap();
-
-    run();
-}
-
-#[tokio::main]
-async fn run() {
-    let client = Client::with_urls(vec!["127.0.0.1:3001"]);
-    let db = Database::new("greptime", client);
-
-    let expr = InsertExpr {
-        schema_name: "public".to_string(),
-        table_name: "demo".to_string(),
-        expr: Some(insert_expr::Expr::Values(insert_expr::Values {
-            values: insert_batches(),
-        })),
-        options: HashMap::default(),
-        region_number: 0,
-    };
-    db.insert(expr).await.unwrap();
-}
-
-fn insert_batches() -> Vec<Vec<u8>> {
-    const SEMANTIC_TAG: i32 = 0;
-    const SEMANTIC_FEILD: i32 = 1;
-    const SEMANTIC_TS: i32 = 2;
-
-    let row_count = 4;
-
-    let host_vals = column::Values {
-        string_values: vec![
-            "host1".to_string(),
-            "host2".to_string(),
-            "host3".to_string(),
-            "host4".to_string(),
-        ],
-        ..Default::default()
-    };
-    let host_column = Column {
-        column_name: "host".to_string(),
-        semantic_type: SEMANTIC_TAG,
-        values: Some(host_vals),
-        null_mask: vec![0],
-        ..Default::default()
-    };
-
-    let cpu_vals = column::Values {
-        f64_values: vec![0.31, 0.41, 0.2],
-        ..Default::default()
-    };
-    let cpu_column = Column {
-        column_name: "cpu".to_string(),
-        semantic_type: SEMANTIC_FEILD,
-        values: Some(cpu_vals),
-        null_mask: vec![2],
-        ..Default::default()
-    };
-
-    let mem_vals = column::Values {
-        f64_values: vec![0.1, 0.2, 0.3],
-        ..Default::default()
-    };
-    let mem_column = Column {
-        column_name: "memory".to_string(),
-        semantic_type: SEMANTIC_FEILD,
-        values: Some(mem_vals),
-        null_mask: vec![4],
-        ..Default::default()
-    };
-
-    let ts_vals = column::Values {
-        i64_values: vec![100, 101, 102, 103],
-        ..Default::default()
-    };
-    let ts_column = Column {
-        column_name: "ts".to_string(),
-        semantic_type: SEMANTIC_TS,
-        values: Some(ts_vals),
-        null_mask: vec![0],
-        ..Default::default()
-    };
-
-    let insert_batch = InsertBatch {
-        columns: vec![host_column, cpu_column, mem_column, ts_column],
-        row_count,
-    };
-    vec![insert_batch.into()]
-}

src/client/examples/logical.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,8 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use api::v1::{ColumnDataType, ColumnDef, CreateExpr};
-use client::admin::Admin;
+use api::v1::{ColumnDataType, ColumnDef, CreateTableExpr, TableId};
 use client::{Client, Database};
 use prost_09::Message;
 use substrait_proto::protobuf::plan_rel::RelType as PlanRelType;
@@ -33,41 +32,41 @@ fn main() {
 async fn run() {
     let client = Client::with_urls(vec!["127.0.0.1:3001"]);
 
-    let create_table_expr = CreateExpr {
-        catalog_name: Some("greptime".to_string()),
-        schema_name: Some("public".to_string()),
+    let create_table_expr = CreateTableExpr {
+        catalog_name: "greptime".to_string(),
+        schema_name: "public".to_string(),
         table_name: "test_logical_dist_exec".to_string(),
-        desc: None,
+        desc: "".to_string(),
         column_defs: vec![
             ColumnDef {
                 name: "timestamp".to_string(),
-                datatype: ColumnDataType::Timestamp as i32,
+                datatype: ColumnDataType::TimestampMillisecond as i32,
                 is_nullable: false,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "key".to_string(),
                 datatype: ColumnDataType::Uint64 as i32,
                 is_nullable: false,
-                default_constraint: None,
+                default_constraint: vec![],
             },
             ColumnDef {
                 name: "value".to_string(),
                 datatype: ColumnDataType::Uint64 as i32,
                 is_nullable: false,
-                default_constraint: None,
+                default_constraint: vec![],
             },
         ],
         time_index: "timestamp".to_string(),
         primary_keys: vec!["key".to_string()],
         create_if_not_exists: false,
         table_options: Default::default(),
-        table_id: Some(1024),
+        table_id: Some(TableId { id: 1024 }),
         region_ids: vec![0],
     };
 
-    let admin = Admin::new("create table", client.clone());
-    let result = admin.create(create_table_expr).await.unwrap();
+    let db = Database::new("create table", client.clone());
+    let result = db.create(create_table_expr).await.unwrap();
     event!(Level::INFO, "create table result: {:#?}", result);
 
     let logical = mock_logical_plan();

src/client/examples/physical.rs

@@ -1,51 +0,0 @@
-// Copyright 2022 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.
-
-use std::sync::Arc;
-
-use client::{Client, Database};
-use common_grpc::MockExecution;
-use datafusion::physical_plan::expressions::Column;
-use datafusion::physical_plan::projection::ProjectionExec;
-use datafusion::physical_plan::{ExecutionPlan, PhysicalExpr};
-use tracing::{event, Level};
-
-fn main() {
-    tracing::subscriber::set_global_default(tracing_subscriber::FmtSubscriber::builder().finish())
-        .unwrap();
-
-    run();
-}
-
-#[tokio::main]
-async fn run() {
-    let client = Client::with_urls(vec!["127.0.0.1:3001"]);
-    let db = Database::new("greptime", client);
-
-    let physical = mock_physical_plan();
-    let result = db.physical_plan(physical, None).await;
-
-    event!(Level::INFO, "result: {:#?}", result);
-}
-
-fn mock_physical_plan() -> Arc<dyn ExecutionPlan> {
-    let id_expr = Arc::new(Column::new("id", 0)) as Arc<dyn PhysicalExpr>;
-    let age_expr = Arc::new(Column::new("age", 2)) as Arc<dyn PhysicalExpr>;
-    let expr = vec![(id_expr, "id".to_string()), (age_expr, "age".to_string())];
-
-    let input =
-        Arc::new(MockExecution::new("mock_input_exec".to_string())) as Arc<dyn ExecutionPlan>;
-    let projection = ProjectionExec::try_new(expr, input).unwrap();
-    Arc::new(projection)
-}

src/client/examples/select.rs

@@ -1,34 +0,0 @@
-// Copyright 2022 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.
-
-use client::{Client, Database, Select};
-use tracing::{event, Level};
-
-fn main() {
-    tracing::subscriber::set_global_default(tracing_subscriber::FmtSubscriber::builder().finish())
-        .unwrap();
-
-    run();
-}
-
-#[tokio::main]
-async fn run() {
-    let client = Client::with_urls(vec!["127.0.0.1:3001"]);
-    let db = Database::new("greptime", client);
-
-    let sql = Select::Sql("select * from demo".to_string());
-    let result = db.select(sql).await.unwrap();
-
-    event!(Level::INFO, "result: {:#?}", result);
-}

src/client/src/admin.rs

@@ -1,125 +0,0 @@
-// Copyright 2022 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.
-
-use api::v1::*;
-use common_error::prelude::StatusCode;
-use common_query::Output;
-use snafu::prelude::*;
-
-use crate::database::PROTOCOL_VERSION;
-use crate::{error, Client, Result};
-
-#[derive(Clone, Debug)]
-pub struct Admin {
-    name: String,
-    client: Client,
-}
-
-impl Admin {
-    pub fn new(name: impl Into<String>, client: Client) -> Self {
-        Self {
-            name: name.into(),
-            client,
-        }
-    }
-
-    pub async fn create(&self, expr: CreateExpr) -> Result<AdminResult> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-        let expr = AdminExpr {
-            header: Some(header),
-            expr: Some(admin_expr::Expr::Create(expr)),
-        };
-        self.do_request(expr).await
-    }
-
-    pub async fn do_request(&self, expr: AdminExpr) -> Result<AdminResult> {
-        // `remove(0)` is safe because of `do_requests`'s invariants.
-        Ok(self.do_requests(vec![expr]).await?.remove(0))
-    }
-
-    pub async fn alter(&self, expr: AlterExpr) -> Result<AdminResult> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-        let expr = AdminExpr {
-            header: Some(header),
-            expr: Some(admin_expr::Expr::Alter(expr)),
-        };
-        Ok(self.do_requests(vec![expr]).await?.remove(0))
-    }
-
-    /// Invariants: the lengths of input vec (`Vec<AdminExpr>`) and output vec (`Vec<AdminResult>`) are equal.
-    async fn do_requests(&self, exprs: Vec<AdminExpr>) -> Result<Vec<AdminResult>> {
-        let expr_count = exprs.len();
-        let req = AdminRequest {
-            name: self.name.clone(),
-            exprs,
-        };
-
-        let resp = self.client.admin(req).await?;
-
-        let results = resp.results;
-        ensure!(
-            results.len() == expr_count,
-            error::MissingResultSnafu {
-                name: "admin_results",
-                expected: expr_count,
-                actual: results.len(),
-            }
-        );
-        Ok(results)
-    }
-
-    pub async fn create_database(&self, expr: CreateDatabaseExpr) -> Result<AdminResult> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-        let expr = AdminExpr {
-            header: Some(header),
-            expr: Some(admin_expr::Expr::CreateDatabase(expr)),
-        };
-        Ok(self.do_requests(vec![expr]).await?.remove(0))
-    }
-}
-
-pub fn admin_result_to_output(admin_result: AdminResult) -> Result<Output> {
-    let header = admin_result.header.context(error::MissingHeaderSnafu)?;
-    if !StatusCode::is_success(header.code) {
-        return error::DatanodeSnafu {
-            code: header.code,
-            msg: header.err_msg,
-        }
-        .fail();
-    }
-
-    let result = admin_result.result.context(error::MissingResultSnafu {
-        name: "result".to_string(),
-        expected: 1_usize,
-        actual: 0_usize,
-    })?;
-    let output = match result {
-        admin_result::Result::Mutate(mutate) => {
-            if mutate.failure != 0 {
-                return error::MutateFailureSnafu {
-                    failure: mutate.failure,
-                }
-                .fail();
-            }
-            Output::AffectedRows(mutate.success as usize)
-        }
-    };
-    Ok(output)
-}

src/client/src/client.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -14,8 +14,7 @@
 
 use std::sync::Arc;
 
-use api::v1::greptime_client::GreptimeClient;
-use api::v1::*;
+use arrow_flight::flight_service_client::FlightServiceClient;
 use common_grpc::channel_manager::ChannelManager;
 use parking_lot::RwLock;
 use snafu::{OptionExt, ResultExt};
@@ -24,6 +23,21 @@ use tonic::transport::Channel;
 use crate::load_balance::{LoadBalance, Loadbalancer};
 use crate::{error, Result};
 
+pub(crate) struct FlightClient {
+    addr: String,
+    client: FlightServiceClient<Channel>,
+}
+
+impl FlightClient {
+    pub(crate) fn addr(&self) -> &str {
+        &self.addr
+    }
+
+    pub(crate) fn mut_inner(&mut self) -> &mut FlightServiceClient<Channel> {
+        &mut self.client
+    }
+}
+
 #[derive(Clone, Debug, Default)]
 pub struct Client {
     inner: Arc<Inner>,
@@ -104,57 +118,23 @@ impl Client {
         self.inner.set_peers(urls);
     }
 
-    pub async fn admin(&self, req: AdminRequest) -> Result<AdminResponse> {
-        let req = BatchRequest {
-            admins: vec![req],
-            ..Default::default()
-        };
-
-        let mut res = self.batch(req).await?;
-        res.admins.pop().context(error::MissingResultSnafu {
-            name: "admins",
-            expected: 1_usize,
-            actual: 0_usize,
-        })
-    }
-
-    pub async fn database(&self, req: DatabaseRequest) -> Result<DatabaseResponse> {
-        let req = BatchRequest {
-            databases: vec![req],
-            ..Default::default()
-        };
-
-        let mut res = self.batch(req).await?;
-        res.databases.pop().context(error::MissingResultSnafu {
-            name: "database",
-            expected: 1_usize,
-            actual: 0_usize,
-        })
-    }
-
-    pub async fn batch(&self, req: BatchRequest) -> Result<BatchResponse> {
-        let peer = self
+    pub(crate) fn make_client(&self) -> Result<FlightClient> {
+        let addr = self
             .inner
             .get_peer()
             .context(error::IllegalGrpcClientStateSnafu {
                 err_msg: "No available peer found",
             })?;
-        let mut client = self.make_client(&peer)?;
-        let result = client
-            .batch(req)
-            .await
-            .context(error::TonicStatusSnafu { addr: peer })?;
-        Ok(result.into_inner())
-    }
 
-    fn make_client(&self, addr: impl AsRef<str>) -> Result<GreptimeClient<Channel>> {
-        let addr = addr.as_ref();
         let channel = self
             .inner
             .channel_manager
-            .get(addr)
-            .context(error::CreateChannelSnafu { addr })?;
-        Ok(GreptimeClient::new(channel))
+            .get(&addr)
+            .context(error::CreateChannelSnafu { addr: &addr })?;
+        Ok(FlightClient {
+            addr,
+            client: FlightServiceClient::new(channel),
+        })
     }
 }
 

src/client/src/database.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,31 +12,26 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use std::sync::Arc;
+use std::str::FromStr;
 
-use api::v1::codec::SelectResult as GrpcSelectResult;
+use api::v1::ddl_request::Expr as DdlExpr;
+use api::v1::greptime_request::Request;
+use api::v1::query_request::Query;
 use api::v1::{
-    object_expr, object_result, select_expr, DatabaseRequest, ExprHeader, InsertExpr,
-    MutateResult as GrpcMutateResult, ObjectExpr, ObjectResult as GrpcObjectResult, PhysicalPlan,
-    SelectExpr,
+    AlterExpr, CreateTableExpr, DdlRequest, DropTableExpr, GreptimeRequest, InsertRequest,
+    QueryRequest,
 };
-use common_error::status_code::StatusCode;
-use common_grpc::{AsExcutionPlan, DefaultAsPlanImpl};
-use common_insert::column_to_vector;
+use arrow_flight::{FlightData, Ticket};
+use common_error::prelude::*;
+use common_grpc::flight::{flight_messages_to_recordbatches, FlightDecoder, FlightMessage};
 use common_query::Output;
-use common_recordbatch::{RecordBatch, RecordBatches};
-use datafusion::physical_plan::ExecutionPlan;
-use datatypes::prelude::*;
-use datatypes::schema::{ColumnSchema, Schema};
-use snafu::{ensure, OptionExt, ResultExt};
+use futures_util::{TryFutureExt, TryStreamExt};
+use prost::Message;
+use snafu::{ensure, ResultExt};
 
-use crate::error::{
-    ColumnToVectorSnafu, ConvertSchemaSnafu, DatanodeSnafu, DecodeSelectSnafu, EncodePhysicalSnafu,
-};
+use crate::error::{ConvertFlightDataSnafu, IllegalFlightMessagesSnafu};
 use crate::{error, Client, Result};
 
-pub const PROTOCOL_VERSION: u32 = 1;
-
 #[derive(Clone, Debug)]
 pub struct Database {
     name: String,
@@ -55,211 +50,128 @@ impl Database {
         &self.name
     }
 
-    pub async fn insert(&self, insert: InsertExpr) -> Result<ObjectResult> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-        let expr = ObjectExpr {
-            header: Some(header),
-            expr: Some(object_expr::Expr::Insert(insert)),
-        };
-        self.object(expr).await?.try_into()
-    }
-
-    pub async fn batch_insert(&self, insert_exprs: Vec<InsertExpr>) -> Result<Vec<ObjectResult>> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-        let obj_exprs = insert_exprs
-            .into_iter()
-            .map(|expr| ObjectExpr {
-                header: Some(header.clone()),
-                expr: Some(object_expr::Expr::Insert(expr)),
-            })
-            .collect();
-        self.objects(obj_exprs)
-            .await?
-            .into_iter()
-            .map(|result| result.try_into())
-            .collect()
-    }
-
-    pub async fn select(&self, expr: Select) -> Result<ObjectResult> {
-        let select_expr = match expr {
-            Select::Sql(sql) => SelectExpr {
-                expr: Some(select_expr::Expr::Sql(sql)),
-            },
-        };
-        self.do_select(select_expr).await
-    }
-
-    pub async fn physical_plan(
-        &self,
-        physical: Arc<dyn ExecutionPlan>,
-        original_ql: Option<String>,
-    ) -> Result<ObjectResult> {
-        let plan = DefaultAsPlanImpl::try_from_physical_plan(physical.clone())
-            .context(EncodePhysicalSnafu { physical })?
-            .bytes;
-        let original_ql = original_ql.unwrap_or_default();
-        let select_expr = SelectExpr {
-            expr: Some(select_expr::Expr::PhysicalPlan(PhysicalPlan {
-                original_ql: original_ql.into_bytes(),
-                plan,
-            })),
-        };
-        self.do_select(select_expr).await
-    }
-
-    pub async fn logical_plan(&self, logical_plan: Vec<u8>) -> Result<ObjectResult> {
-        let select_expr = SelectExpr {
-            expr: Some(select_expr::Expr::LogicalPlan(logical_plan)),
-        };
-        self.do_select(select_expr).await
-    }
-
-    async fn do_select(&self, select_expr: SelectExpr) -> Result<ObjectResult> {
-        let header = ExprHeader {
-            version: PROTOCOL_VERSION,
-        };
-
-        let expr = ObjectExpr {
-            header: Some(header),
-            expr: Some(object_expr::Expr::Select(select_expr)),
-        };
-
-        let obj_result = self.object(expr).await?;
-        obj_result.try_into()
-    }
-
-    pub async fn object(&self, expr: ObjectExpr) -> Result<GrpcObjectResult> {
-        let res = self.objects(vec![expr]).await?.pop().unwrap();
-        Ok(res)
-    }
-
-    async fn objects(&self, exprs: Vec<ObjectExpr>) -> Result<Vec<GrpcObjectResult>> {
-        let expr_count = exprs.len();
-        let req = DatabaseRequest {
-            name: self.name.clone(),
-            exprs,
-        };
-
-        let res = self.client.database(req).await?;
-        let res = res.results;
-
-        ensure!(
-            res.len() == expr_count,
-            error::MissingResultSnafu {
-                name: "object_results",
-                expected: expr_count,
-                actual: res.len(),
-            }
-        );
-
-        Ok(res)
-    }
-}
-
-#[derive(Debug)]
-pub enum ObjectResult {
-    Select(GrpcSelectResult),
-    Mutate(GrpcMutateResult),
-}
-
-impl TryFrom<api::v1::ObjectResult> for ObjectResult {
-    type Error = error::Error;
-
-    fn try_from(object_result: api::v1::ObjectResult) -> std::result::Result<Self, Self::Error> {
-        let header = object_result.header.context(error::MissingHeaderSnafu)?;
-        if !StatusCode::is_success(header.code) {
-            return DatanodeSnafu {
-                code: header.code,
-                msg: header.err_msg,
-            }
-            .fail();
-        }
-
-        let obj_result = object_result.result.context(error::MissingResultSnafu {
-            name: "result".to_string(),
-            expected: 1_usize,
-            actual: 0_usize,
-        })?;
-        Ok(match obj_result {
-            object_result::Result::Select(select) => {
-                let result = (*select.raw_data).try_into().context(DecodeSelectSnafu)?;
-                ObjectResult::Select(result)
-            }
-            object_result::Result::Mutate(mutate) => ObjectResult::Mutate(mutate),
+    pub async fn insert(&self, request: InsertRequest) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Insert(request)),
         })
+        .await
     }
-}
 
-pub enum Select {
-    Sql(String),
-}
+    pub async fn sql(&self, sql: &str) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Query(QueryRequest {
+                query: Some(Query::Sql(sql.to_string())),
+            })),
+        })
+        .await
+    }
 
-impl TryFrom<ObjectResult> for Output {
-    type Error = error::Error;
+    pub async fn logical_plan(&self, logical_plan: Vec<u8>) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Query(QueryRequest {
+                query: Some(Query::LogicalPlan(logical_plan)),
+            })),
+        })
+        .await
+    }
 
-    fn try_from(value: ObjectResult) -> Result<Self> {
-        let output = match value {
-            ObjectResult::Select(select) => {
-                let vectors = select
-                    .columns
-                    .iter()
-                    .map(|column| {
-                        column_to_vector(column, select.row_count).context(ColumnToVectorSnafu)
+    pub async fn create(&self, expr: CreateTableExpr) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Ddl(DdlRequest {
+                expr: Some(DdlExpr::CreateTable(expr)),
+            })),
+        })
+        .await
+    }
+
+    pub async fn alter(&self, expr: AlterExpr) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Ddl(DdlRequest {
+                expr: Some(DdlExpr::Alter(expr)),
+            })),
+        })
+        .await
+    }
+
+    pub async fn drop_table(&self, expr: DropTableExpr) -> Result<Output> {
+        self.do_get(GreptimeRequest {
+            request: Some(Request::Ddl(DdlRequest {
+                expr: Some(DdlExpr::DropTable(expr)),
+            })),
+        })
+        .await
+    }
+
+    async fn do_get(&self, request: GreptimeRequest) -> Result<Output> {
+        let mut client = self.client.make_client()?;
+
+        // TODO(LFC): Streaming get flight data.
+        let flight_data: Vec<FlightData> = client
+            .mut_inner()
+            .do_get(Ticket {
+                ticket: request.encode_to_vec(),
+            })
+            .and_then(|response| response.into_inner().try_collect())
+            .await
+            .map_err(|e| {
+                let code = get_metadata_value(&e, INNER_ERROR_CODE)
+                    .and_then(|s| StatusCode::from_str(&s).ok())
+                    .unwrap_or(StatusCode::Unknown);
+                let msg = get_metadata_value(&e, INNER_ERROR_MSG).unwrap_or(e.to_string());
+                error::ExternalSnafu { code, msg }
+                    .fail::<()>()
+                    .map_err(BoxedError::new)
+                    .context(error::FlightGetSnafu {
+                        tonic_code: e.code(),
+                        addr: client.addr(),
                     })
-                    .collect::<Result<Vec<VectorRef>>>()?;
+                    .unwrap_err()
+            })?;
 
-                let column_schemas = select
-                    .columns
-                    .iter()
-                    .zip(vectors.iter())
-                    .map(|(column, vector)| {
-                        let datatype = vector.data_type();
-                        // nullable or not, does not affect the output
-                        ColumnSchema::new(&column.column_name, datatype, true)
-                    })
-                    .collect::<Vec<ColumnSchema>>();
+        let decoder = &mut FlightDecoder::default();
+        let flight_messages = flight_data
+            .into_iter()
+            .map(|x| decoder.try_decode(x).context(ConvertFlightDataSnafu))
+            .collect::<Result<Vec<_>>>()?;
 
-                let schema = Arc::new(Schema::try_new(column_schemas).context(ConvertSchemaSnafu)?);
-                let recordbatches = if vectors.is_empty() {
-                    RecordBatches::try_new(schema, vec![])
-                } else {
-                    RecordBatch::new(schema, vectors)
-                        .and_then(|batch| RecordBatches::try_new(batch.schema.clone(), vec![batch]))
+        let output = if let Some(FlightMessage::AffectedRows(rows)) = flight_messages.get(0) {
+            ensure!(
+                flight_messages.len() == 1,
+                IllegalFlightMessagesSnafu {
+                    reason: "Expect 'AffectedRows' Flight messages to be one and only!"
                 }
-                .context(error::CreateRecordBatchesSnafu)?;
-                Output::RecordBatches(recordbatches)
-            }
-            ObjectResult::Mutate(mutate) => {
-                if mutate.failure != 0 {
-                    return error::MutateFailureSnafu {
-                        failure: mutate.failure,
-                    }
-                    .fail();
-                }
-                Output::AffectedRows(mutate.success as usize)
-            }
+            );
+            Output::AffectedRows(*rows)
+        } else {
+            let recordbatches = flight_messages_to_recordbatches(flight_messages)
+                .context(ConvertFlightDataSnafu)?;
+            Output::RecordBatches(recordbatches)
         };
         Ok(output)
     }
 }
 
+fn get_metadata_value(e: &tonic::Status, key: &str) -> Option<String> {
+    e.metadata()
+        .get(key)
+        .and_then(|v| String::from_utf8(v.as_bytes().to_vec()).ok())
+}
+
 #[cfg(test)]
 mod tests {
+    use std::sync::Arc;
+
     use api::helper::ColumnDataTypeWrapper;
     use api::v1::Column;
-    use datanode::server::grpc::select::{null_mask, values};
+    use common_grpc::select::{null_mask, values};
+    use common_grpc_expr::column_to_vector;
+    use datatypes::prelude::{Vector, VectorRef};
     use datatypes::vectors::{
         BinaryVector, BooleanVector, DateTimeVector, DateVector, Float32Vector, Float64Vector,
         Int16Vector, Int32Vector, Int64Vector, Int8Vector, StringVector, UInt16Vector,
         UInt32Vector, UInt64Vector, UInt8Vector,
     };
 
-    use super::*;
-
     #[test]
     fn test_column_to_vector() {
         let mut column = create_test_column(Arc::new(BooleanVector::from(vec![true])));
@@ -335,12 +247,11 @@ mod tests {
 
     fn create_test_column(vector: VectorRef) -> Column {
         let wrapper: ColumnDataTypeWrapper = vector.data_type().try_into().unwrap();
-        let array = vector.to_arrow_array();
         Column {
             column_name: "test".to_string(),
             semantic_type: 1,
-            values: Some(values(&[array.clone()]).unwrap()),
-            null_mask: null_mask(&vec![array], vector.len()),
+            values: Some(values(&[vector.clone()]).unwrap()),
+            null_mask: null_mask(&[vector.clone()], vector.len()),
             datatype: wrapper.datatype() as i32,
         }
     }

src/client/src/error.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -13,67 +13,43 @@
 // limitations under the License.
 
 use std::any::Any;
-use std::sync::Arc;
 
-use api::serde::DecodeError;
 use common_error::prelude::*;
-use datafusion::physical_plan::ExecutionPlan;
+use tonic::Code;
 
 #[derive(Debug, Snafu)]
 #[snafu(visibility(pub))]
 pub enum Error {
-    #[snafu(display("Connect failed to {}, source: {}", url, source))]
-    ConnectFailed {
-        url: String,
-        source: tonic::transport::Error,
+    #[snafu(display("Illegal Flight messages, reason: {}", reason))]
+    IllegalFlightMessages {
+        reason: String,
         backtrace: Backtrace,
     },
 
-    #[snafu(display("Missing {}, expected {}, actual {}", name, expected, actual))]
-    MissingResult {
-        name: String,
-        expected: usize,
-        actual: usize,
-    },
-
-    #[snafu(display("Missing result header"))]
-    MissingHeader,
-
-    #[snafu(display("Tonic internal error, addr: {}, source: {}", addr, source))]
-    TonicStatus {
+    #[snafu(display(
+        "Failed to do Flight get, addr: {}, code: {}, source: {}",
+        addr,
+        tonic_code,
+        source
+    ))]
+    FlightGet {
         addr: String,
-        source: tonic::Status,
-        backtrace: Backtrace,
+        tonic_code: Code,
+        source: BoxedError,
     },
 
-    #[snafu(display("Fail to decode select result, source: {}", source))]
-    DecodeSelect { source: DecodeError },
-
-    #[snafu(display("Error occurred on the data node, code: {}, msg: {}", code, msg))]
-    Datanode { code: u32, msg: String },
-
-    #[snafu(display("Failed to encode physical plan: {:?}, source: {}", physical, source))]
-    EncodePhysical {
-        physical: Arc<dyn ExecutionPlan>,
+    #[snafu(display("Failed to convert FlightData, source: {}", source))]
+    ConvertFlightData {
         #[snafu(backtrace)]
         source: common_grpc::Error,
     },
 
-    #[snafu(display("Mutate result has failure {}", failure))]
-    MutateFailure { failure: u32, backtrace: Backtrace },
-
     #[snafu(display("Column datatype error, source: {}", source))]
     ColumnDataType {
         #[snafu(backtrace)]
         source: api::error::Error,
     },
 
-    #[snafu(display("Failed to create RecordBatches, source: {}", source))]
-    CreateRecordBatches {
-        #[snafu(backtrace)]
-        source: common_recordbatch::error::Error,
-    },
-
     #[snafu(display("Illegal GRPC client state: {}", err_msg))]
     IllegalGrpcClientState {
         err_msg: String,
@@ -83,12 +59,6 @@ pub enum Error {
     #[snafu(display("Missing required field in protobuf, field: {}", field))]
     MissingField { field: String, backtrace: Backtrace },
 
-    #[snafu(display("Failed to convert schema, source: {}", source))]
-    ConvertSchema {
-        #[snafu(backtrace)]
-        source: datatypes::error::Error,
-    },
-
     #[snafu(display(
         "Failed to create gRPC channel, peer address: {}, source: {}",
         addr,
@@ -100,11 +70,9 @@ pub enum Error {
         source: common_grpc::error::Error,
     },
 
-    #[snafu(display("Failed to convert column to vector, source: {}", source))]
-    ColumnToVector {
-        #[snafu(backtrace)]
-        source: common_insert::error::Error,
-    },
+    /// Error deserialized from gRPC metadata
+    #[snafu(display("{}", msg))]
+    ExternalError { code: StatusCode, msg: String },
 }
 
 pub type Result<T> = std::result::Result<T, Error>;
@@ -112,21 +80,15 @@ pub type Result<T> = std::result::Result<T, Error>;
 impl ErrorExt for Error {
     fn status_code(&self) -> StatusCode {
         match self {
-            Error::ConnectFailed { .. }
-            | Error::MissingResult { .. }
-            | Error::MissingHeader { .. }
-            | Error::TonicStatus { .. }
-            | Error::DecodeSelect { .. }
-            | Error::Datanode { .. }
-            | Error::EncodePhysical { .. }
-            | Error::MutateFailure { .. }
+            Error::IllegalFlightMessages { .. }
             | Error::ColumnDataType { .. }
             | Error::MissingField { .. } => StatusCode::Internal,
-            Error::ConvertSchema { source } => source.status_code(),
-            Error::CreateRecordBatches { source } => source.status_code(),
-            Error::CreateChannel { source, .. } => source.status_code(),
+            Error::FlightGet { source, .. } => source.status_code(),
+            Error::CreateChannel { source, .. } | Error::ConvertFlightData { source } => {
+                source.status_code()
+            }
             Error::IllegalGrpcClientState { .. } => StatusCode::Unexpected,
-            Error::ColumnToVector { source, .. } => source.status_code(),
+            Error::ExternalError { code, .. } => *code,
         }
     }
 

src/client/src/lib.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,7 +12,6 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-pub mod admin;
 mod client;
 mod database;
 mod error;
@@ -21,5 +20,5 @@ pub mod load_balance;
 pub use api;
 
 pub use self::client::Client;
-pub use self::database::{Database, ObjectResult, Select};
+pub use self::database::Database;
 pub use self::error::{Error, Result};

src/client/src/load_balance.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/cmd/Cargo.toml

@@ -1,15 +1,16 @@
 [package]
 name = "cmd"
-version = "0.1.0"
-edition = "2021"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 default-run = "greptime"
-license = "Apache-2.0"
 
 [[bin]]
 name = "greptime"
 path = "src/bin/greptime.rs"
 
 [dependencies]
+anymap = "1.0.0-beta.2"
 clap = { version = "3.1", features = ["derive"] }
 common-error = { path = "../common/error" }
 common-telemetry = { path = "../common/telemetry", features = [
@@ -17,13 +18,18 @@ common-telemetry = { path = "../common/telemetry", features = [
 ] }
 datanode = { path = "../datanode" }
 frontend = { path = "../frontend" }
-futures = "0.3"
+futures.workspace = true
+meta-client = { path = "../meta-client" }
 meta-srv = { path = "../meta-srv" }
-serde = "1.0"
-snafu = { version = "0.7", features = ["backtraces"] }
+serde.workspace = true
+servers = { path = "../servers" }
+snafu.workspace = true
 tokio = { version = "1.18", features = ["full"] }
 toml = "0.5"
 
 [dev-dependencies]
-serde = "1.0"
+serde.workspace = true
 tempdir = "0.3"
+
+[build-dependencies]
+build-data = "0.1.3"

src/cmd/build.rs

@@ -0,0 +1,29 @@
+// 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.
+
+const DEFAULT_VALUE: &str = "unknown";
+fn main() {
+    println!(
+        "cargo:rustc-env=GIT_COMMIT={}",
+        build_data::get_git_commit().unwrap_or_else(|_| DEFAULT_VALUE.to_string())
+    );
+    println!(
+        "cargo:rustc-env=GIT_BRANCH={}",
+        build_data::get_git_branch().unwrap_or_else(|_| DEFAULT_VALUE.to_string())
+    );
+    println!(
+        "cargo:rustc-env=GIT_DIRTY={}",
+        build_data::get_git_dirty().map_or(DEFAULT_VALUE.to_string(), |v| v.to_string())
+    );
+}

src/cmd/src/bin/greptime.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -20,7 +20,7 @@ use cmd::{datanode, frontend, metasrv, standalone};
 use common_telemetry::logging::{error, info};
 
 #[derive(Parser)]
-#[clap(name = "greptimedb")]
+#[clap(name = "greptimedb", version = print_version())]
 struct Command {
     #[clap(long, default_value = "/tmp/greptimedb/logs")]
     log_dir: String,
@@ -70,6 +70,19 @@ impl fmt::Display for SubCommand {
     }
 }
 
+fn print_version() -> &'static str {
+    concat!(
+        "\nbranch: ",
+        env!("GIT_BRANCH"),
+        "\ncommit: ",
+        env!("GIT_COMMIT"),
+        "\ndirty: ",
+        env!("GIT_DIRTY"),
+        "\nversion: ",
+        env!("CARGO_PKG_VERSION")
+    )
+}
+
 #[tokio::main]
 async fn main() -> Result<()> {
     let cmd = Command::parse();

src/cmd/src/datanode.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -14,8 +14,9 @@
 
 use clap::Parser;
 use common_telemetry::logging;
-use datanode::datanode::{Datanode, DatanodeOptions};
-use frontend::frontend::Mode;
+use datanode::datanode::{Datanode, DatanodeOptions, ObjectStoreConfig};
+use meta_client::MetaClientOpts;
+use servers::Mode;
 use snafu::ResultExt;
 
 use crate::error::{Error, MissingConfigSnafu, Result, StartDatanodeSnafu};
@@ -46,7 +47,7 @@ impl SubCommand {
     }
 }
 
-#[derive(Debug, Parser)]
+#[derive(Debug, Parser, Default)]
 struct StartCommand {
     #[clap(long)]
     node_id: Option<u64>,
@@ -58,6 +59,10 @@ struct StartCommand {
     metasrv_addr: Option<String>,
     #[clap(short, long)]
     config_file: Option<String>,
+    #[clap(long)]
+    data_dir: Option<String>,
+    #[clap(long)]
+    wal_dir: Option<String>,
 }
 
 impl StartCommand {
@@ -93,31 +98,34 @@ impl TryFrom<StartCommand> for DatanodeOptions {
             opts.mysql_addr = addr;
         }
 
-        match (cmd.metasrv_addr, cmd.node_id) {
-            (Some(meta_addr), Some(node_id)) => {
-                // Running mode is only set to Distributed when
-                // both metasrv addr and node id are set in
-                // commandline options
-                opts.meta_client_opts.metasrv_addr = meta_addr.clone();
-                opts.node_id = node_id;
-                opts.metasrv_addr = Some(vec![meta_addr]);
-                opts.mode = Mode::Distributed;
-            }
-            (None, None) => {
-                opts.mode = Mode::Standalone;
-            }
-            (None, Some(_)) => {
-                return MissingConfigSnafu {
-                    msg: "Missing metasrv address option",
-                }
-                .fail();
-            }
-            (Some(_), None) => {
-                return MissingConfigSnafu {
-                    msg: "Missing node id option",
-                }
-                .fail();
+        if let Some(node_id) = cmd.node_id {
+            opts.node_id = Some(node_id);
+        }
+
+        if let Some(meta_addr) = cmd.metasrv_addr {
+            opts.meta_client_opts
+                .get_or_insert_with(MetaClientOpts::default)
+                .metasrv_addrs = meta_addr
+                .split(',')
+                .map(&str::trim)
+                .map(&str::to_string)
+                .collect::<_>();
+            opts.mode = Mode::Distributed;
+        }
+
+        if let (Mode::Distributed, None) = (&opts.mode, &opts.node_id) {
+            return MissingConfigSnafu {
+                msg: "Missing node id option",
             }
+            .fail();
+        }
+
+        if let Some(data_dir) = cmd.data_dir {
+            opts.storage = ObjectStoreConfig::File { data_dir };
+        }
+
+        if let Some(wal_dir) = cmd.wal_dir {
+            opts.wal.dir = wal_dir;
         }
         Ok(opts)
     }
@@ -126,41 +134,44 @@ impl TryFrom<StartCommand> for DatanodeOptions {
 #[cfg(test)]
 mod tests {
     use std::assert_matches::assert_matches;
+    use std::time::Duration;
 
     use datanode::datanode::ObjectStoreConfig;
-    use frontend::frontend::Mode;
+    use servers::Mode;
 
     use super::*;
 
     #[test]
     fn test_read_from_config_file() {
         let cmd = StartCommand {
-            node_id: None,
-            rpc_addr: None,
-            mysql_addr: None,
-            metasrv_addr: None,
             config_file: Some(format!(
                 "{}/../../config/datanode.example.toml",
                 std::env::current_dir().unwrap().as_path().to_str().unwrap()
             )),
+            ..Default::default()
         };
         let options: DatanodeOptions = cmd.try_into().unwrap();
-        assert_eq!("0.0.0.0:3001".to_string(), options.rpc_addr);
-        assert_eq!("/tmp/greptimedb/wal".to_string(), options.wal_dir);
-        assert_eq!("0.0.0.0:3306".to_string(), options.mysql_addr);
+        assert_eq!("127.0.0.1:3001".to_string(), options.rpc_addr);
+        assert_eq!("/tmp/greptimedb/wal".to_string(), options.wal.dir);
+        assert_eq!("127.0.0.1:4406".to_string(), options.mysql_addr);
         assert_eq!(4, options.mysql_runtime_size);
-        assert_eq!(
-            "1.1.1.1:3002".to_string(),
-            options.meta_client_opts.metasrv_addr
-        );
-        assert_eq!(5000, options.meta_client_opts.connect_timeout_millis);
-        assert_eq!(3000, options.meta_client_opts.timeout_millis);
-        assert!(!options.meta_client_opts.tcp_nodelay);
+        let MetaClientOpts {
+            metasrv_addrs: metasrv_addr,
+            timeout_millis,
+            connect_timeout_millis,
+            tcp_nodelay,
+        } = options.meta_client_opts.unwrap();
+
+        assert_eq!(vec!["127.0.0.1:3002".to_string()], metasrv_addr);
+        assert_eq!(5000, connect_timeout_millis);
+        assert_eq!(3000, timeout_millis);
+        assert!(!tcp_nodelay);
 
         match options.storage {
             ObjectStoreConfig::File { data_dir } => {
                 assert_eq!("/tmp/greptimedb/data/".to_string(), data_dir)
             }
+            ObjectStoreConfig::S3 { .. } => unreachable!(),
         };
     }
 
@@ -168,43 +179,59 @@ mod tests {
     fn test_try_from_cmd() {
         assert_eq!(
             Mode::Standalone,
-            DatanodeOptions::try_from(StartCommand {
-                node_id: None,
-                rpc_addr: None,
-                mysql_addr: None,
-                metasrv_addr: None,
-                config_file: None
-            })
-            .unwrap()
-            .mode
+            DatanodeOptions::try_from(StartCommand::default())
+                .unwrap()
+                .mode
         );
 
         let mode = DatanodeOptions::try_from(StartCommand {
             node_id: Some(42),
-            rpc_addr: None,
-            mysql_addr: None,
             metasrv_addr: Some("127.0.0.1:3002".to_string()),
-            config_file: None,
+            ..Default::default()
         })
         .unwrap()
         .mode;
         assert_matches!(mode, Mode::Distributed);
 
         assert!(DatanodeOptions::try_from(StartCommand {
-            node_id: None,
-            rpc_addr: None,
-            mysql_addr: None,
             metasrv_addr: Some("127.0.0.1:3002".to_string()),
-            config_file: None,
+            ..Default::default()
         })
         .is_err());
-        assert!(DatanodeOptions::try_from(StartCommand {
+
+        // Providing node_id but leave metasrv_addr absent is ok since metasrv_addr has default value
+        DatanodeOptions::try_from(StartCommand {
             node_id: Some(42),
-            rpc_addr: None,
-            mysql_addr: None,
-            metasrv_addr: None,
-            config_file: None,
+            ..Default::default()
         })
-        .is_err());
+        .unwrap();
+    }
+
+    #[test]
+    fn test_merge_config() {
+        let dn_opts = DatanodeOptions::try_from(StartCommand {
+            config_file: Some(format!(
+                "{}/../../config/datanode.example.toml",
+                std::env::current_dir().unwrap().as_path().to_str().unwrap()
+            )),
+            ..Default::default()
+        })
+        .unwrap();
+        assert_eq!("/tmp/greptimedb/wal", dn_opts.wal.dir);
+        assert_eq!(Duration::from_secs(600), dn_opts.wal.purge_interval);
+        assert_eq!(1024 * 1024 * 1024, dn_opts.wal.file_size.0);
+        assert_eq!(1024 * 1024 * 1024 * 50, dn_opts.wal.purge_threshold.0);
+        assert!(!dn_opts.wal.sync_write);
+        assert_eq!(Some(42), dn_opts.node_id);
+        let MetaClientOpts {
+            metasrv_addrs: metasrv_addr,
+            timeout_millis,
+            connect_timeout_millis,
+            tcp_nodelay,
+        } = dn_opts.meta_client_opts.unwrap();
+        assert_eq!(vec!["127.0.0.1:3002".to_string()], metasrv_addr);
+        assert_eq!(3000, timeout_millis);
+        assert_eq!(5000, connect_timeout_millis);
+        assert!(!tcp_nodelay);
     }
 }

src/cmd/src/error.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -25,12 +25,6 @@ pub enum Error {
         source: datanode::error::Error,
     },
 
-    #[snafu(display("Failed to build frontend, source: {}", source))]
-    BuildFrontend {
-        #[snafu(backtrace)]
-        source: frontend::error::Error,
-    },
-
     #[snafu(display("Failed to start frontend, source: {}", source))]
     StartFrontend {
         #[snafu(backtrace)]
@@ -61,6 +55,12 @@ pub enum Error {
 
     #[snafu(display("Illegal config: {}", msg))]
     IllegalConfig { msg: String, backtrace: Backtrace },
+
+    #[snafu(display("Illegal auth config: {}", source))]
+    IllegalAuthConfig {
+        #[snafu(backtrace)]
+        source: servers::auth::Error,
+    },
 }
 
 pub type Result<T> = std::result::Result<T, Error>;
@@ -75,7 +75,7 @@ impl ErrorExt for Error {
                 StatusCode::InvalidArguments
             }
             Error::IllegalConfig { .. } => StatusCode::InvalidArguments,
-            Error::BuildFrontend { source, .. } => source.status_code(),
+            Error::IllegalAuthConfig { .. } => StatusCode::InvalidArguments,
         }
     }
 
@@ -97,10 +97,7 @@ mod tests {
     #[test]
     fn test_start_node_error() {
         fn throw_datanode_error() -> StdResult<datanode::error::Error> {
-            datanode::error::MissingFieldSnafu {
-                field: "test_field",
-            }
-            .fail()
+            datanode::error::MissingNodeIdSnafu {}.fail()
         }
 
         let e = throw_datanode_error()

src/cmd/src/frontend.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,17 +12,25 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+use std::sync::Arc;
+
 use clap::Parser;
-use frontend::frontend::{Frontend, FrontendOptions, Mode};
+use frontend::frontend::{Frontend, FrontendOptions};
 use frontend::grpc::GrpcOptions;
 use frontend::influxdb::InfluxdbOptions;
 use frontend::instance::Instance;
 use frontend::mysql::MysqlOptions;
 use frontend::opentsdb::OpentsdbOptions;
 use frontend::postgres::PostgresOptions;
+use frontend::Plugins;
+use meta_client::MetaClientOpts;
+use servers::auth::UserProviderRef;
+use servers::http::HttpOptions;
+use servers::tls::{TlsMode, TlsOption};
+use servers::{auth, Mode};
 use snafu::ResultExt;
 
-use crate::error::{self, Result};
+use crate::error::{self, IllegalAuthConfigSnafu, Result};
 use crate::toml_loader;
 
 #[derive(Parser)]
@@ -68,21 +76,41 @@ pub struct StartCommand {
     influxdb_enable: Option<bool>,
     #[clap(long)]
     metasrv_addr: Option<String>,
+    #[clap(long)]
+    tls_mode: Option<TlsMode>,
+    #[clap(long)]
+    tls_cert_path: Option<String>,
+    #[clap(long)]
+    tls_key_path: Option<String>,
+    #[clap(long)]
+    user_provider: Option<String>,
 }
 
 impl StartCommand {
     async fn run(self) -> Result<()> {
+        let plugins = Arc::new(load_frontend_plugins(&self.user_provider)?);
         let opts: FrontendOptions = self.try_into()?;
-        let mut frontend = Frontend::new(
-            opts.clone(),
-            Instance::try_new(&opts)
-                .await
-                .context(error::StartFrontendSnafu)?,
-        );
+
+        let mut instance = Instance::try_new_distributed(&opts)
+            .await
+            .context(error::StartFrontendSnafu)?;
+        instance.set_plugins(plugins.clone());
+
+        let mut frontend = Frontend::new(opts, instance, plugins);
         frontend.start().await.context(error::StartFrontendSnafu)
     }
 }
 
+pub fn load_frontend_plugins(user_provider: &Option<String>) -> Result<Plugins> {
+    let mut plugins = Plugins::new();
+
+    if let Some(provider) = user_provider {
+        let provider = auth::user_provider_from_option(provider).context(IllegalAuthConfigSnafu)?;
+        plugins.insert::<UserProviderRef>(provider);
+    }
+    Ok(plugins)
+}
+
 impl TryFrom<StartCommand> for FrontendOptions {
     type Error = error::Error;
 
@@ -93,8 +121,13 @@ impl TryFrom<StartCommand> for FrontendOptions {
             FrontendOptions::default()
         };
 
+        let tls_option = TlsOption::new(cmd.tls_mode, cmd.tls_cert_path, cmd.tls_key_path);
+
         if let Some(addr) = cmd.http_addr {
-            opts.http_addr = Some(addr);
+            opts.http_options = Some(HttpOptions {
+                addr,
+                ..Default::default()
+            });
         }
         if let Some(addr) = cmd.grpc_addr {
             opts.grpc_options = Some(GrpcOptions {
@@ -105,12 +138,14 @@ impl TryFrom<StartCommand> for FrontendOptions {
         if let Some(addr) = cmd.mysql_addr {
             opts.mysql_options = Some(MysqlOptions {
                 addr,
+                tls: tls_option.clone(),
                 ..Default::default()
             });
         }
         if let Some(addr) = cmd.postgres_addr {
             opts.postgres_options = Some(PostgresOptions {
                 addr,
+                tls: tls_option,
                 ..Default::default()
             });
         }
@@ -124,13 +159,13 @@ impl TryFrom<StartCommand> for FrontendOptions {
             opts.influxdb_options = Some(InfluxdbOptions { enable });
         }
         if let Some(metasrv_addr) = cmd.metasrv_addr {
-            opts.metasrv_addr = Some(
-                metasrv_addr
-                    .split(',')
-                    .into_iter()
-                    .map(|x| x.trim().to_string())
-                    .collect::<Vec<String>>(),
-            );
+            opts.meta_client_opts
+                .get_or_insert_with(MetaClientOpts::default)
+                .metasrv_addrs = metasrv_addr
+                .split(',')
+                .map(&str::trim)
+                .map(&str::to_string)
+                .collect::<Vec<_>>();
             opts.mode = Mode::Distributed;
         }
         Ok(opts)
@@ -139,6 +174,10 @@ impl TryFrom<StartCommand> for FrontendOptions {
 
 #[cfg(test)]
 mod tests {
+    use std::time::Duration;
+
+    use servers::auth::{Identity, Password, UserProviderRef};
+
     use super::*;
 
     #[test]
@@ -152,10 +191,14 @@ mod tests {
             influxdb_enable: Some(false),
             config_file: None,
             metasrv_addr: None,
+            tls_mode: None,
+            tls_cert_path: None,
+            tls_key_path: None,
+            user_provider: None,
         };
 
         let opts: FrontendOptions = command.try_into().unwrap();
-        assert_eq!(opts.http_addr, Some("127.0.0.1:1234".to_string()));
+        assert_eq!(opts.http_options.as_ref().unwrap().addr, "127.0.0.1:1234");
         assert_eq!(opts.mysql_options.as_ref().unwrap().addr, "127.0.0.1:5678");
         assert_eq!(
             opts.postgres_options.as_ref().unwrap().addr,
@@ -186,4 +229,66 @@ mod tests {
 
         assert!(!opts.influxdb_options.unwrap().enable);
     }
+
+    #[test]
+    fn test_read_from_config_file() {
+        let command = StartCommand {
+            http_addr: None,
+            grpc_addr: None,
+            mysql_addr: None,
+            postgres_addr: None,
+            opentsdb_addr: None,
+            influxdb_enable: None,
+            config_file: Some(format!(
+                "{}/../../config/frontend.example.toml",
+                std::env::current_dir().unwrap().as_path().to_str().unwrap()
+            )),
+            metasrv_addr: None,
+            tls_mode: None,
+            tls_cert_path: None,
+            tls_key_path: None,
+            user_provider: None,
+        };
+
+        let fe_opts = FrontendOptions::try_from(command).unwrap();
+        assert_eq!(Mode::Distributed, fe_opts.mode);
+        assert_eq!(
+            "127.0.0.1:4000".to_string(),
+            fe_opts.http_options.as_ref().unwrap().addr
+        );
+        assert_eq!(
+            Duration::from_secs(30),
+            fe_opts.http_options.as_ref().unwrap().timeout
+        );
+    }
+
+    #[tokio::test]
+    async fn test_try_from_start_command_to_anymap() {
+        let command = StartCommand {
+            http_addr: None,
+            grpc_addr: None,
+            mysql_addr: None,
+            postgres_addr: None,
+            opentsdb_addr: None,
+            influxdb_enable: None,
+            config_file: None,
+            metasrv_addr: None,
+            tls_mode: None,
+            tls_cert_path: None,
+            tls_key_path: None,
+            user_provider: Some("static_user_provider:cmd:test=test".to_string()),
+        };
+
+        let plugins = load_frontend_plugins(&command.user_provider);
+        assert!(plugins.is_ok());
+        let plugins = plugins.unwrap();
+        let provider = plugins.get::<UserProviderRef>();
+        assert!(provider.is_some());
+
+        let provider = provider.unwrap();
+        let result = provider
+            .auth(Identity::UserId("test", None), Password::PlainText("test"))
+            .await;
+        assert!(result.is_ok());
+    }
 }

src/cmd/src/lib.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/cmd/src/metasrv.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -104,13 +104,13 @@ mod tests {
     fn test_read_from_cmd() {
         let cmd = StartCommand {
             bind_addr: Some("127.0.0.1:3002".to_string()),
-            server_addr: Some("0.0.0.0:3002".to_string()),
+            server_addr: Some("127.0.0.1:3002".to_string()),
             store_addr: Some("127.0.0.1:2380".to_string()),
             config_file: None,
         };
         let options: MetaSrvOptions = cmd.try_into().unwrap();
         assert_eq!("127.0.0.1:3002".to_string(), options.bind_addr);
-        assert_eq!("0.0.0.0:3002".to_string(), options.server_addr);
+        assert_eq!("127.0.0.1:3002".to_string(), options.server_addr);
         assert_eq!("127.0.0.1:2380".to_string(), options.store_addr);
     }
 
@@ -127,8 +127,8 @@ mod tests {
         };
         let options: MetaSrvOptions = cmd.try_into().unwrap();
         assert_eq!("127.0.0.1:3002".to_string(), options.bind_addr);
-        assert_eq!("0.0.0.0:3002".to_string(), options.server_addr);
-        assert_eq!("127.0.0.1:2380".to_string(), options.store_addr);
-        assert_eq!(30, options.datanode_lease_secs);
+        assert_eq!("127.0.0.1:3002".to_string(), options.server_addr);
+        assert_eq!("127.0.0.1:2379".to_string(), options.store_addr);
+        assert_eq!(15, options.datanode_lease_secs);
     }
 }

src/cmd/src/standalone.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,11 +12,13 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+use std::sync::Arc;
+
 use clap::Parser;
 use common_telemetry::info;
-use datanode::datanode::{Datanode, DatanodeOptions, ObjectStoreConfig};
+use datanode::datanode::{Datanode, DatanodeOptions, ObjectStoreConfig, WalConfig};
 use datanode::instance::InstanceRef;
-use frontend::frontend::{Frontend, FrontendOptions, Mode};
+use frontend::frontend::{Frontend, FrontendOptions};
 use frontend::grpc::GrpcOptions;
 use frontend::influxdb::InfluxdbOptions;
 use frontend::instance::Instance as FeInstance;
@@ -24,13 +26,15 @@ use frontend::mysql::MysqlOptions;
 use frontend::opentsdb::OpentsdbOptions;
 use frontend::postgres::PostgresOptions;
 use frontend::prometheus::PrometheusOptions;
+use frontend::Plugins;
 use serde::{Deserialize, Serialize};
+use servers::http::HttpOptions;
+use servers::tls::{TlsMode, TlsOption};
+use servers::Mode;
 use snafu::ResultExt;
-use tokio::try_join;
 
-use crate::error::{
-    BuildFrontendSnafu, Error, IllegalConfigSnafu, Result, StartDatanodeSnafu, StartFrontendSnafu,
-};
+use crate::error::{Error, IllegalConfigSnafu, Result, StartDatanodeSnafu, StartFrontendSnafu};
+use crate::frontend::load_frontend_plugins;
 use crate::toml_loader;
 
 #[derive(Parser)]
@@ -59,8 +63,9 @@ impl SubCommand {
 }
 
 #[derive(Clone, Debug, Serialize, Deserialize)]
+#[serde(default)]
 pub struct StandaloneOptions {
-    pub http_addr: Option<String>,
+    pub http_options: Option<HttpOptions>,
     pub grpc_options: Option<GrpcOptions>,
     pub mysql_options: Option<MysqlOptions>,
     pub postgres_options: Option<PostgresOptions>,
@@ -68,16 +73,15 @@ pub struct StandaloneOptions {
     pub influxdb_options: Option<InfluxdbOptions>,
     pub prometheus_options: Option<PrometheusOptions>,
     pub mode: Mode,
-    pub wal_dir: String,
+    pub wal: WalConfig,
     pub storage: ObjectStoreConfig,
-    pub datanode_mysql_addr: String,
-    pub datanode_mysql_runtime_size: usize,
+    pub enable_memory_catalog: bool,
 }
 
 impl Default for StandaloneOptions {
     fn default() -> Self {
         Self {
-            http_addr: Some("0.0.0.0:4000".to_string()),
+            http_options: Some(HttpOptions::default()),
             grpc_options: Some(GrpcOptions::default()),
             mysql_options: Some(MysqlOptions::default()),
             postgres_options: Some(PostgresOptions::default()),
@@ -85,10 +89,9 @@ impl Default for StandaloneOptions {
             influxdb_options: Some(InfluxdbOptions::default()),
             prometheus_options: Some(PrometheusOptions::default()),
             mode: Mode::Standalone,
-            wal_dir: "/tmp/greptimedb/wal".to_string(),
+            wal: WalConfig::default(),
             storage: ObjectStoreConfig::default(),
-            datanode_mysql_addr: "0.0.0.0:3306".to_string(),
-            datanode_mysql_runtime_size: 4,
+            enable_memory_catalog: false,
         }
     }
 }
@@ -96,7 +99,7 @@ impl Default for StandaloneOptions {
 impl StandaloneOptions {
     fn frontend_options(self) -> FrontendOptions {
         FrontendOptions {
-            http_addr: self.http_addr,
+            http_options: self.http_options,
             grpc_options: self.grpc_options,
             mysql_options: self.mysql_options,
             postgres_options: self.postgres_options,
@@ -104,17 +107,15 @@ impl StandaloneOptions {
             influxdb_options: self.influxdb_options,
             prometheus_options: self.prometheus_options,
             mode: self.mode,
-            datanode_rpc_addr: "127.0.0.1:3001".to_string(),
-            metasrv_addr: None,
+            meta_client_opts: None,
         }
     }
 
     fn datanode_options(self) -> DatanodeOptions {
         DatanodeOptions {
-            wal_dir: self.wal_dir,
+            wal: self.wal,
             storage: self.storage,
-            mysql_addr: self.datanode_mysql_addr,
-            mysql_runtime_size: self.datanode_mysql_runtime_size,
+            enable_memory_catalog: self.enable_memory_catalog,
             ..Default::default()
         }
     }
@@ -136,18 +137,31 @@ struct StartCommand {
     influxdb_enable: bool,
     #[clap(short, long)]
     config_file: Option<String>,
+    #[clap(short = 'm', long = "memory-catalog")]
+    enable_memory_catalog: bool,
+    #[clap(long)]
+    tls_mode: Option<TlsMode>,
+    #[clap(long)]
+    tls_cert_path: Option<String>,
+    #[clap(long)]
+    tls_key_path: Option<String>,
+    #[clap(long)]
+    user_provider: Option<String>,
 }
 
 impl StartCommand {
     async fn run(self) -> Result<()> {
+        let enable_memory_catalog = self.enable_memory_catalog;
         let config_file = self.config_file.clone();
+        let plugins = Arc::new(load_frontend_plugins(&self.user_provider)?);
         let fe_opts = FrontendOptions::try_from(self)?;
         let dn_opts: DatanodeOptions = {
-            let opts: StandaloneOptions = if let Some(path) = config_file {
+            let mut opts: StandaloneOptions = if let Some(path) = config_file {
                 toml_loader::from_file!(&path)?
             } else {
                 StandaloneOptions::default()
             };
+            opts.enable_memory_catalog = enable_memory_catalog;
             opts.datanode_options()
         };
 
@@ -159,13 +173,16 @@ impl StartCommand {
         let mut datanode = Datanode::new(dn_opts.clone())
             .await
             .context(StartDatanodeSnafu)?;
-        let mut frontend = build_frontend(fe_opts, &dn_opts, datanode.get_instance()).await?;
+        let mut frontend = build_frontend(fe_opts, plugins, datanode.get_instance()).await?;
 
-        try_join!(
-            async { datanode.start().await.context(StartDatanodeSnafu) },
-            async { frontend.start().await.context(StartFrontendSnafu) }
-        )?;
+        // Start datanode instance before starting services, to avoid requests come in before internal components are started.
+        datanode
+            .start_instance()
+            .await
+            .context(StartDatanodeSnafu)?;
+        info!("Datanode instance started");
 
+        frontend.start().await.context(StartFrontendSnafu)?;
         Ok(())
     }
 }
@@ -173,20 +190,13 @@ impl StartCommand {
 /// Build frontend instance in standalone mode
 async fn build_frontend(
     fe_opts: FrontendOptions,
-    dn_opts: &DatanodeOptions,
+    plugins: Arc<Plugins>,
     datanode_instance: InstanceRef,
 ) -> Result<Frontend<FeInstance>> {
-    let grpc_server_addr = &dn_opts.rpc_addr;
-    info!(
-        "Build frontend with datanode gRPC addr: {}",
-        grpc_server_addr
-    );
-    let mut frontend_instance = FeInstance::try_new(&fe_opts)
-        .await
-        .context(BuildFrontendSnafu)?;
-    frontend_instance.set_catalog_manager(datanode_instance.catalog_manager().clone());
+    let mut frontend_instance = FeInstance::new_standalone(datanode_instance.clone());
     frontend_instance.set_script_handler(datanode_instance);
-    Ok(Frontend::new(fe_opts, frontend_instance))
+    frontend_instance.set_plugins(plugins.clone());
+    Ok(Frontend::new(fe_opts, frontend_instance, plugins))
 }
 
 impl TryFrom<StartCommand> for FrontendOptions {
@@ -204,7 +214,10 @@ impl TryFrom<StartCommand> for FrontendOptions {
         opts.mode = Mode::Standalone;
 
         if let Some(addr) = cmd.http_addr {
-            opts.http_addr = Some(addr);
+            opts.http_options = Some(HttpOptions {
+                addr,
+                ..Default::default()
+            });
         }
         if let Some(addr) = cmd.rpc_addr {
             // frontend grpc addr conflict with datanode default grpc addr
@@ -212,8 +225,7 @@ impl TryFrom<StartCommand> for FrontendOptions {
             if addr == datanode_grpc_addr {
                 return IllegalConfigSnafu {
                     msg: format!(
-                        "gRPC listen address conflicts with datanode reserved gRPC addr: {}",
-                        datanode_grpc_addr
+                        "gRPC listen address conflicts with datanode reserved gRPC addr: {datanode_grpc_addr}",
                     ),
                 }
                 .fail();
@@ -248,12 +260,28 @@ impl TryFrom<StartCommand> for FrontendOptions {
             opts.influxdb_options = Some(InfluxdbOptions { enable: true });
         }
 
+        let tls_option = TlsOption::new(cmd.tls_mode, cmd.tls_cert_path, cmd.tls_key_path);
+
+        if let Some(mut mysql_options) = opts.mysql_options {
+            mysql_options.tls = tls_option.clone();
+            opts.mysql_options = Some(mysql_options);
+        }
+
+        if let Some(mut postgres_options) = opts.postgres_options {
+            postgres_options.tls = tls_option;
+            opts.postgres_options = Some(postgres_options);
+        }
+
         Ok(opts)
     }
 }
 
 #[cfg(test)]
 mod tests {
+    use std::time::Duration;
+
+    use servers::auth::{Identity, Password, UserProviderRef};
+
     use super::*;
 
     #[test]
@@ -269,18 +297,61 @@ mod tests {
                 std::env::current_dir().unwrap().as_path().to_str().unwrap()
             )),
             influxdb_enable: false,
+            enable_memory_catalog: false,
+            tls_mode: None,
+            tls_cert_path: None,
+            tls_key_path: None,
+            user_provider: None,
         };
 
         let fe_opts = FrontendOptions::try_from(cmd).unwrap();
         assert_eq!(Mode::Standalone, fe_opts.mode);
-        assert_eq!("127.0.0.1:3001".to_string(), fe_opts.datanode_rpc_addr);
-        assert_eq!(Some("0.0.0.0:4000".to_string()), fe_opts.http_addr);
         assert_eq!(
-            "0.0.0.0:4001".to_string(),
+            "127.0.0.1:4000".to_string(),
+            fe_opts.http_options.as_ref().unwrap().addr
+        );
+        assert_eq!(
+            Duration::from_secs(30),
+            fe_opts.http_options.as_ref().unwrap().timeout
+        );
+        assert_eq!(
+            "127.0.0.1:4001".to_string(),
             fe_opts.grpc_options.unwrap().addr
         );
-        assert_eq!("0.0.0.0:4002", fe_opts.mysql_options.as_ref().unwrap().addr);
+        assert_eq!(
+            "127.0.0.1:4002",
+            fe_opts.mysql_options.as_ref().unwrap().addr
+        );
         assert_eq!(2, fe_opts.mysql_options.as_ref().unwrap().runtime_size);
         assert!(fe_opts.influxdb_options.as_ref().unwrap().enable);
     }
+
+    #[tokio::test]
+    async fn test_try_from_start_command_to_anymap() {
+        let command = StartCommand {
+            http_addr: None,
+            rpc_addr: None,
+            mysql_addr: None,
+            postgres_addr: None,
+            opentsdb_addr: None,
+            config_file: None,
+            influxdb_enable: false,
+            enable_memory_catalog: false,
+            tls_mode: None,
+            tls_cert_path: None,
+            tls_key_path: None,
+            user_provider: Some("static_user_provider:cmd:test=test".to_string()),
+        };
+
+        let plugins = load_frontend_plugins(&command.user_provider);
+        assert!(plugins.is_ok());
+        let plugins = plugins.unwrap();
+        let provider = plugins.get::<UserProviderRef>();
+        assert!(provider.is_some());
+        let provider = provider.unwrap();
+        let result = provider
+            .auth(Identity::UserId("test", None), Password::PlainText("test"))
+            .await;
+        assert!(result.is_ok());
+    }
 }

src/cmd/src/toml_loader.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -37,12 +37,23 @@ mod tests {
     use crate::error::Result;
 
     #[derive(Clone, PartialEq, Debug, Deserialize, Serialize)]
+    #[serde(default)]
     struct MockConfig {
         path: String,
         port: u32,
         host: String,
     }
 
+    impl Default for MockConfig {
+        fn default() -> Self {
+            Self {
+                path: "test".to_string(),
+                port: 0,
+                host: "localhost".to_string(),
+            }
+        }
+    }
+
     #[test]
     fn test_from_file() -> Result<()> {
         let config = MockConfig {
@@ -63,6 +74,21 @@ mod tests {
         let loaded_config: MockConfig = from_file!(&test_file)?;
         assert_eq!(loaded_config, config);
 
+        // Only host in file
+        let mut file = File::create(&test_file).unwrap();
+        file.write_all("host='greptime.test'\n".as_bytes()).unwrap();
+
+        let loaded_config: MockConfig = from_file!(&test_file)?;
+        assert_eq!(loaded_config.host, "greptime.test");
+        assert_eq!(loaded_config.port, 0);
+        assert_eq!(loaded_config.path, "test");
+
+        // Truncate the file.
+        let file = File::create(&test_file).unwrap();
+        file.set_len(0).unwrap();
+        let loaded_config: MockConfig = from_file!(&test_file)?;
+        assert_eq!(loaded_config, MockConfig::default());
+
         Ok(())
     }
 }

src/common/base/Cargo.toml

@@ -1,8 +1,8 @@
 [package]
 name = "common-base"
-version = "0.1.0"
-edition = "2021"
-license = "Apache-2.0"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 
 [dependencies]
 bitvec = "1.0"
@@ -10,4 +10,7 @@ bytes = { version = "1.1", features = ["serde"] }
 common-error = { path = "../error" }
 paste = "1.0"
 serde = { version = "1.0", features = ["derive"] }
-snafu = { version = "0.7", features = ["backtraces"] }
+snafu.workspace = true
+
+[dev-dependencies]
+toml = "0.5"

src/common/base/src/bit_vec.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,8 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use bitvec::prelude as bv;
+pub use bitvec::prelude;
 
 // `Lsb0` provides the best codegen for bit manipulation,
 // see https://github.com/bitvecto-rs/bitvec/blob/main/doc/order/Lsb0.md
-pub type BitVec = bv::BitVec<u8, bv::Lsb0>;
+pub type BitVec = prelude::BitVec<u8>;

src/common/base/src/buffer.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/common/base/src/bytes.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,

src/common/base/src/lib.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,
@@ -12,8 +12,10 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-pub mod bitset;
+pub mod bit_vec;
 pub mod buffer;
 pub mod bytes;
+#[allow(clippy::all)]
+pub mod readable_size;
 
-pub use bitset::BitVec;
+pub use bit_vec::BitVec;

src/common/base/src/readable_size.rs

@@ -0,0 +1,321 @@
+// Copyright (c) 2017-present, PingCAP, Inc. Licensed under Apache-2.0.
+
+// 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.
+
+// This file is copied from https://github.com/tikv/raft-engine/blob/8dd2a39f359ff16f5295f35343f626e0c10132fa/src/util.rs without any modification.
+
+use std::fmt;
+use std::fmt::{Display, Write};
+use std::ops::{Div, Mul};
+use std::str::FromStr;
+
+use serde::de::{Unexpected, Visitor};
+use serde::{de, Deserialize, Deserializer, Serialize, Serializer};
+
+const UNIT: u64 = 1;
+
+const BINARY_DATA_MAGNITUDE: u64 = 1024;
+pub const B: u64 = UNIT;
+pub const KIB: u64 = B * BINARY_DATA_MAGNITUDE;
+pub const MIB: u64 = KIB * BINARY_DATA_MAGNITUDE;
+pub const GIB: u64 = MIB * BINARY_DATA_MAGNITUDE;
+pub const TIB: u64 = GIB * BINARY_DATA_MAGNITUDE;
+pub const PIB: u64 = TIB * BINARY_DATA_MAGNITUDE;
+
+#[derive(Clone, Debug, Copy, PartialEq, Eq, PartialOrd)]
+pub struct ReadableSize(pub u64);
+
+impl ReadableSize {
+    pub const fn kb(count: u64) -> ReadableSize {
+        ReadableSize(count * KIB)
+    }
+
+    pub const fn mb(count: u64) -> ReadableSize {
+        ReadableSize(count * MIB)
+    }
+
+    pub const fn gb(count: u64) -> ReadableSize {
+        ReadableSize(count * GIB)
+    }
+
+    pub const fn as_mb(self) -> u64 {
+        self.0 / MIB
+    }
+}
+
+impl Div<u64> for ReadableSize {
+    type Output = ReadableSize;
+
+    fn div(self, rhs: u64) -> ReadableSize {
+        ReadableSize(self.0 / rhs)
+    }
+}
+
+impl Div<ReadableSize> for ReadableSize {
+    type Output = u64;
+
+    fn div(self, rhs: ReadableSize) -> u64 {
+        self.0 / rhs.0
+    }
+}
+
+impl Mul<u64> for ReadableSize {
+    type Output = ReadableSize;
+
+    fn mul(self, rhs: u64) -> ReadableSize {
+        ReadableSize(self.0 * rhs)
+    }
+}
+
+impl Serialize for ReadableSize {
+    fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
+    where
+        S: Serializer,
+    {
+        let size = self.0;
+        let mut buffer = String::new();
+        if size == 0 {
+            write!(buffer, "{}KiB", size).unwrap();
+        } else if size % PIB == 0 {
+            write!(buffer, "{}PiB", size / PIB).unwrap();
+        } else if size % TIB == 0 {
+            write!(buffer, "{}TiB", size / TIB).unwrap();
+        } else if size % GIB as u64 == 0 {
+            write!(buffer, "{}GiB", size / GIB).unwrap();
+        } else if size % MIB as u64 == 0 {
+            write!(buffer, "{}MiB", size / MIB).unwrap();
+        } else if size % KIB as u64 == 0 {
+            write!(buffer, "{}KiB", size / KIB).unwrap();
+        } else {
+            return serializer.serialize_u64(size);
+        }
+        serializer.serialize_str(&buffer)
+    }
+}
+
+impl FromStr for ReadableSize {
+    type Err = String;
+
+    // This method parses value in binary unit.
+    fn from_str(s: &str) -> Result<ReadableSize, String> {
+        let size_str = s.trim();
+        if size_str.is_empty() {
+            return Err(format!("{:?} is not a valid size.", s));
+        }
+
+        if !size_str.is_ascii() {
+            return Err(format!("ASCII string is expected, but got {:?}", s));
+        }
+
+        // size: digits and '.' as decimal separator
+        let size_len = size_str
+            .to_string()
+            .chars()
+            .take_while(|c| char::is_ascii_digit(c) || ['.', 'e', 'E', '-', '+'].contains(c))
+            .count();
+
+        // unit: alphabetic characters
+        let (size, unit) = size_str.split_at(size_len);
+
+        let unit = match unit.trim() {
+            "K" | "KB" | "KiB" => KIB,
+            "M" | "MB" | "MiB" => MIB,
+            "G" | "GB" | "GiB" => GIB,
+            "T" | "TB" | "TiB" => TIB,
+            "P" | "PB" | "PiB" => PIB,
+            "B" | "" => B,
+            _ => {
+                return Err(format!(
+                    "only B, KB, KiB, MB, MiB, GB, GiB, TB, TiB, PB, and PiB are supported: {:?}",
+                    s
+                ));
+            }
+        };
+
+        match size.parse::<f64>() {
+            Ok(n) => Ok(ReadableSize((n * unit as f64) as u64)),
+            Err(_) => Err(format!("invalid size string: {:?}", s)),
+        }
+    }
+}
+
+impl Display for ReadableSize {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        if self.0 >= PIB {
+            write!(f, "{:.1}PiB", self.0 as f64 / PIB as f64)
+        } else if self.0 >= TIB {
+            write!(f, "{:.1}TiB", self.0 as f64 / TIB as f64)
+        } else if self.0 >= GIB {
+            write!(f, "{:.1}GiB", self.0 as f64 / GIB as f64)
+        } else if self.0 >= MIB {
+            write!(f, "{:.1}MiB", self.0 as f64 / MIB as f64)
+        } else if self.0 >= KIB {
+            write!(f, "{:.1}KiB", self.0 as f64 / KIB as f64)
+        } else {
+            write!(f, "{}B", self.0)
+        }
+    }
+}
+
+impl<'de> Deserialize<'de> for ReadableSize {
+    fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
+    where
+        D: Deserializer<'de>,
+    {
+        struct SizeVisitor;
+
+        impl<'de> Visitor<'de> for SizeVisitor {
+            type Value = ReadableSize;
+
+            fn expecting(&self, formatter: &mut fmt::Formatter<'_>) -> fmt::Result {
+                formatter.write_str("valid size")
+            }
+
+            fn visit_i64<E>(self, size: i64) -> Result<ReadableSize, E>
+            where
+                E: de::Error,
+            {
+                if size >= 0 {
+                    self.visit_u64(size as u64)
+                } else {
+                    Err(E::invalid_value(Unexpected::Signed(size), &self))
+                }
+            }
+
+            fn visit_u64<E>(self, size: u64) -> Result<ReadableSize, E>
+            where
+                E: de::Error,
+            {
+                Ok(ReadableSize(size))
+            }
+
+            fn visit_str<E>(self, size_str: &str) -> Result<ReadableSize, E>
+            where
+                E: de::Error,
+            {
+                size_str.parse().map_err(E::custom)
+            }
+        }
+
+        deserializer.deserialize_any(SizeVisitor)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_readable_size() {
+        let s = ReadableSize::kb(2);
+        assert_eq!(s.0, 2048);
+        assert_eq!(s.as_mb(), 0);
+        let s = ReadableSize::mb(2);
+        assert_eq!(s.0, 2 * 1024 * 1024);
+        assert_eq!(s.as_mb(), 2);
+        let s = ReadableSize::gb(2);
+        assert_eq!(s.0, 2 * 1024 * 1024 * 1024);
+        assert_eq!(s.as_mb(), 2048);
+
+        assert_eq!((ReadableSize::mb(2) / 2).0, MIB);
+        assert_eq!((ReadableSize::mb(1) / 2).0, 512 * KIB);
+        assert_eq!(ReadableSize::mb(2) / ReadableSize::kb(1), 2048);
+    }
+
+    #[test]
+    fn test_parse_readable_size() {
+        #[derive(Serialize, Deserialize)]
+        struct SizeHolder {
+            s: ReadableSize,
+        }
+
+        let legal_cases = vec![
+            (0, "0KiB"),
+            (2 * KIB, "2KiB"),
+            (4 * MIB, "4MiB"),
+            (5 * GIB, "5GiB"),
+            (7 * TIB, "7TiB"),
+            (11 * PIB, "11PiB"),
+        ];
+        for (size, exp) in legal_cases {
+            let c = SizeHolder {
+                s: ReadableSize(size),
+            };
+            let res_str = toml::to_string(&c).unwrap();
+            let exp_str = format!("s = {:?}\n", exp);
+            assert_eq!(res_str, exp_str);
+            let res_size: SizeHolder = toml::from_str(&exp_str).unwrap();
+            assert_eq!(res_size.s.0, size);
+        }
+
+        let c = SizeHolder {
+            s: ReadableSize(512),
+        };
+        let res_str = toml::to_string(&c).unwrap();
+        assert_eq!(res_str, "s = 512\n");
+        let res_size: SizeHolder = toml::from_str(&res_str).unwrap();
+        assert_eq!(res_size.s.0, c.s.0);
+
+        let decode_cases = vec![
+            (" 0.5 PB", PIB / 2),
+            ("0.5 TB", TIB / 2),
+            ("0.5GB ", GIB / 2),
+            ("0.5MB", MIB / 2),
+            ("0.5KB", KIB / 2),
+            ("0.5P", PIB / 2),
+            ("0.5T", TIB / 2),
+            ("0.5G", GIB / 2),
+            ("0.5M", MIB / 2),
+            ("0.5K", KIB / 2),
+            ("23", 23),
+            ("1", 1),
+            ("1024B", KIB),
+            // units with binary prefixes
+            (" 0.5 PiB", PIB / 2),
+            ("1PiB", PIB),
+            ("0.5 TiB", TIB / 2),
+            ("2 TiB", TIB * 2),
+            ("0.5GiB ", GIB / 2),
+            ("787GiB ", GIB * 787),
+            ("0.5MiB", MIB / 2),
+            ("3MiB", MIB * 3),
+            ("0.5KiB", KIB / 2),
+            ("1 KiB", KIB),
+            // scientific notation
+            ("0.5e6 B", B * 500000),
+            ("0.5E6 B", B * 500000),
+            ("1e6B", B * 1000000),
+            ("8E6B", B * 8000000),
+            ("8e7", B * 80000000),
+            ("1e-1MB", MIB / 10),
+            ("1e+1MB", MIB * 10),
+            ("0e+10MB", 0),
+        ];
+        for (src, exp) in decode_cases {
+            let src = format!("s = {:?}", src);
+            let res: SizeHolder = toml::from_str(&src).unwrap();
+            assert_eq!(res.s.0, exp);
+        }
+
+        let illegal_cases = vec![
+            "0.5kb", "0.5kB", "0.5Kb", "0.5k", "0.5g", "b", "gb", "1b", "B", "1K24B", " 5_KB",
+            "4B7", "5M_",
+        ];
+        for src in illegal_cases {
+            let src_str = format!("s = {:?}", src);
+            assert!(toml::from_str::<SizeHolder>(&src_str).is_err(), "{}", src);
+        }
+    }
+}

src/common/base/tests/buffer_tests.rs

@@ -1,10 +1,10 @@
-// Copyright 2022 Greptime Team
+// 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
+//     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,