build: install ca-certificates in docker image building (#807)

refactor: install ca-certificates in docker image building

Signed-off-by: zyy17 <zyylsxm@gmail.com>

Signed-off-by: zyy17 <zyylsxm@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

@@ -1,25 +1,44 @@
 on:
   pull_request:
     types: [opened, synchronize, reopened, ready_for_review]
+    paths-ignore:
+      - 'docs/**'
+      - 'config/**'
+      - '**.md'
+      - '.dockerignore'
+      - 'docker/**'
+      - '.gitignore'
   push:
     branches:
       - "main"
       - "develop"
+    paths-ignore:
+      - 'docs/**'
+      - 'config/**'
+      - '**.md'
+      - '.dockerignore'
+      - 'docker/**'
+      - '.gitignore'
   workflow_dispatch:
 
 name: Code coverage
 
 env:
-  RUST_TOOLCHAIN: nightly-2022-07-14
+  RUST_TOOLCHAIN: nightly-2022-12-20
 
 jobs:
   coverage:
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    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: KyleMayes/install-llvm-action@v1
+        with:
+          version: "14.0"
       - name: Install toolchain
         uses: dtolnay/rust-toolchain@master
         with:
@@ -27,10 +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
@@ -38,6 +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

@@ -1,6 +1,12 @@
 on:
   pull_request:
     types: [opened, synchronize, reopened, ready_for_review]
+    paths-ignore:
+      - 'docs/**'
+      - 'config/**'
+      - '**.md'
+      - '.dockerignore'
+      - 'docker/**'
   push:
     branches:
       - develop
@@ -8,19 +14,25 @@ on:
     paths-ignore:
       - 'docs/**'
       - 'config/**'
-      - '.github/**'
       - '**.md'
-      - '**.yml'
       - '.dockerignore'
       - 'docker/**'
+      - '.gitignore'
   workflow_dispatch:
 
-name: Continuous integration for developing
+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
@@ -29,6 +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 }}
@@ -37,44 +51,64 @@ jobs:
       - name: Run cargo check
         run: cargo check --workspace --all-targets
 
-  test:
-    name: Test Suite
+  toml:
+    name: Toml Check
     if: github.event.pull_request.draft == false
     runs-on: ubuntu-latest
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
-      - name: Cache LLVM and Clang
-        id: cache-llvm
-        uses: actions/cache@v3
-        with:
-          path: ./llvm
-          key: llvm
-      - uses: arduino/setup-protoc@v1
-      - uses: KyleMayes/install-llvm-action@v1
-        with:
-          version: "14.0"
-          cached: ${{ steps.cache-llvm.outputs.cache-hit }}
       - uses: dtolnay/rust-toolchain@master
         with:
           toolchain: ${{ env.RUST_TOOLCHAIN }}
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
-      - name: Cleanup disk
-        uses: curoky/cleanup-disk-action@v2.0
-        with:
-          retain: 'rust,llvm'
-      - name: Install latest nextest release
-        uses: taiki-e/install-action@nextest
-      - name: Run tests
-        run: cargo nextest run
-        env:
-          CARGO_BUILD_RUSTFLAGS: "-C link-arg=-fuse-ld=lld"
-          RUST_BACKTRACE: 1
-          GT_S3_BUCKET: ${{ secrets.S3_BUCKET }}
-          GT_S3_ACCESS_KEY_ID: ${{ secrets.S3_ACCESS_KEY_ID }}
-          GT_S3_ACCESS_KEY: ${{ secrets.S3_ACCESS_KEY }}
-          UNITTEST_LOG_DIR: "__unittest_logs"
+      - 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
+  #   if: github.event.pull_request.draft == false
+  #   runs-on: ubuntu-latest
+  #   timeout-minutes: 60
+  #   steps:
+  #     - uses: actions/checkout@v3
+  #     - name: Cache LLVM and Clang
+  #       id: cache-llvm
+  #       uses: actions/cache@v3
+  #       with:
+  #         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"
+  #         cached: ${{ steps.cache-llvm.outputs.cache-hit }}
+  #     - uses: dtolnay/rust-toolchain@master
+  #       with:
+  #         toolchain: ${{ env.RUST_TOOLCHAIN }}
+  #     - name: Rust Cache
+  #       uses: Swatinem/rust-cache@v2
+  #     - name: Cleanup disk
+  #       uses: curoky/cleanup-disk-action@v2.0
+  #       with:
+  #         retain: 'rust,llvm'
+  #     - name: Install latest nextest release
+  #       uses: taiki-e/install-action@nextest
+  #     - name: Run tests
+  #       run: cargo nextest run
+  #       env:
+  #         CARGO_BUILD_RUSTFLAGS: "-C link-arg=-fuse-ld=lld"
+  #         RUST_BACKTRACE: 1
+  #         GT_S3_BUCKET: ${{ secrets.S3_BUCKET }}
+  #         GT_S3_ACCESS_KEY_ID: ${{ secrets.S3_ACCESS_KEY_ID }}
+  #         GT_S3_ACCESS_KEY: ${{ secrets.S3_ACCESS_KEY }}
+  #         UNITTEST_LOG_DIR: "__unittest_logs"
 
   fmt:
     name: Rustfmt
@@ -84,6 +118,8 @@ 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 }}
@@ -101,6 +137,8 @@ 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/license.yaml

@@ -0,0 +1,16 @@
+name: License checker
+
+on:
+  push:
+    branches:
+    - develop
+  pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
+jobs:
+  license-header-check:
+    runs-on: ubuntu-latest
+    name: license-header-check
+    steps:
+    - uses: actions/checkout@v2
+    - name: Check License Header
+      uses: apache/skywalking-eyes/header@main

.github/workflows/release.yml

@@ -2,12 +2,21 @@ 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
 
 jobs:
   build:
@@ -17,10 +26,10 @@ jobs:
         # The file format is greptime-<os>-<arch>
         include:
           - arch: x86_64-unknown-linux-gnu
-            os: ubuntu-latest
+            os: ubuntu-latest-16-cores
             file: greptime-linux-amd64
           - arch: aarch64-unknown-linux-gnu
-            os: ubuntu-latest
+            os: ubuntu-latest-16-cores
             file: greptime-linux-arm64
           - arch: aarch64-apple-darwin
             os: macos-latest
@@ -106,8 +115,32 @@ 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 +178,12 @@ 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 +191,20 @@ 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 +221,6 @@ 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

.licenserc.yaml

@@ -0,0 +1,14 @@
+header:
+  license:
+    spdx-id: Apache-2.0
+    copyright-owner: Greptime Team
+
+  paths:
+    - "**/*.rs"
+    - "**/*.py"
+
+  comment: on-failure
+
+dependency:
+  files:
+    - Cargo.toml

.pre-commit-config.yaml

@@ -9,7 +9,7 @@ repos:
     rev: e6a795bc6b2c0958f9ef52af4863bbd7cc17238f
     hooks:
     -    id: cargo-sort
-         args: ["--workspace", "--print"]
+         args: ["--workspace"]
 
 -   repo: https://github.com/doublify/pre-commit-rust
     rev: v1.0

CONTRIBUTING.md

@@ -1,12 +1,55 @@
-# Contributing to GreptimeDB
+# Welcome!
 
-Much appreciate for your interest in contributing to GreptimeDB! This document list some guidelines for contributing to our code base.
+Thanks a lot for considering contributing to GreptimeDB. We believe people like you would make GreptimeDB a great product. We intend to build a community where individuals can have open talks, show respect for one another, and speak with true ❤️. Meanwhile, we are to keep transparency and make your effort count here.
 
-To learn about the design of GreptimeDB, please refer to the [design docs](https://github.com/GrepTimeTeam/docs).
+Read the guidelines, and they can help you get started. Communicate with respect to developers maintaining and developing the project. In return, they should reciprocate that respect by addressing your issue, reviewing changes, as well as helping finalize and merge your pull requests.
 
-## Pull Requests
+Follow our [README](https://github.com/GreptimeTeam/greptimedb#readme) to get the whole picture of the project. To learn about the design of GreptimeDB, please refer to the [design docs](https://github.com/GrepTimeTeam/docs).
+
+## Your First Contribution
+
+It can feel intimidating to contribute to a complex project, but it can also be exciting and fun. These general notes will help everyone participate in this communal activity.
+
+- Follow the [Code of Conduct](https://github.com/GreptimeTeam/greptimedb/blob/develop/CODE_OF_CONDUCT.md)
+- Small changes make huge differences. We will happily accept a PR making a single character change if it helps move forward. Don't wait to have everything working.
+- Check the closed issues before opening your issue.
+- Try to follow the existing style of the code.
+- More importantly, when in doubt, ask away.
+
+Pull requests are great, but we accept all kinds of other help if you like. Such as
+
+- Write tutorials or blog posts. Blog, speak about, or create tutorials about one of GreptimeDB's many features. Mention [@greptime](https://twitter.com/greptime) on Twitter and email info@greptime.com so we can give pointers and tips and help you spread the word by promoting your content on Greptime communication channels.
+- Improve the documentation. [Submit documentation](http://github.com/greptimeTeam/docs/) updates, enhancements, designs, or bug fixes, and fixing any spelling or grammar errors will be very much appreciated.
+- Present at meetups and conferences about your GreptimeDB projects. Your unique challenges and successes in building things with GreptimeDB can provide great speaking material. We'd love to review your talk abstract, so get in touch with us if you'd like some help!
+- Submit bug reports. To report a bug or a security issue, you can [open a new GitHub issue](https://github.com/GrepTimeTeam/greptimedb/issues/new).
+- Speak up feature requests. Send feedback is a great way for us to understand your different use cases of GreptimeDB better. If you want to share your experience with GreptimeDB, or if you want to discuss any ideas, you can start a discussion on [GitHub discussions](https://github.com/GreptimeTeam/greptimedb/discussions), chat with the Greptime team on [Slack](https://greptime.com/slack), or you can tweet [@greptime](https://twitter.com/greptime) on Twitter.
+
+## Code of Conduct
+
+Also, there are things that we are not looking for because they don't match the goals of the product or benefit the community. Please read [Code of Conduct](https://github.com/GreptimeTeam/greptimedb/blob/develop/CODE_OF_CONDUCT.md); we hope everyone can keep good manners and become an honored member.
+
+## License
+
+GreptimeDB uses the [Apache 2.0 license](https://github.com/GreptimeTeam/greptimedb/blob/master/LICENSE) to strike a balance between open contributions and allowing you to use the software however you want.
+
+## Getting Started
+
+### Submitting Issues
+
+- Check if an issue already exists. Before filing an issue report, see whether it's already covered. Use the search bar and check out existing issues.
+- File an issue:
+  - To report a bug, a security issue, or anything that you think is a problem and that isn't under the radar, go ahead and [open a new GitHub issue](https://github.com/GrepTimeTeam/greptimedb/issues/new).
+  - In the given templates, look for the one that suits you.
+  - If you bump into anything, reach out to our [Slack](https://greptime.com/slack) for a wider audience and ask for help.
+- What happens after:
+  - Once we spot a new issue, we identify and categorize it as soon as possible.
+  - Usually, it gets assigned to other developers. Follow up and see what folks are talking about and how they take care of it.
+  - Please be patient and offer as much information as you can to help reach a solution or a consensus. You are not alone and embrace team power.
 
 ### Before PR
+
+- To ensure that community is free and confident in its ability to use your contributions, please sign the Contributor License Agreement (CLA) which will be incorporated in the pull request process.
+- Make sure all your codes are formatted and follow the [coding style](https://pingcap.github.io/style-guide/rust/).
 - Make sure all unit tests are passed.
 - Make sure all clippy warnings are fixed (you can check it locally by running `cargo clippy --workspace --all-targets -- -D warnings -D clippy::print_stdout -D clippy::print_stderr`).
 
@@ -38,19 +81,31 @@ now `pre-commit` will run automatically on `git commit`.
 
 ### Title
 
-The titles of pull requests should be prefixed with one of the change types listed in [Conventional Commits specification](https://www.conventionalcommits.org/en/v1.0.0) 
-like `feat`/`fix`/`docs`, with a concise summary of code change follows. The following scope field is optional, you can fill it with the name of sub-crate if the pull request only changes one, or just leave it blank.
+The titles of pull requests should be prefixed with category names listed in [Conventional Commits specification](https://www.conventionalcommits.org/en/v1.0.0)
+like `feat`/`fix`/`docs`, with a concise summary of code change following. DO NOT use last commit message as pull request title.
 
 ### Description
 
-- If your pull request is small, like a typo fix, feel free to go brief.
+- Feel free to go brief if your pull request is small, like a typo fix.
 - But if it contains large code change, make sure to state the motivation/design details of this PR so that reviewers can understand what you're trying to do.
 - If the PR contains any breaking change or API change, make sure that is clearly listed in your description.
 
-## Getting help
+### Commit Messages
+
+All commit messages SHOULD adhere to the [Conventional Commits specification](https://conventionalcommits.org/).
+
+## Getting Help
 
 There are many ways to get help when you're stuck. It is recommended to ask for help by opening an issue, with a detailed description
-of what you were trying to do and what went wrong. You can also reach for help in our Slack channel.
+of what you were trying to do and what went wrong. You can also reach for help in our [Slack channel](https://greptime.com/slack).
 
-## Bug report
-To report a bug or a security issue, you can [open a new GitHub issue](https://github.com/GrepTimeTeam/greptimedb/issues/new).
+## Community
+
+The core team will be thrilled if you participate in any way you like. When you are stuck, try ask for help by filing an issue, with a detailed description of what you were trying to do and what went wrong. If you have any questions or if you would like to get involved in our community, please check out:
+- [GreptimeDB Community Slack](https://greptime.com/slack)
+- [GreptimeDB Github Discussions](https://github.com/GreptimeTeam/greptimedb/discussions)
+
+Also, see some extra GreptimeDB content:
+- [GreptimeDB Docs](https://greptime.com/docs)
+- [Learn GreptimeDB](https://greptime.com/products/db)
+- [Greptime Inc. Website](https://greptime.com)

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 +24,38 @@ 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-schema = { version = "29.0", features = ["serde"] }
+# 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" }
+parquet = "29.0"
+sqlparser = "0.28"
+
 [profile.release]
 debug = true

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

@@ -1,101 +1,100 @@
-# GreptimeDB
+<p align="center">
+  <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>
 
-[![codecov](https://codecov.io/gh/GrepTimeTeam/greptimedb/branch/develop/graph/badge.svg?token=FITFDI3J3C)](https://codecov.io/gh/GrepTimeTeam/greptimedb)
 
-GreptimeDB: the next-generation hybrid timeseries/analytics processing database in the cloud.
+<h3 align="center">
+    The next-generation hybrid timeseries/analytics processing database in the cloud
+</h3>
 
-## Getting Started
+<p align="center">
+    <a href="https://codecov.io/gh/GrepTimeTeam/greptimedb"><img src="https://codecov.io/gh/GrepTimeTeam/greptimedb/branch/develop/graph/badge.svg?token=FITFDI3J3C"></img></a>
+    &nbsp;
+    <a href="https://github.com/GreptimeTeam/greptimedb/actions/workflows/develop.yml"><img src="https://github.com/GreptimeTeam/greptimedb/actions/workflows/develop.yml/badge.svg" alt="CI"></img></a>
+    &nbsp;
+    <a href="https://github.com/greptimeTeam/greptimedb/blob/develop/LICENSE"><img src="https://img.shields.io/github/license/greptimeTeam/greptimedb"></a>
+</p>
 
-### Prerequisites
+<p align="center">
+    <a href="https://twitter.com/greptime"><img src="https://img.shields.io/badge/twitter-follow_us-1d9bf0.svg"></a>
+    &nbsp;
+    <a href="https://www.linkedin.com/company/greptime/"><img src="https://img.shields.io/badge/linkedin-connect_with_us-0a66c2.svg"></a>
+</p>
 
-To compile GreptimeDB from source, you'll need the following:
-- Rust
-- Protobuf
+## What is GreptimeDB
 
-#### Rust
+GreptimeDB is an open-source time-series database with a special focus on
+scalability, analytical capabilities and efficiency. It's designed to work on
+infrastructure of the cloud era, and users benefit from its elasticity and commodity
+storage.
 
-The easiest way to install Rust is to use [`rustup`](https://rustup.rs/), which will check our `rust-toolchain` file and install correct Rust version for you.
+Our core developers have been building time-series data platform
+for years. Based on their best-practices, GreptimeDB is born to give you:
 
-#### Protobuf
+- A standalone binary that scales to highly-available distributed cluster, providing a transparent experience for cluster users
+- Optimized columnar layout for handling time-series data; compacted, compressed, stored on various storage backends
+- Flexible index options, tackling high cardinality issues down
+- Distributed, parallel query execution, leveraging elastic computing resource
+- Native SQL, and Python scripting for advanced analytical scenarios
+- Widely adopted database protocols and APIs
+- Extensible table engine architecture for extensive workloads
 
-`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/).
+## Quick Start
 
-### Build the Docker Image
+### Build
+
+#### Build from Source
+
+To compile GreptimeDB from source, you'll need:
+
+- C/C++ Toolchain: provides basic tools for compiling and linking. This is
+  available as `build-essential` on ubuntu and similar name on other platforms.
+- Rust: the easiest way to install Rust is to use
+  [`rustup`](https://rustup.rs/), which will check our `rust-toolchain` file and
+  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/).
+  **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
+
+A docker image with necessary dependencies is provided:
 
 ```
 docker build --network host -f docker/Dockerfile -t greptimedb .
 ```
 
-## Usage
+### Run
 
-### Start Datanode
+Start GreptimeDB from source code, in standalone mode:
 
 ```
-// Start datanode with default options.
-cargo run -- datanode start
-
-OR
-
-// Start datanode with `http-addr` option.
-cargo run -- datanode start --http-addr=0.0.0.0:9999
-
-OR
-
-// Start datanode with `log-dir` and `log-level` options.
-cargo run -- --log-dir=logs --log-level=debug datanode start
+cargo run -- standalone start
 ```
 
-Start datanode with config file:
+Or if you built from docker:
 
 ```
-cargo run -- --log-dir=logs --log-level=debug datanode start -c ./config/datanode.example.toml
+docker run -p 4002:4002 -v "$(pwd):/tmp/greptimedb" greptime/greptimedb standalone start
 ```
 
-Start datanode by runing docker container:
+For more startup options, greptimedb's **distributed mode** and information
+about Kubernetes deployment, check our [docs](https://docs.greptime.com/).
 
-```
-docker run -p 3000:3000 \
--p 3001:3001 \
--p 3306:3306 \
-greptimedb
-```
+### Connect
 
-### Start Frontend
-
-Frontend should connect to Datanode, so **Datanode must have been started** at first!
-
-```
-// Connects to local Datanode at its default GRPC port: 3001
-
-// Start Frontend with default options.
-cargo run -- frontend start
-
-OR
-
-// Start Frontend with `mysql-addr` option.
-cargo run -- frontend start --mysql-addr=0.0.0.0:9999
-
-OR
-
-// Start datanode with `log-dir` and `log-level` options.
-cargo run -- --log-dir=logs --log-level=debug frontend start
-```
-
-Start datanode with config file:
-
-```
-cargo run -- --log-dir=logs --log-level=debug frontend start -c ./config/frontend.example.toml
-```
-
-### SQL Operations
-
-1. Connecting DB by [mysql client](https://dev.mysql.com/downloads/mysql/):
+1. Connect to GreptimeDB via standard [MySQL
+   client](https://dev.mysql.com/downloads/mysql/):
 
    ```
-   # The datanode listen on port 3306 by default.
-   mysql -h 127.0.0.1 -P 3306
+   # The standalone instance listen on port 4002 by default.
+   mysql -h 127.0.0.1 -P 4002
    ```
 
 2. Create table:
@@ -110,29 +109,95 @@ cargo run -- --log-dir=logs --log-level=debug frontend start -c ./config/fronten
      PRIMARY KEY(host)) ENGINE=mito WITH(regions=1);
    ```
 
-3. Insert data:
+3. Insert some data:
 
    ```SQL
-   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host1', 66.6, 1024, 1660897955);
-   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host2', 77.7, 2048, 1660897956);
-   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host3', 88.8, 4096, 1660897957);
+   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host1', 66.6, 1024, 1660897955000);
+   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host2', 77.7, 2048, 1660897956000);
+   INSERT INTO monitor(host, cpu, memory, ts) VALUES ('host3', 88.8, 4096, 1660897957000);
    ```
 
-4. Query data:
+4. Query the data:
 
    ```SQL
-   mysql> SELECT * FROM monitor;
-   +-------+------------+------+--------+
-   | host  | ts         | cpu  | memory |
-   +-------+------------+------+--------+
-   | host1 | 1660897955 | 66.6 |   1024 |
-   | host2 | 1660897956 | 77.7 |   2048 |
-   | host3 | 1660897957 | 88.8 |   4096 |
-   +-------+------------+------+--------+
+   SELECT * FROM monitor;
+   ```
+
+   ```TEXT
+   +-------+---------------------+------+--------+
+   | host  | ts                  | cpu  | memory |
+   +-------+---------------------+------+--------+
+   | host1 | 2022-08-19 08:32:35 | 66.6 |   1024 |
+   | host2 | 2022-08-19 08:32:36 | 77.7 |   2048 |
+   | host3 | 2022-08-19 08:32:37 | 88.8 |   4096 |
+   +-------+---------------------+------+--------+
    3 rows in set (0.01 sec)
    ```
-   You can delete your data by removing `/tmp/greptimedb`.
+
+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)
+
+## Project Status
+
+This project is in its early stage and under heavy development. We move fast and
+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
+ask for help by filling an issue with a detailed description of what you were trying to do
+and what went wrong. If you have any questions or if you would like to get involved in our
+community, please check out:
+
+- GreptimeDB Community on [Slack](https://greptime.com/slack)
+- GreptimeDB GitHub [Discussions](https://github.com/GreptimeTeam/greptimedb/discussions)
+- Greptime official [Website](https://greptime.com)
+
+In addition, you may:
+
+- 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)
+
+## License
+
+GreptimeDB uses the [Apache 2.0 license][1] to strike a balance between
+open contributions and allowing you to use the software however you want.
+
+[1]: <https://github.com/greptimeTeam/greptimedb/blob/develop/LICENSE>
 
 ## Contributing
 
 Please refer to [contribution guidelines](CONTRIBUTING.md) for more information.
+
+## Acknowledgement
+- GreptimeDB uses [Apache Arrow](https://arrow.apache.org/) as the memory model and [Apache Parquet](https://parquet.apache.org/) as the persistent file format.
+- 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

@@ -1,13 +1,14 @@
 [package]
 name = "benchmarks"
-version = "0.1.0"
-edition = "2021"
+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 = "*" }
+parquet.workspace = true
 tokio = { version = "1.21", features = ["full"] }

benchmarks/src/bin/nyc-taxi.rs

@@ -1,35 +1,36 @@
+// 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 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,
-    path::{Path, PathBuf},
-    sync::Arc,
-    time::Instant,
-};
+use std::collections::HashMap;
+use std::path::{Path, PathBuf};
+use std::time::Instant;
 
-use arrow::{
-    array::{ArrayRef, PrimitiveArray, StringArray, TimestampNanosecondArray},
-    datatypes::{DataType, Float64Type, Int64Type},
-    record_batch::RecordBatch,
-};
+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,
-    api::v1::{
-        codec::InsertBatch, column::Values, insert_expr, Column, ColumnDataType, ColumnDef,
-        CreateExpr, InsertExpr,
-    },
-    Client, Database, Select,
-};
+use client::admin::Admin;
+use client::api::v1::column::Values;
+use client::api::v1::{Column, ColumnDataType, ColumnDef, CreateTableExpr, InsertExpr, TableId};
+use client::{Client, Database, Select};
 use indicatif::{MultiProgress, ProgressBar, ProgressStyle};
-use parquet::{
-    arrow::{ArrowReader, ParquetFileArrowReader},
-    file::{reader::FileReader, serialized_reader::SerializedFileReader},
-};
+use parquet::arrow::arrow_reader::ParquetRecordBatchReaderBuilder;
 use tokio::task::JoinSet;
 
 const DATABASE_NAME: &str = "greptime";
@@ -81,27 +82,30 @@ 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 (columns, row_count) = convert_record_batch(record_batch);
         let insert_expr = InsertExpr {
+            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();
@@ -110,14 +114,11 @@ async fn write_data(
         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();
@@ -135,10 +136,7 @@ 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 {
@@ -209,139 +207,142 @@ 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(),
         primary_keys: vec!["VendorID".to_string()],
         create_if_not_exists: false,
         table_options: Default::default(),
+        region_ids: vec![0],
+        table_id: Some(TableId { id: 0 }),
     }
 }
 
@@ -350,12 +351,12 @@ 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
@@ -368,7 +369,7 @@ async fn do_write(args: &Args, client: &Client) {
     let mut write_jobs = JoinSet::new();
 
     let create_table_result = admin.create(create_table_expr()).await;
-    println!("Create table result: {:?}", create_table_result);
+    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}",
@@ -401,7 +402,7 @@ 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();

codecov.yml

@@ -1,10 +1,10 @@
 # codecov config
 coverage:
   status:
-    patch: off # disable patch status
     project:
       default:
-        enable: yes
         threshold: 1%
+    patch: off
 ignore:
   - "**/error*.rs" # ignore all error.rs files
+  - "tests/runner/*.rs" # ignore integration test runner

component/script/python/example/calc_rv.py

@@ -1,71 +0,0 @@
-import sys
-# for annoying releative import beyond top-level package
-sys.path.insert(0, "../")
-from greptime import mock_tester, coprocessor, greptime as gt_builtin
-from greptime.greptime import interval, vector, log, prev, sqrt, datetime
-import greptime.greptime as greptime
-import json
-import numpy as np
-
-
-def data_sample(k_lines, symbol, density=5 * 30 * 86400):
-    """
-    Only return close data for simplicty for now
-    """
-    k_lines = k_lines["result"] if k_lines["ret_msg"] == "OK" else None
-    if k_lines is None:
-        raise Exception("Expect a `OK`ed message")
-    close = [float(i["close"]) for i in k_lines]
-
-    return interval(close, density, "prev")
-
-
-def as_table(kline: list):
-    col_len = len(kline)
-    ret = {
-        k: vector([fn(row[k]) for row in kline], str(ty))
-        for k, fn, ty in
-        [
-            ("symbol", str, "str"),
-            ("period", str, "str"),
-            ("open_time", int, "int"),
-            ("open", float, "float"),
-            ("high", float, "float"),
-            ("low", float, "float"),
-            ("close", float, "float")
-        ]
-    }
-    return ret
-
-@coprocessor(args=["open_time", "close"], returns=[
-    "rv_7d",
-    "rv_15d",
-    "rv_30d",
-    "rv_60d",
-    "rv_90d",
-    "rv_180d"
-])
-def calc_rvs(open_time, close):
-    from greptime import vector, log, prev, sqrt, datetime, pow, sum, last
-    import greptime as g
-    def calc_rv(close, open_time, time, interval):
-        mask = (open_time < time) & (open_time > time - interval)
-        close = close[mask]
-        open_time = open_time[mask]
-        close = g.interval(open_time, close, datetime("10m"), lambda x:last(x))
-
-        avg_time_interval = (open_time[-1] - open_time[0])/(len(open_time)-1)
-        ref = log(close/prev(close))
-        var = sum(pow(ref, 2)/(len(ref)-1))
-        return sqrt(var/avg_time_interval)
-
-    # how to get env var,
-    # maybe through accessing scope and serde then send to remote?
-    timepoint = open_time[-1]
-    rv_7d = vector([calc_rv(close, open_time, timepoint, datetime("7d"))])
-    rv_15d = vector([calc_rv(close, open_time, timepoint, datetime("15d"))])
-    rv_30d = vector([calc_rv(close, open_time, timepoint, datetime("30d"))])
-    rv_60d = vector([calc_rv(close, open_time, timepoint, datetime("60d"))])
-    rv_90d = vector([calc_rv(close, open_time, timepoint, datetime("90d"))])
-    rv_180d = vector([calc_rv(close, open_time, timepoint, datetime("180d"))])
-    return rv_7d, rv_15d, rv_30d, rv_60d, rv_90d, rv_180d

component/script/python/example/fetch_kline.sh

@@ -1 +0,0 @@
-curl "https://api.bybit.com/v2/public/index-price-kline?symbol=BTCUSD&interval=1&limit=$1&from=1581231260" > kline.json

component/script/python/example/kline.json

@@ -1,108 +0,0 @@
-{
-    "ret_code": 0,
-    "ret_msg": "OK",
-    "ext_code": "",
-    "ext_info": "",
-    "result": [
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 300,
-            "open": "10107",
-            "high": "10109.34",
-            "low": "10106.71",
-            "close": "10106.79"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 900,
-            "open": "10106.79",
-            "high": "10109.27",
-            "low": "10105.92",
-            "close": "10106.09"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 1200,
-            "open": "10106.09",
-            "high": "10108.75",
-            "low": "10104.66",
-            "close": "10108.73"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 1800,
-            "open": "10108.73",
-            "high": "10109.52",
-            "low": "10106.07",
-            "close": "10106.38"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 2400,
-            "open": "10106.38",
-            "high": "10109.48",
-            "low": "10104.81",
-            "close": "10106.95"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 3000,
-            "open": "10106.95",
-            "high": "10109.48",
-            "low": "10106.6",
-            "close": "10107.55"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 3600,
-            "open": "10107.55",
-            "high": "10109.28",
-            "low": "10104.68",
-            "close": "10104.68"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 4200,
-            "open": "10104.68",
-            "high": "10109.18",
-            "low": "10104.14",
-            "close": "10108.8"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 4800,
-            "open": "10108.8",
-            "high": "10117.36",
-            "low": "10108.8",
-            "close": "10115.96"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 5400,
-            "open": "10115.96",
-            "high": "10119.19",
-            "low": "10115.96",
-            "close": "10117.08"
-        },
-        {
-            "symbol": "BTCUSD",
-            "period": "1",
-            "open_time": 6000,
-            "open": "10117.08",
-            "high": "10120.73",
-            "low": "10116.96",
-            "close": "10120.43"
-        }
-    ],
-    "time_now": "1661225351.158190"
-}

component/script/python/greptime/__init__.py

@@ -1,4 +0,0 @@
-from .greptime import coprocessor, copr
-from .greptime import vector, log, prev, next, first, last, sqrt, pow, datetime, sum, interval
-from .mock import mock_tester
-from .cfg import set_conn_addr, get_conn_addr

component/script/python/greptime/cfg.py

@@ -1,11 +0,0 @@
-GREPTIME_DB_CONN_ADDRESS = "localhost:3000"
-"""The Global Variable for address for conntect to database"""
-
-def set_conn_addr(addr: str):
-    """set database address to given `addr`"""
-    global GREPTIME_DB_CONN_ADDRESS
-    GREPTIME_DB_CONN_ADDRESS = addr
-
-def get_conn_addr()->str:
-    global GREPTIME_DB_CONN_ADDRESS
-    return GREPTIME_DB_CONN_ADDRESS

component/script/python/greptime/greptime.py

@@ -1,207 +0,0 @@
-"""
-Be note that this is a mock library, if not connected to database,
-it can only run on mock data and mock function which is supported by numpy
-"""
-import functools
-import numpy as np
-import json
-from urllib import request
-import inspect
-import requests
-
-from .cfg import set_conn_addr, get_conn_addr
-
-log = np.log
-sum = np.nansum
-sqrt = np.sqrt
-pow = np.power
-nan = np.nan
-
-
-class TimeStamp(str):
-    """
-    TODO: impl date time
-    """
-    pass
-
-
-class i32(int):
-    """
-    For Python Coprocessor Type Annotation ONLY
-    A signed 32-bit integer.
-    """
-
-    def __repr__(self) -> str:
-        return "i32"
-
-
-class i64(int):
-    """
-    For Python Coprocessor Type Annotation ONLY
-    A signed 64-bit integer.
-    """
-
-    def __repr__(self) -> str:
-        return "i64"
-
-
-class f32(float):
-    """
-    For Python Coprocessor Type Annotation ONLY
-    A 32-bit floating point number.
-    """
-
-    def __repr__(self) -> str:
-        return "f32"
-
-
-class f64(float):
-    """
-    For Python Coprocessor Type Annotation ONLY
-    A 64-bit floating point number.
-    """
-
-    def __repr__(self) -> str:
-        return "f64"
-
-
-class vector(np.ndarray):
-    """
-    A compact Vector with all elements of same Data type.
-    """
-    _datatype: str | None = None
-
-    def __new__(
-        cls,
-        lst,
-        dtype=None
-    ) -> ...:
-        self = np.asarray(lst).view(cls)
-        self._datatype = dtype
-        return self
-
-    def __str__(self) -> str:
-        return "vector({}, \"{}\")".format(super().__str__(), self.datatype())
-
-    def datatype(self):
-        return self._datatype
-
-    def filter(self, lst_bool):
-        return self[lst_bool]
-
-def last(lst):
-    return lst[-1]
-
-def first(lst):
-    return lst[0]
-
-def prev(lst):
-    ret = np.zeros(len(lst))
-    ret[1:] = lst[0:-1]
-    ret[0] = nan
-    return ret
-
-def next(lst):
-    ret = np.zeros(len(lst))
-    ret[:-1] = lst[1:]
-    ret[-1] = nan
-    return ret
-
-def interval(ts: vector, arr: vector, duration: int, func):
-    """
-    Note that this is a mock function with same functionailty to the actual Python Coprocessor
-    `arr` is a vector of integral or temporal type.
-    """
-    start = np.min(ts)
-    end = np.max(ts)
-    masks = [(ts >= i) & (ts <= (i+duration)) for i in range(start, end, duration)]
-    lst_res = [func(arr[mask]) for mask in masks]
-    return lst_res
-
-
-def factor(unit: str) -> int:
-    if unit == "d":
-        return 24 * 60 * 60
-    elif unit == "h":
-        return 60 * 60
-    elif unit == "m":
-        return 60
-    elif unit == "s":
-        return 1
-    else:
-        raise Exception("Only d,h,m,s, found{}".format(unit))
-
-
-def datetime(input_time: str) -> int:
-    """
-    support `d`(day) `h`(hour) `m`(minute) `s`(second)
-
-    support format:
-    `12s` `7d` `12d2h7m`
-    """
-
-    prev = 0
-    cur = 0
-    state = "Num"
-    parse_res = []
-    for idx, ch in enumerate(input_time):
-        if ch.isdigit():
-            cur = idx
-
-            if state != "Num":
-                parse_res.append((state, input_time[prev:cur], (prev, cur)))
-                prev = idx
-                state = "Num"
-        else:
-            cur = idx
-            if state != "Symbol":
-                parse_res.append((state, input_time[prev:cur], (prev, cur)))
-                prev = idx
-                state = "Symbol"
-    parse_res.append((state, input_time[prev:cur+1], (prev, cur+1)))
-
-    cur_idx = 0
-    res_time = 0
-    while cur_idx < len(parse_res):
-        pair = parse_res[cur_idx]
-        if pair[0] == "Num":
-            val = int(pair[1])
-            nxt = parse_res[cur_idx+1]
-            res_time += val * factor(nxt[1])
-            cur_idx += 2
-        else:
-            raise Exception("Two symbol in a row is impossible")
-
-    return res_time
-
-
-def coprocessor(args=None, returns=None, sql=None):
-    """
-    The actual coprocessor, which will connect to database and update
-    whatever function decorated with `@coprocessor(args=[...], returns=[...], sql=...)`
-    """
-    def decorator_copr(func):
-        @functools.wraps(func)
-        def wrapper_do_actual(*args, **kwargs):
-            if len(args)!=0 or len(kwargs)!=0:
-                raise Exception("Expect call with no arguements(for all args are given by coprocessor itself)")
-            source = inspect.getsource(func)
-            url = "http://{}/v1/scripts".format(get_conn_addr())
-            print("Posting to {}".format(url))
-            data = {
-                    "script": source,
-                    "engine": None,
-                }
-
-            res = requests.post(
-                url,
-                headers={"Content-Type": "application/json"},
-                json=data
-            )
-            return res
-        return wrapper_do_actual
-    return decorator_copr
-
-
-# make a alias for short
-copr = coprocessor

component/script/python/greptime/mock.py

@@ -1,82 +0,0 @@
-"""
-Note this is a mock library, if not connected to database, 
-it can only run on mock data and support by numpy
-"""
-from typing import Any
-import numpy as np
-from .greptime import i32,i64,f32,f64, vector, interval, prev, datetime, log, sum, sqrt, pow, nan, copr, coprocessor
-
-import inspect
-import functools
-import ast
-
-
-
-def mock_tester(
-    func,
-    env:dict, 
-    table=None
-):
-    """
-    Mock tester helper function,
-    What it does is replace `@coprocessor` with `@mock_cpor` and add a keyword `env=env`
-    like `@mock_copr(args=...,returns=...,env=env)`
-    """
-    code = inspect.getsource(func)
-    tree = ast.parse(code)
-    tree = HackyReplaceDecorator("env").visit(tree)
-    new_func = tree.body[0]
-    fn_name = new_func.name
-
-    code_obj = compile(tree, "<embedded>", "exec")
-    exec(code_obj)
-
-    ret = eval("{}()".format(fn_name))
-    return ret
-
-def mock_copr(args, returns, sql=None, env:None|dict=None):
-    """
-    This should not be used directly by user
-    """
-    def decorator_copr(func):
-        @functools.wraps(func)
-        def wrapper_do_actual(*fn_args, **fn_kwargs):
-
-            real_args = [env[name] for name in args]
-            ret = func(*real_args)
-            return ret
-        
-        return wrapper_do_actual
-    return decorator_copr
-
-class HackyReplaceDecorator(ast.NodeTransformer):
-    """
-    This class accept a `env` dict for environment to extract args from,
-    and put `env` dict in the param list of `mock_copr` decorator, i.e:
-
-    a `@copr(args=["a", "b"], returns=["c"])` with call like mock_helper(abc, env={"a":2, "b":3})
-    
-    will be transform into `@mock_copr(args=["a", "b"], returns=["c"], env={"a":2, "b":3})`
-    """
-    def __init__(self, env: str) -> None:
-        # just for add `env` keyword
-        self.env = env
-
-    def visit_FunctionDef(self, node: ast.FunctionDef) -> Any:
-        new_node = node
-        decorator_list = new_node.decorator_list
-        if len(decorator_list)!=1:
-            return node
-
-        deco = decorator_list[0]
-        if deco.func.id!="coprocessor" and deco.func.id !="copr":
-            raise Exception("Expect a @copr or @coprocessor, found {}.".format(deco.func.id))
-        deco.func = ast.Name(id="mock_copr", ctx=ast.Load())
-        new_kw = ast.keyword(arg="env", value=ast.Name(id=self.env, ctx=ast.Load()))
-        deco.keywords.append(new_kw)
-
-        # Tie up loose ends in the AST.
-        ast.copy_location(new_node, node)
-        ast.fix_missing_locations(new_node)
-        self.generic_visit(node)
-        return new_node

component/script/python/test.py

@@ -1,60 +0,0 @@
-from example.calc_rv import as_table, calc_rvs
-from greptime import coprocessor, set_conn_addr, get_conn_addr, mock_tester
-import sys
-import json
-import requests
-'''
-To run this script, you need to first start a http server of greptime, and 
-`
-python3 component/script/python/test.py 地址:端口
-`
-
-'''
-@coprocessor(sql='select number from numbers limit 10', args=['number'], returns=['n'])
-def test(n):
-    return n+2
-
-def init_table(close, open_time):
-    req_init = "/v1/sql?sql=create table k_line (close double, open_time bigint, TIME INDEX (open_time))"
-    print(get_db(req_init).text)
-    for c1, c2 in zip(close, open_time):
-        req = "/v1/sql?sql=INSERT INTO k_line(close, open_time) VALUES ({}, {})".format(c1, c2)
-        print(get_db(req).text)
-    print(get_db("/v1/sql?sql=select * from k_line").text)
-
-def get_db(req:str):
-    return requests.get("http://{}{}".format(get_conn_addr(), req))
-
-if __name__ == "__main__":
-    with open("component/script/python/example/kline.json", "r") as kline_file:
-        kline = json.load(kline_file)
-        table = as_table(kline["result"])
-        close = table["close"]
-        open_time = table["open_time"]
-        env = {"close":close, "open_time": open_time}
-
-        res = mock_tester(calc_rvs, env=env)
-        print("Mock result:", [i[0] for i in res])
-        exit()
-    if len(sys.argv)!=2:
-        raise Exception("Expect only one address as cmd's args")
-    set_conn_addr(sys.argv[1])
-    res = test()
-    print(res.headers)
-    print(res.text)
-    with open("component/script/python/example/kline.json", "r") as kline_file:
-        kline = json.load(kline_file)
-        # vec = vector([1,2,3], int)
-        # print(vec, vec.datatype())
-        table = as_table(kline["result"])
-        # print(table)
-        close = table["close"]
-        open_time = table["open_time"]
-        init_table(close, open_time)
-
-        real = calc_rvs()
-        print(real)
-        try:
-            print(real.text["error"])
-        except:
-            print(real.text)

config/datanode.example.toml

@@ -1,22 +1,26 @@
 node_id = 42
-http_addr = '0.0.0.0:3000'
-rpc_addr = '0.0.0.0:3001'
+mode = 'distributed'
+rpc_addr = '127.0.0.1:3001'
 wal_dir = '/tmp/greptimedb/wal'
 rpc_runtime_size = 8
-mode = "standalone"
-mysql_addr = '0.0.0.0:3306'
+mysql_addr = '127.0.0.1:4406'
 mysql_runtime_size = 4
+enable_memory_catalog = false
 
-# applied when postgres feature enbaled
-postgres_addr = '0.0.0.0:5432'
-postgres_runtime_size = 4
+[wal]
+dir = "/tmp/greptimedb/wal"
+file_size = 1073741824
+purge_interval = 600
+purge_threshold = 53687091200
+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 = true
+tcp_nodelay = false

config/frontend.example.toml

@@ -1,4 +1,12 @@
-http_addr = '0.0.0.0:4000'
-grpc_addr = '0.0.0.0:4001'
-mysql_addr = '0.0.0.0:4003'
-mysql_runtime_size = 4
+mode = 'distributed'
+datanode_rpc_addr = '127.0.0.1:3001'
+
+[http_options]
+addr = '127.0.0.1:4000'
+timeout = "30s"
+
+[meta_client_opts]
+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

@@ -0,0 +1,44 @@
+node_id = 0
+mode = 'standalone'
+enable_memory_catalog = false
+
+[http_options]
+addr = '127.0.0.1:4000'
+timeout = "30s"
+
+[wal]
+dir = "/tmp/greptimedb/wal"
+file_size = 1073741824
+purge_interval = 600
+purge_threshold = 53687091200
+read_batch_size = 128
+sync_write = false
+
+
+[storage]
+type = 'File'
+data_dir = '/tmp/greptimedb/data/'
+
+[grpc_options]
+addr = '127.0.0.1:4001'
+runtime_size = 8
+
+[mysql_options]
+addr = '127.0.0.1:4002'
+runtime_size = 2
+
+[influxdb_options]
+enable = true
+
+[opentsdb_options]
+addr = '127.0.0.1:4242'
+enable = true
+runtime_size = 2
+
+[prometheus_options]
+enable = true
+
+[postgres_options]
+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

rustfmt.toml

@@ -1,3 +1,2 @@
 group_imports = "StdExternalCrate"
-
-
+imports_granularity = "Module"

scripts/install.sh

@@ -7,7 +7,7 @@ ARCH_TYPE=
 VERSION=${1:-latest}
 GITHUB_ORG=GreptimeTeam
 GITHUB_REPO=greptimedb
-BIN=greptimedb
+BIN=greptime
 
 get_os_type() {
     os_type="$(uname -s)"

src/api/Cargo.toml

@@ -1,11 +1,12 @@
 [package]
 name = "api"
-version = "0.1.0"
-edition = "2021"
-# 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]
 common-base = { path = "../common/base" }
+common-error = { path = "../common/error" }
 common-time = { path = "../common/time" }
 datatypes = { path = "../datatypes" }
 prost = "0.11"

src/api/build.rs

@@ -1,3 +1,17 @@
+// 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::path::PathBuf;
 
 fn main() {
@@ -6,9 +20,7 @@ fn main() {
         .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/meta/common.proto",
                 "greptime/v1/meta/heartbeat.proto",

src/api/greptime/v1/admin.proto

@@ -17,8 +17,10 @@ message AdminResponse {
 message AdminExpr {
   ExprHeader header = 1;
   oneof expr {
-    CreateExpr create = 2;
+    CreateTableExpr create_table = 2;
     AlterExpr alter = 3;
+    CreateDatabaseExpr create_database = 4;
+    DropTableExpr drop_table = 5;
   }
 }
 
@@ -29,27 +31,58 @@ message AdminResult {
   }
 }
 
-message CreateExpr {
-  optional string catalog_name = 1;
-  optional string schema_name = 2;
+message CreateTableExpr {
+  string catalog_name = 1;
+  string schema_name = 2;
   string table_name = 3;
-  optional string desc = 4;
+  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 {
-  optional string catalog_name = 1;
-  optional string schema_name = 2;
+  string catalog_name = 1;
+  string schema_name = 2;
   string table_name = 3;
   oneof kind {
-    AddColumn add_column = 4;
+    AddColumns add_columns = 4;
+    DropColumns drop_columns = 5;
   }
 }
 
+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;
+}
+
+message AddColumns {
+  repeated AddColumn add_columns = 1;
+}
+
+message DropColumns {
+  repeated DropColumn drop_columns = 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/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/database.proto

@@ -2,6 +2,7 @@ syntax = "proto3";
 
 package greptime.v1;
 
+import "greptime/v1/column.proto";
 import "greptime/v1/common.proto";
 
 message DatabaseRequest {
@@ -28,36 +29,23 @@ message SelectExpr {
   oneof expr {
     string sql = 1;
     bytes logical_plan = 2;
-    PhysicalPlan physical_plan = 15;
   }
 }
 
-message PhysicalPlan {
-  bytes original_ql = 1;
-  bytes plan = 2;
-}
-
 message InsertExpr {
-  string table_name = 1;
+  string schema_name = 1;
+  string table_name = 2;
 
-  message Values {
-    repeated bytes values = 1;
-  }
+  // Data is represented here.
+  repeated Column columns = 3;
 
-  oneof expr {
-    Values values = 2;
+  // The row_count of all columns, which include null and non-null values.
+  //
+  // Note: the row_count of all columns in a InsertExpr 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 = 3;
-  }
-  
-  map<string, bytes> options = 4;
+  // The region number of current insert request.
+  uint32 region_number = 5;
 }
 
 // TODO(jiachun)

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

@@ -39,7 +39,7 @@ message NodeStat {
   uint64 wcus = 2;
   // Table number in this node
   uint64 table_num = 3;
-  // Regon number in this node
+  // Region number in this node
   uint64 region_num = 4;
 
   double cpu_usage = 5;

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;
@@ -69,6 +77,7 @@ message Table {
 }
 
 message Region {
+  // TODO(LFC): Maybe use message RegionNumber?
   uint64 id = 1;
   string name = 2;
   Partition partition = 3;

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/src/error.rs

@@ -1,6 +1,24 @@
+// 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::any::Any;
+
+use common_error::ext::ErrorExt;
+use common_error::prelude::StatusCode;
 use datatypes::prelude::ConcreteDataType;
 use snafu::prelude::*;
-use snafu::Backtrace;
+use snafu::{Backtrace, ErrorCompat};
 
 pub type Result<T> = std::result::Result<T, Error>;
 
@@ -15,4 +33,44 @@ 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 {
+    fn status_code(&self) -> StatusCode {
+        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> {
+        ErrorCompat::backtrace(self)
+    }
+
+    fn as_any(&self) -> &dyn Any {
+        self
+    }
 }

src/api/src/helper.rs

@@ -1,14 +1,28 @@
+// 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_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::*;
 
 use crate::error::{self, Result};
 use crate::v1::column::Values;
-use crate::v1::Column;
-use crate::v1::ColumnDataType;
+use crate::v1::{Column, ColumnDataType};
 
 #[derive(Debug, PartialEq, Eq)]
 pub struct ColumnDataTypeWrapper(ColumnDataType);
@@ -43,7 +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()
+            }
         }
     }
 }
@@ -68,7 +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()
             }
@@ -140,8 +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()
             },
         }
@@ -174,9 +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();
@@ -187,7 +230,10 @@ impl Column {
 mod tests {
     use std::sync::Arc;
 
-    use datatypes::vectors::BooleanVector;
+    use datatypes::vectors::{
+        BooleanVector, TimestampMicrosecondVector, TimestampMillisecondVector,
+        TimestampNanosecondVector, TimestampSecondVector,
+    };
 
     use super::*;
 
@@ -245,8 +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());
     }
 
@@ -313,8 +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()
         );
     }
 
@@ -381,8 +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()
         );
@@ -399,7 +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

@@ -1,6 +1,21 @@
+// 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.
+
 pub mod error;
 pub mod helper;
 pub mod prometheus;
+pub mod result;
 pub mod serde;
 pub mod v1;
 

src/api/src/prometheus.rs

@@ -1,3 +1,17 @@
+// 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.
+
 #![allow(clippy::derive_partial_eq_without_eq)]
 
 pub mod remote {

src/api/src/result.rs

@@ -1,23 +1,39 @@
-use api::v1::{
-    admin_result, codec::SelectResult, object_result, AdminResult, MutateResult, ObjectResult,
-    ResultHeader, SelectResult as SelectResultRaw,
-};
+// 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(crate) struct ObjectResultBuilder {
+pub struct ObjectResultBuilder {
     version: u32,
     code: u32,
     err_msg: Option<String>,
     result: Option<Body>,
 }
 
-pub(crate) enum Body {
+pub enum Body {
     Mutate((Success, Failure)),
     Select(SelectResult),
 }
@@ -80,7 +96,7 @@ impl ObjectResultBuilder {
     }
 }
 
-pub(crate) fn build_err_result(err: &impl ErrorExt) -> ObjectResult {
+pub fn build_err_result(err: &impl ErrorExt) -> ObjectResult {
     ObjectResultBuilder::new()
         .status_code(err.status_code() as u32)
         .err_msg(err.to_string())
@@ -88,7 +104,7 @@ pub(crate) fn build_err_result(err: &impl ErrorExt) -> ObjectResult {
 }
 
 #[derive(Debug)]
-pub(crate) struct AdminResultBuilder {
+pub struct AdminResultBuilder {
     version: u32,
     code: u32,
     err_msg: Option<String>,
@@ -144,11 +160,11 @@ impl Default for AdminResultBuilder {
 
 #[cfg(test)]
 mod tests {
-    use api::v1::{object_result, MutateResult};
     use common_error::status_code::StatusCode;
 
     use super::*;
-    use crate::error::UnsupportedExprSnafu;
+    use crate::error::UnknownColumnDataTypeSnafu;
+    use crate::v1::{object_result, MutateResult};
 
     #[test]
     fn test_object_result_builder() {
@@ -175,14 +191,13 @@ mod tests {
 
     #[test]
     fn test_build_err_result() {
-        let err = UnsupportedExprSnafu { name: "select" }.build();
+        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::Internal as u32, header.code);
-        assert_eq!("Unsupported expr type: select", header.err_msg);
+        assert_eq!(StatusCode::InvalidArguments as u32, header.code);
         assert!(result.is_none());
     }
 }

src/api/src/serde.rs

@@ -1,9 +1,20 @@
+// 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.
+
 pub use prost::DecodeError;
 use prost::Message;
 
-use crate::v1::codec::InsertBatch;
-use crate::v1::codec::PhysicalPlanNode;
-use crate::v1::codec::RegionNumber;
 use crate::v1::codec::SelectResult;
 use crate::v1::meta::TableRouteValue;
 
@@ -25,10 +36,7 @@ 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)]
@@ -36,57 +44,10 @@ mod tests {
     use std::ops::Deref;
 
     use crate::v1::codec::*;
-    use crate::v1::column;
-    use crate::v1::Column;
+    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();
@@ -133,35 +94,6 @@ mod tests {
         );
     }
 
-    #[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],

src/api/src/v1.rs

@@ -1,3 +1,17 @@
+// 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.
+
 #![allow(clippy::derive_partial_eq_without_eq)]
 tonic::include_proto!("greptime.v1");
 
@@ -7,4 +21,5 @@ 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 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 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

@@ -1,8 +1,21 @@
+// 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.
+
 tonic::include_proto!("greptime.v1.meta");
 
 use std::collections::HashMap;
-use std::hash::Hash;
-use std::hash::Hasher;
+use std::hash::{Hash, Hasher};
 
 pub const PROTOCOL_VERSION: u64 = 1;
 
@@ -71,11 +84,22 @@ impl ResponseHeader {
             error: Some(error),
         }
     }
+
+    #[inline]
+    pub fn is_not_leader(&self) -> bool {
+        if let Some(error) = &self.error {
+            if error.code == ErrorCode::NotLeader as i32 {
+                return true;
+            }
+        }
+        false
+    }
 }
 
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
 pub enum ErrorCode {
     NoActiveDatanodes = 1,
+    NotLeader = 2,
 }
 
 impl Error {
@@ -86,6 +110,24 @@ impl Error {
             err_msg: "No active datanodes".to_string(),
         }
     }
+
+    #[inline]
+    pub fn is_not_leader() -> Self {
+        Self {
+            code: ErrorCode::NotLeader as i32,
+            err_msg: "Current server is not leader".to_string(),
+        }
+    }
+}
+
+impl HeartbeatResponse {
+    #[inline]
+    pub fn is_not_leader(&self) -> bool {
+        if let Some(header) = &self.header {
+            return header.is_not_leader();
+        }
+        false
+    }
 }
 
 macro_rules! gen_set_header {
@@ -103,10 +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

@@ -1,8 +1,8 @@
 [package]
 name = "catalog"
-version = "0.1.0"
-edition = "2021"
-# 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" }
@@ -18,15 +18,12 @@ 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"
 lazy_static = "1.4"
 meta-client = { path = "../meta-client" }
-opendal = "0.17"
 regex = "1.6"
 serde = "1.0"
 serde_json = "1.0"
@@ -38,9 +35,8 @@ tokio = { version = "1.18", features = ["full"] }
 [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"] }

src/catalog/src/error.rs

@@ -1,9 +1,23 @@
+// 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::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};
 
@@ -37,14 +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))]
@@ -80,15 +92,27 @@ pub enum Error {
         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))]
+    SchemaExists {
+        schema: String,
+        backtrace: Backtrace,
+    },
+
     #[snafu(display("Failed to register table"))]
     RegisterTable {
         #[snafu(backtrace)]
         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,
@@ -112,7 +136,7 @@ pub enum Error {
         "Failed to insert table creation record to system catalog, source: {}",
         source
     ))]
-    InsertTableRecord {
+    InsertCatalogRecord {
         #[snafu(backtrace)]
         source: table::error::Error,
     },
@@ -165,21 +189,8 @@ pub enum Error {
         source: meta_client::error::Error,
     },
 
-    #[snafu(display("Failed to bump table id"))]
-    BumpTableId { msg: String, backtrace: Backtrace },
-
-    #[snafu(display("Failed to parse table id from metasrv, data: {:?}", data))]
-    ParseTableId { data: String, backtrace: Backtrace },
-
-    #[snafu(display("Failed to deserialize partition rule from string: {:?}", data))]
-    DeserializePartitionRule {
-        data: String,
-        source: serde_json::error::Error,
-        backtrace: Backtrace,
-    },
-
-    #[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,
     },
@@ -209,28 +220,29 @@ 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,
 
             Error::OpenSystemCatalog { source, .. }
             | Error::CreateSystemCatalog { source, .. }
-            | Error::InsertTableRecord { source, .. }
+            | Error::InsertCatalogRecord { source, .. }
             | Error::OpenTable { source, .. }
             | Error::CreateTable { source, .. } => source.status_code(),
             Error::MetaSrv { source, .. } => source.status_code(),
             Error::SystemCatalogTableScan { source } => source.status_code(),
             Error::SystemCatalogTableScanExec { source } => source.status_code(),
             Error::InvalidTableSchema { source, .. } => source.status_code(),
-            Error::BumpTableId { .. } | Error::ParseTableId { .. } => {
-                StatusCode::StorageUnavailable
-            }
-            Error::DeserializePartitionRule { .. } => StatusCode::Unexpected,
-            Error::InvalidSchemaInCatalog { .. } => StatusCode::Unexpected,
+            Error::InvalidTableInfoInCatalog { .. } => StatusCode::Unexpected,
             Error::Internal { source, .. } => source.status_code(),
+
+            Error::Unimplemented { .. } => StatusCode::Unsupported,
         }
     }
 
@@ -252,7 +264,6 @@ impl From<Error> for DataFusionError {
 #[cfg(test)]
 mod tests {
     use common_error::mock::MockError;
-    use datatypes::arrow::datatypes::DataType;
     use snafu::GenerateImplicitData;
 
     use super::*;
@@ -301,11 +312,8 @@ 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

@@ -1,54 +1,70 @@
+// 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 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-Z_]+)$",
-        TABLE_GLOBAL_KEY_PREFIX
+        "^{TABLE_GLOBAL_KEY_PREFIX}-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})-({ALPHANUMERICS_NAME_PATTERN})$"
     ))
     .unwrap();
 }
 
 lazy_static! {
     static ref TABLE_REGIONAL_KEY_PATTERN: Regex = Regex::new(&format!(
-        "^{}-([a-zA-Z_]+)-([a-zA-Z_]+)-([a-zA-Z_]+)-([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(
@@ -56,8 +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()
     )
@@ -112,18 +127,20 @@ 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>>,
-    /// Node id -> region ids
-    pub meta: RawTableMeta,
-    /// Partition rules for table
-    pub partition_rules: String,
+    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.
@@ -245,6 +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)?
@@ -266,6 +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::*;
 
@@ -326,15 +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,
-            partition_rules: "{}".to_string(),
+            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

@@ -1,6 +1,21 @@
+// 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.
+
 #![feature(assert_matches)]
 
 use std::any::Any;
+use std::fmt::{Debug, Formatter};
 use std::sync::Arc;
 
 use common_telemetry::info;
@@ -14,6 +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;
@@ -69,17 +85,24 @@ pub trait CatalogManager: CatalogList {
     /// Starts a catalog manager.
     async fn start(&self) -> Result<()>;
 
-    /// Returns next available table id.
-    async fn next_table_id(&self) -> Result<TableId>;
+    /// Registers a table within given catalog/schema to catalog manager,
+    /// returns whether the table registered.
+    async fn register_table(&self, request: RegisterTableRequest) -> Result<bool>;
 
-    /// Registers a table given given catalog/schema to catalog manager,
-    /// returns table registered.
-    async fn register_table(&self, request: RegisterTableRequest) -> 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>;
 
     /// Register a system table, should be called before starting the manager.
     async fn register_system_table(&self, request: RegisterSystemTableRequest)
         -> error::Result<()>;
 
+    fn schema(&self, catalog: &str, schema: &str) -> Result<Option<SchemaProviderRef>>;
+
     /// Returns the table by catalog, schema and table name.
     fn table(&self, catalog: &str, schema: &str, table_name: &str) -> Result<Option<TableRef>>;
 }
@@ -107,9 +130,34 @@ 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(Clone)]
+pub struct DeregisterTableRequest {
+    pub catalog: String,
+    pub schema: String,
+    pub table_name: String,
+}
+
+#[derive(Debug, Clone)]
+pub struct RegisterSchemaRequest {
+    pub catalog: String,
+    pub schema: String,
+}
+
 /// 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 {
@@ -139,8 +187,7 @@ 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
@@ -152,7 +199,7 @@ 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 {

src/catalog/src/local.rs

@@ -1,3 +1,17 @@
+// 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.
+
 pub mod manager;
 pub mod memory;
 

src/catalog/src/local/manager.rs

@@ -1,3 +1,17 @@
+// 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::any::Any;
 use std::sync::atomic::{AtomicU32, Ordering};
 use std::sync::Arc;
@@ -7,7 +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;
@@ -16,13 +30,14 @@ use table::engine::{EngineContext, TableEngineRef};
 use table::metadata::TableId;
 use table::requests::OpenTableRequest;
 use table::table::numbers::NumbersTable;
+use table::table::TableIdProvider;
 use table::TableRef;
 
 use crate::error::{
-    CatalogNotFoundSnafu, IllegalManagerStateSnafu, OpenTableSnafu, SchemaNotFoundSnafu,
-    SystemCatalogSnafu, SystemCatalogTypeMismatchSnafu, TableExistsSnafu, TableNotFoundSnafu,
+    CatalogNotFoundSnafu, IllegalManagerStateSnafu, OpenTableSnafu, ReadSystemCatalogSnafu, Result,
+    SchemaExistsSnafu, SchemaNotFoundSnafu, SystemCatalogSnafu, SystemCatalogTypeMismatchSnafu,
+    TableExistsSnafu, TableNotFoundSnafu, UnimplementedSnafu,
 };
-use crate::error::{ReadSystemCatalogSnafu, Result};
 use crate::local::memory::{MemoryCatalogManager, MemoryCatalogProvider, MemorySchemaProvider};
 use crate::system::{
     decode_system_catalog, Entry, SystemCatalogTable, TableEntry, ENTRY_TYPE_INDEX, KEY_INDEX,
@@ -31,8 +46,8 @@ use crate::system::{
 use crate::tables::SystemCatalog;
 use crate::{
     format_full_table_name, handle_system_table_request, CatalogList, CatalogManager,
-    CatalogProvider, CatalogProviderRef, RegisterSystemTableRequest, RegisterTableRequest,
-    SchemaProvider,
+    CatalogProvider, CatalogProviderRef, DeregisterTableRequest, RegisterSchemaRequest,
+    RegisterSystemTableRequest, RegisterTableRequest, SchemaProvider, SchemaProviderRef,
 };
 
 /// A `CatalogManager` consists of a system catalog and a bunch of user catalogs.
@@ -42,6 +57,7 @@ pub struct LocalCatalogManager {
     engine: TableEngineRef,
     next_table_id: AtomicU32,
     init_lock: Mutex<bool>,
+    register_lock: Mutex<()>,
     system_table_requests: Mutex<Vec<RegisterSystemTableRequest>>,
 }
 
@@ -61,6 +77,7 @@ 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()),
         })
     }
@@ -128,27 +145,34 @@ 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());
@@ -226,6 +250,7 @@ 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
@@ -278,6 +303,13 @@ impl CatalogList for LocalCatalogManager {
     }
 }
 
+#[async_trait::async_trait]
+impl TableIdProvider for LocalCatalogManager {
+    async fn next_table_id(&self) -> table::Result<TableId> {
+        Ok(self.next_table_id.fetch_add(1, Ordering::Relaxed))
+    }
+}
+
 #[async_trait::async_trait]
 impl CatalogManager for LocalCatalogManager {
     /// Start [LocalCatalogManager] to load all information from system catalog table.
@@ -286,12 +318,7 @@ impl CatalogManager for LocalCatalogManager {
         self.init().await
     }
 
-    #[inline]
-    async fn next_table_id(&self) -> Result<TableId> {
-        Ok(self.next_table_id.fetch_add(1, Ordering::Relaxed))
-    }
-
-    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!(
@@ -311,27 +338,82 @@ impl CatalogManager for LocalCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                schema_info: format!("{catalog_name}.{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();
         }
+    }
 
-        self.system
-            .register_table(
-                catalog_name.clone(),
-                schema_name.clone(),
-                request.table_name.clone(),
-                request.table_id,
-            )
-            .await?;
+    async fn deregister_table(&self, _request: DeregisterTableRequest) -> Result<bool> {
+        UnimplementedSnafu {
+            operation: "deregister table",
+        }
+        .fail()
+    }
 
-        schema.register_table(request.table_name, request.table)?;
-        Ok(1)
+    async fn register_schema(&self, request: RegisterSchemaRequest) -> 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 _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)
+        }
     }
 
     async fn register_system_table(&self, request: RegisterSystemTableRequest) -> Result<()> {
@@ -348,6 +430,15 @@ impl CatalogManager for LocalCatalogManager {
         Ok(())
     }
 
+    fn schema(&self, catalog: &str, schema: &str) -> Result<Option<SchemaProviderRef>> {
+        self.catalogs
+            .catalog(catalog)?
+            .context(CatalogNotFoundSnafu {
+                catalog_name: catalog,
+            })?
+            .schema(schema)
+    }
+
     fn table(
         &self,
         catalog_name: &str,
@@ -361,7 +452,7 @@ impl CatalogManager for LocalCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                schema_info: format!("{catalog_name}.{schema_name}"),
             })?;
         schema.table(table_name)
     }

src/catalog/src/local/memory.rs

@@ -1,20 +1,35 @@
+// 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::any::Any;
 use std::collections::hash_map::Entry;
 use std::collections::HashMap;
 use std::sync::atomic::{AtomicU32, Ordering};
-use std::sync::Arc;
-use std::sync::RwLock;
+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::schema::SchemaProvider;
 use crate::{
-    CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef, RegisterSystemTableRequest,
-    RegisterTableRequest, SchemaProviderRef,
+    CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef, DeregisterTableRequest,
+    RegisterSchemaRequest, RegisterSystemTableRequest, RegisterTableRequest, SchemaProviderRef,
 };
 
 /// Simple in-memory list of catalogs
@@ -41,6 +56,13 @@ impl Default for MemoryCatalogManager {
     }
 }
 
+#[async_trait::async_trait]
+impl TableIdProvider for MemoryCatalogManager {
+    async fn next_table_id(&self) -> table::error::Result<TableId> {
+        Ok(self.table_id.fetch_add(1, Ordering::Relaxed))
+    }
+}
+
 #[async_trait::async_trait]
 impl CatalogManager for MemoryCatalogManager {
     async fn start(&self) -> Result<()> {
@@ -48,11 +70,7 @@ impl CatalogManager for MemoryCatalogManager {
         Ok(())
     }
 
-    async fn next_table_id(&self) -> Result<TableId> {
-        Ok(self.table_id.fetch_add(1, Ordering::Relaxed))
-    }
-
-    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)
@@ -67,11 +85,50 @@ impl CatalogManager for MemoryCatalogManager {
             })?;
         schema
             .register_table(request.table_name, request.table)
-            .map(|v| if v.is_some() { 0 } else { 1 })
+            .map(|v| v.is_none())
+    }
+
+    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 {
+                schema_info: format!("{}.{}", &request.catalog, &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)
+            .context(CatalogNotFoundSnafu {
+                catalog_name: &request.catalog,
+            })?;
+        catalog.register_schema(request.schema, Arc::new(MemorySchemaProvider::new()))?;
+        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>> {
+        let catalogs = self.catalogs.read().unwrap();
+        if let Some(c) = catalogs.get(catalog) {
+            c.schema(schema)
+        } else {
+            Ok(None)
+        }
     }
 
     fn table(&self, catalog: &str, schema: &str, table_name: &str) -> Result<Option<TableRef>> {
@@ -214,11 +271,21 @@ 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 deregister_table(&self, name: &str) -> Result<Option<TableRef>> {
@@ -278,7 +345,7 @@ 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());
@@ -303,4 +370,34 @@ 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

@@ -1,3 +1,17 @@
+// 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::fmt::Debug;
 use std::pin::Pin;
 use std::sync::Arc;

src/catalog/src/remote/client.rs

@@ -1,4 +1,19 @@
+// 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::fmt::Debug;
+use std::sync::Arc;
 
 use async_stream::stream;
 use common_telemetry::info;
@@ -10,7 +25,7 @@ use crate::error::{Error, MetaSrvSnafu};
 use crate::remote::{Kv, KvBackend, ValueIter};
 #[derive(Debug)]
 pub struct MetaKvBackend {
-    pub client: MetaClient,
+    pub client: Arc<MetaClient>,
 }
 
 /// Implement `KvBackend` trait for `MetaKvBackend` instead of opendal's `Accessor` since

src/catalog/src/remote/manager.rs

@@ -1,19 +1,26 @@
+// 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::any::Any;
 use std::collections::HashMap;
 use std::pin::Pin;
 use std::sync::Arc;
-use std::time::Duration;
 
 use arc_swap::ArcSwap;
 use async_stream::stream;
-use backoff::exponential::ExponentialBackoffBuilder;
-use backoff::ExponentialBackoff;
 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, error, info};
+use common_telemetry::{debug, info};
 use futures::Stream;
 use futures_util::StreamExt;
 use snafu::{OptionExt, ResultExt};
@@ -25,14 +32,18 @@ use table::TableRef;
 use tokio::sync::Mutex;
 
 use crate::error::{
-    BumpTableIdSnafu, CatalogNotFoundSnafu, CreateTableSnafu, InvalidCatalogValueSnafu,
-    OpenTableSnafu, ParseTableIdSnafu, SchemaNotFoundSnafu, TableExistsSnafu,
+    CatalogNotFoundSnafu, CreateTableSnafu, InvalidCatalogValueSnafu, InvalidTableSchemaSnafu,
+    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::error::{InvalidTableSchemaSnafu, Result};
 use crate::remote::{Kv, KvBackendRef};
 use crate::{
     handle_system_table_request, CatalogList, CatalogManager, CatalogProvider, CatalogProviderRef,
-    RegisterSystemTableRequest, RegisterTableRequest, SchemaProvider, SchemaProviderRef,
+    DeregisterTableRequest, RegisterSchemaRequest, RegisterSystemTableRequest,
+    RegisterTableRequest, SchemaProvider, SchemaProviderRef,
 };
 
 /// Catalog manager based on metasrv.
@@ -65,6 +76,7 @@ impl RemoteCatalogManager {
 
     fn new_catalog_provider(&self, catalog_name: &str) -> CatalogProviderRef {
         Arc::new(RemoteCatalogProvider {
+            node_id: self.node_id,
             catalog_name: catalog_name.to_string(),
             backend: self.backend.clone(),
             schemas: Default::default(),
@@ -126,8 +138,6 @@ impl RemoteCatalogManager {
     }
 
     /// Iterate over all table entries on metasrv
-    /// TODO(hl): table entries with different version is not currently considered.
-    /// Ideally deprecated table entry must be deleted when deregistering from catalog.
     async fn iter_remote_tables(
         &self,
         catalog_name: &str,
@@ -144,10 +154,10 @@ 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)?;
 
-                debug!(
+                info!(
                     "Found catalog table entry, key: {}, value: {:?}",
                     table_key, table_value
                 );
@@ -232,17 +242,21 @@ impl RemoteCatalogManager {
         schema: SchemaProviderRef,
         mut max_table_id: TableId,
     ) -> Result<()> {
+        info!("initializing tables in {}.{}", catalog_name, schema_name);
+        let mut table_num = 0;
         let mut tables = self.iter_remote_tables(catalog_name, schema_name).await;
         while let Some(r) = tables.next().await {
             let (table_key, table_value) = r?;
             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!(
+            "initialized tables in {}.{}, total: {}",
+            catalog_name, schema_name, table_num
+        );
         Ok(())
     }
 
@@ -294,44 +308,61 @@ 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) => Ok(table),
+            Some(table) => {
+                info!(
+                    "Table opened: {}.{}.{}",
+                    catalog_name, schema_name, table_name
+                );
+                Ok(table)
+            }
             None => {
+                info!(
+                    "Try create table: {}.{}.{}",
+                    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(),
@@ -343,7 +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
                         ),
                     })
             }
@@ -377,64 +408,7 @@ impl CatalogManager for RemoteCatalogManager {
         Ok(())
     }
 
-    /// Bump table id in a CAS manner with backoff.
-    async fn next_table_id(&self) -> Result<TableId> {
-        let key = common_catalog::consts::TABLE_ID_KEY_PREFIX.as_bytes();
-        let op = || async {
-            // TODO(hl): optimize this get
-            let (prev, prev_bytes) = match self.backend.get(key).await? {
-                None => (MIN_USER_TABLE_ID, vec![]),
-                Some(kv) => (parse_table_id(&kv.1)?, kv.1),
-            };
-
-            match self
-                .backend
-                .compare_and_set(key, &prev_bytes, &(prev + 1).to_le_bytes())
-                .await
-            {
-                Ok(cas_res) => match cas_res {
-                    Ok(_) => Ok(prev),
-                    Err(e) => {
-                        info!("Table id {:?} already occupied", e);
-                        Err(backoff::Error::transient(
-                            BumpTableIdSnafu {
-                                msg: "Table id occupied",
-                            }
-                            .build(),
-                        ))
-                    }
-                },
-                Err(e) => {
-                    error!(e;"Failed to CAS table id");
-                    Err(backoff::Error::permanent(
-                        BumpTableIdSnafu {
-                            msg: format!("Failed to perform CAS operation: {:?}", e),
-                        }
-                        .build(),
-                    ))
-                }
-            }
-        };
-
-        let retry_policy: ExponentialBackoff = ExponentialBackoffBuilder::new()
-            .with_initial_interval(Duration::from_millis(4))
-            .with_multiplier(2.0)
-            .with_max_interval(Duration::from_millis(1000))
-            .with_max_elapsed_time(Some(Duration::from_millis(3000)))
-            .build();
-
-        backoff::future::retry(retry_policy, op).await.map_err(|e| {
-            BumpTableIdSnafu {
-                msg: format!(
-                    "Bump table id exceeds max fail times, last error msg: {:?}",
-                    e
-                ),
-            }
-            .build()
-        })
-    }
-
-    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 {
@@ -453,7 +427,25 @@ impl CatalogManager for RemoteCatalogManager {
             .fail();
         }
         schema_provider.register_table(request.table_name, request.table)?;
-        Ok(1)
+        Ok(true)
+    }
+
+    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 {
+            catalog_name: &catalog_name,
+        })?;
+        let schema_provider = self.new_schema_provider(&catalog_name, &schema_name);
+        catalog_provider.register_schema(schema_name, schema_provider)?;
+        Ok(true)
     }
 
     async fn register_system_table(&self, request: RegisterSystemTableRequest) -> Result<()> {
@@ -462,6 +454,14 @@ impl CatalogManager for RemoteCatalogManager {
         Ok(())
     }
 
+    fn schema(&self, catalog: &str, schema: &str) -> Result<Option<SchemaProviderRef>> {
+        self.catalog(catalog)?
+            .context(CatalogNotFoundSnafu {
+                catalog_name: catalog,
+            })?
+            .schema(schema)
+    }
+
     fn table(
         &self,
         catalog_name: &str,
@@ -474,7 +474,7 @@ impl CatalogManager for RemoteCatalogManager {
         let schema = catalog
             .schema(schema_name)?
             .with_context(|| SchemaNotFoundSnafu {
-                schema_info: format!("{}.{}", catalog_name, schema_name),
+                schema_info: format!("{catalog_name}.{schema_name}"),
             })?;
         schema.table(table_name)
     }
@@ -530,6 +530,7 @@ impl CatalogList for RemoteCatalogManager {
 }
 
 pub struct RemoteCatalogProvider {
+    node_id: u64,
     catalog_name: String,
     backend: KvBackendRef,
     schemas: Arc<ArcSwap<HashMap<String, SchemaProviderRef>>>,
@@ -537,8 +538,9 @@ pub struct RemoteCatalogProvider {
 }
 
 impl RemoteCatalogProvider {
-    pub fn new(catalog_name: String, backend: KvBackendRef) -> Self {
+    pub fn new(catalog_name: String, backend: KvBackendRef, node_id: u64) -> Self {
         Self {
+            node_id,
             catalog_name,
             backend,
             schemas: Default::default(),
@@ -546,6 +548,48 @@ impl RemoteCatalogProvider {
         }
     }
 
+    pub fn refresh_schemas(&self) -> Result<()> {
+        let schemas = self.schemas.clone();
+        let schema_prefix = build_schema_prefix(&self.catalog_name);
+        let catalog_name = self.catalog_name.clone();
+        let mutex = self.mutex.clone();
+        let backend = self.backend.clone();
+        let node_id = self.node_id;
+
+        std::thread::spawn(move || {
+            common_runtime::block_on_write(async move {
+                let _guard = mutex.lock().await;
+                let prev_schemas = schemas.load();
+                let mut new_schemas = HashMap::with_capacity(prev_schemas.len() + 1);
+                new_schemas.clone_from(&prev_schemas);
+
+                let mut remote_schemas = backend.range(schema_prefix.as_bytes());
+                while let Some(r) = remote_schemas.next().await {
+                    let Kv(k, _) = r?;
+                    let schema_key = SchemaKey::parse(&String::from_utf8_lossy(&k))
+                        .context(InvalidCatalogValueSnafu)?;
+                    if !new_schemas.contains_key(&schema_key.schema_name) {
+                        new_schemas.insert(
+                            schema_key.schema_name.clone(),
+                            Arc::new(RemoteSchemaProvider::new(
+                                catalog_name.clone(),
+                                schema_key.schema_name,
+                                node_id,
+                                backend.clone(),
+                            )),
+                        );
+                    }
+                }
+                schemas.store(Arc::new(new_schemas));
+                Ok(())
+            })
+        })
+        .join()
+        .unwrap()?;
+
+        Ok(())
+    }
+
     fn build_schema_key(&self, schema_name: impl AsRef<str>) -> SchemaKey {
         SchemaKey {
             catalog_name: self.catalog_name.clone(),
@@ -560,6 +604,7 @@ impl CatalogProvider for RemoteCatalogProvider {
     }
 
     fn schema_names(&self) -> Result<Vec<String>> {
+        self.refresh_schemas()?;
         Ok(self.schemas.load().keys().cloned().collect::<Vec<_>>())
     }
 
@@ -598,20 +643,12 @@ impl CatalogProvider for RemoteCatalogProvider {
     }
 
     fn schema(&self, name: &str) -> Result<Option<Arc<dyn SchemaProvider>>> {
+        // TODO(hl): We should refresh whole catalog before calling datafusion's query engine.
+        self.refresh_schemas()?;
         Ok(self.schemas.load().get(name).cloned())
     }
 }
 
-/// Parse u8 slice to `TableId`
-fn parse_table_id(val: &[u8]) -> Result<TableId> {
-    Ok(TableId::from_le_bytes(val.try_into().map_err(|_| {
-        ParseTableIdSnafu {
-            data: format!("{:?}", val),
-        }
-        .build()
-    })?))
-}
-
 pub struct RemoteSchemaProvider {
     catalog_name: String,
     schema_name: String,
@@ -733,17 +770,3 @@ impl SchemaProvider for RemoteSchemaProvider {
         Ok(self.tables.load().contains_key(name))
     }
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn test_parse_table_id() {
-        assert_eq!(12, parse_table_id(&12_i32.to_le_bytes()).unwrap());
-        let mut data = vec![];
-        data.extend_from_slice(&12_i32.to_le_bytes());
-        data.push(0);
-        assert!(parse_table_id(&data).is_err());
-    }
-}

src/catalog/src/schema.rs

@@ -1,3 +1,17 @@
+// 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::any::Any;
 use std::sync::Arc;
 

src/catalog/src/system.rs

@@ -1,21 +1,33 @@
+// 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::any::Any;
 use std::collections::HashMap;
 use std::sync::Arc;
 
 use common_catalog::consts::{
-    INFORMATION_SCHEMA_NAME, SYSTEM_CATALOG_NAME, SYSTEM_CATALOG_TABLE_ID,
-    SYSTEM_CATALOG_TABLE_NAME,
+    DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, INFORMATION_SCHEMA_NAME, SYSTEM_CATALOG_NAME,
+    SYSTEM_CATALOG_TABLE_ID, SYSTEM_CATALOG_TABLE_NAME,
 };
 use common_query::logical_plan::Expr;
-use common_query::physical_plan::PhysicalPlanRef;
-use common_query::physical_plan::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};
@@ -30,7 +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 {
@@ -50,7 +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> {
@@ -74,6 +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();
@@ -97,7 +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(),
             };
@@ -114,14 +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)
     }
@@ -149,9 +161,10 @@ fn build_system_catalog_schema() -> Schema {
         ),
         ColumnSchema::new(
             "timestamp".to_string(),
-            ConcreteDataType::timestamp_millis_datatype(),
+            ConcreteDataType::timestamp_millisecond_datatype(),
             false,
-        ),
+        )
+        .with_time_index(true),
         ColumnSchema::new(
             "value".to_string(),
             ConcreteDataType::binary_datatype(),
@@ -159,66 +172,78 @@ 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,
         ),
     ];
 
     // The schema of this table must be valid.
-    SchemaBuilder::try_from(cols)
-        .unwrap()
-        .timestamp_index(Some(2))
-        .build()
-        .unwrap()
+    SchemaBuilder::try_from(cols).unwrap().build().unwrap()
 }
 
 pub fn build_table_insert_request(full_table_name: String, table_id: TableId) -> InsertRequest {
+    build_insert_request(
+        EntryType::Table,
+        full_table_name.as_bytes(),
+        serde_json::to_string(&TableEntryValue { table_id })
+            .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}");
+    build_insert_request(
+        EntryType::Schema,
+        full_schema_name.as_bytes(),
+        serde_json::to_string(&SchemaEntryValue {})
+            .unwrap()
+            .as_bytes(),
+    )
+}
+
+pub fn build_insert_request(entry_type: EntryType, key: &[u8], value: &[u8]) -> InsertRequest {
     let mut columns_values = HashMap::with_capacity(6);
     columns_values.insert(
         "entry_type".to_string(),
-        Arc::new(UInt8Vector::from_slice(&[EntryType::Table as u8])) as _,
+        Arc::new(UInt8Vector::from_slice(&[entry_type as u8])) as _,
     );
 
     columns_values.insert(
         "key".to_string(),
-        Arc::new(BinaryVector::from_slice(&[full_table_name.as_bytes()])) as _,
+        Arc::new(BinaryVector::from_slice(&[key])) as _,
     );
 
     // 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(
         "value".to_string(),
-        Arc::new(BinaryVector::from_slice(&[serde_json::to_string(
-            &TableEntryValue { table_id },
-        )
-        .unwrap()
-        .as_bytes()])) as _,
+        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 {
+        catalog_name: DEFAULT_CATALOG_NAME.to_string(),
+        schema_name: DEFAULT_SCHEMA_NAME.to_string(),
         table_name: SYSTEM_CATALOG_TABLE_NAME.to_string(),
         columns_values,
     }
@@ -324,6 +349,9 @@ pub struct SchemaEntry {
     pub schema_name: String,
 }
 
+#[derive(Debug, Serialize, Deserialize, PartialEq, Eq)]
+pub struct SchemaEntryValue;
+
 #[derive(Debug, PartialEq, Eq, Ord, PartialOrd)]
 pub struct TableEntry {
     pub catalog_name: String,
@@ -339,20 +367,20 @@ pub struct TableEntryValue {
 
 #[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()),
@@ -362,7 +390,7 @@ mod tests {
         if let Entry::Catalog(e) = entry {
             assert_eq!("some_catalog", e.catalog_name);
         } else {
-            panic!("Unexpected type: {:?}", entry);
+            panic!("Unexpected type: {entry:?}");
         }
     }
 
@@ -379,7 +407,7 @@ 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:?}");
         }
     }
 
@@ -398,7 +426,7 @@ mod tests {
             assert_eq!("some_table", e.table_name);
             assert_eq!(42, e.table_id);
         } else {
-            panic!("Unexpected type: {:?}", entry);
+            panic!("Unexpected type: {entry:?}");
         }
     }
 
@@ -424,7 +452,7 @@ 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

@@ -1,3 +1,17 @@
+// 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.
+
 // The `tables` table in system catalog keeps a record of all tables created by user.
 
 use std::any::Any;
@@ -12,9 +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;
@@ -24,8 +38,8 @@ use table::metadata::{TableId, TableInfoRef};
 use table::table::scan::SimpleTableScan;
 use table::{Table, TableRef};
 
-use crate::error::{Error, InsertTableRecordSnafu};
-use crate::system::{build_table_insert_request, SystemCatalogTable};
+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,
 };
@@ -63,7 +77,7 @@ 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> {
@@ -135,26 +149,33 @@ 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(),
     ]
 }
 
@@ -254,7 +275,20 @@ impl SystemCatalog {
             .system
             .insert(request)
             .await
-            .context(InsertTableRecordSnafu)
+            .context(InsertCatalogRecordSnafu)
+    }
+
+    pub async fn register_schema(
+        &self,
+        catalog: String,
+        schema: String,
+    ) -> crate::error::Result<usize> {
+        let request = build_schema_insert_request(catalog, schema);
+        self.information_schema
+            .system
+            .insert(request)
+            .await
+            .context(InsertCatalogRecordSnafu)
     }
 }
 
@@ -313,9 +347,7 @@ 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;
 
@@ -338,58 +370,48 @@ 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 +1,131 @@
+// 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.
+
+#[cfg(test)]
+mod tests {
+    use std::sync::Arc;
+
+    use catalog::local::LocalCatalogManager;
+    use catalog::{CatalogManager, RegisterTableRequest};
+    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_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

@@ -1,3 +1,17 @@
+// 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::btree_map::Entry;
 use std::collections::{BTreeMap, HashMap};
 use std::fmt::{Display, Formatter};
@@ -13,7 +27,7 @@ use datatypes::data_type::ConcreteDataType;
 use datatypes::schema::{ColumnSchema, Schema};
 use datatypes::vectors::StringVector;
 use serde::Serializer;
-use table::engine::{EngineContext, TableEngine};
+use table::engine::{EngineContext, TableEngine, TableReference};
 use table::metadata::TableId;
 use table::requests::{AlterTableRequest, CreateTableRequest, DropTableRequest, OpenTableRequest};
 use table::test_util::MemTable;
@@ -151,6 +165,7 @@ impl TableEngine for MockTableEngine {
             table_id,
             catalog_name,
             schema_name,
+            vec![0],
         )) as Arc<_>;
 
         let mut tables = self.tables.write().await;
@@ -174,19 +189,35 @@ impl TableEngine for MockTableEngine {
         unimplemented!()
     }
 
-    fn get_table(&self, _ctx: &EngineContext, name: &str) -> table::Result<Option<TableRef>> {
-        futures::executor::block_on(async { Ok(self.tables.read().await.get(name).cloned()) })
+    fn get_table(
+        &self,
+        _ctx: &EngineContext,
+        table_ref: &TableReference,
+    ) -> table::Result<Option<TableRef>> {
+        futures::executor::block_on(async {
+            Ok(self
+                .tables
+                .read()
+                .await
+                .get(&table_ref.to_string())
+                .cloned())
+        })
     }
 
-    fn table_exists(&self, _ctx: &EngineContext, name: &str) -> bool {
-        futures::executor::block_on(async { self.tables.read().await.contains_key(name) })
+    fn table_exists(&self, _ctx: &EngineContext, table_ref: &TableReference) -> bool {
+        futures::executor::block_on(async {
+            self.tables
+                .read()
+                .await
+                .contains_key(&table_ref.to_string())
+        })
     }
 
     async fn drop_table(
         &self,
         _ctx: &EngineContext,
         _request: DropTableRequest,
-    ) -> table::Result<()> {
+    ) -> table::Result<bool> {
         unimplemented!()
     }
 }

src/catalog/tests/remote_catalog_tests.rs

@@ -1,3 +1,17 @@
+// 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.
+
 #![feature(assert_matches)]
 
 mod mock;
@@ -8,12 +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::{CatalogManager, CatalogManagerRef, RegisterTableRequest};
-    use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, MIN_USER_TABLE_ID};
-    use common_catalog::{CatalogKey, CatalogValue, SchemaKey, SchemaValue};
+    use catalog::{CatalogList, CatalogManager, RegisterTableRequest};
+    use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
     use datatypes::schema::Schema;
     use futures_util::StreamExt;
     use table::engine::{EngineContext, TableEngineRef};
@@ -61,7 +75,9 @@ mod tests {
         );
     }
 
-    async fn prepare_components(node_id: u64) -> (KvBackendRef, TableEngineRef, CatalogManagerRef) {
+    async fn prepare_components(
+        node_id: u64,
+    ) -> (KvBackendRef, TableEngineRef, Arc<RemoteCatalogManager>) {
         let backend = Arc::new(MockKvBackend::default()) as KvBackendRef;
         let table_engine = Arc::new(MockTableEngine::default());
         let catalog_manager =
@@ -186,7 +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
@@ -207,6 +223,7 @@ mod tests {
         let catalog = Arc::new(RemoteCatalogProvider::new(
             catalog_name.clone(),
             backend.clone(),
+            node_id,
         ));
 
         // register catalog to catalog manager
@@ -270,26 +287,11 @@ 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()]),
             new_catalog.schema_names().unwrap().into_iter().collect()
         )
     }
-
-    #[tokio::test]
-    async fn test_next_table_id() {
-        let node_id = 42;
-        let (_, _, catalog_manager) = prepare_components(node_id).await;
-        assert_eq!(
-            MIN_USER_TABLE_ID,
-            catalog_manager.next_table_id().await.unwrap()
-        );
-
-        assert_eq!(
-            MIN_USER_TABLE_ID + 1,
-            catalog_manager.next_table_id().await.unwrap()
-        );
-    }
 }

src/client/Cargo.toml

@@ -1,8 +1,8 @@
 [package]
 name = "client"
-version = "0.1.0"
-edition = "2021"
-# 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" }
@@ -10,12 +10,11 @@ async-stream = "0.3"
 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-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"
 parking_lot = "0.12"

src/client/examples/insert.rs

@@ -1,6 +1,18 @@
-use std::collections::HashMap;
+// 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::{codec::InsertBatch, *};
+use api::v1::*;
 use client::{Client, Database};
 
 fn main() {
@@ -15,19 +27,21 @@ async fn run() {
     let client = Client::with_urls(vec!["127.0.0.1:3001"]);
     let db = Database::new("greptime", client);
 
+    let (columns, row_count) = insert_data();
+
     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,
+        columns,
+        row_count,
     };
     db.insert(expr).await.unwrap();
 }
 
-fn insert_batches() -> Vec<Vec<u8>> {
+fn insert_data() -> (Vec<Column>, u32) {
     const SEMANTIC_TAG: i32 = 0;
-    const SEMANTIC_FEILD: i32 = 1;
+    const SEMANTIC_FIELD: i32 = 1;
     const SEMANTIC_TS: i32 = 2;
 
     let row_count = 4;
@@ -55,7 +69,7 @@ fn insert_batches() -> Vec<Vec<u8>> {
     };
     let cpu_column = Column {
         column_name: "cpu".to_string(),
-        semantic_type: SEMANTIC_FEILD,
+        semantic_type: SEMANTIC_FIELD,
         values: Some(cpu_vals),
         null_mask: vec![2],
         ..Default::default()
@@ -67,7 +81,7 @@ fn insert_batches() -> Vec<Vec<u8>> {
     };
     let mem_column = Column {
         column_name: "memory".to_string(),
-        semantic_type: SEMANTIC_FEILD,
+        semantic_type: SEMANTIC_FIELD,
         values: Some(mem_vals),
         null_mask: vec![4],
         ..Default::default()
@@ -85,9 +99,8 @@ fn insert_batches() -> Vec<Vec<u8>> {
         ..Default::default()
     };
 
-    let insert_batch = InsertBatch {
-        columns: vec![host_column, cpu_column, mem_column, ts_column],
+    (
+        vec![host_column, cpu_column, mem_column, ts_column],
         row_count,
-    };
-    vec![insert_batch.into()]
+    )
 }

src/client/examples/logical.rs

@@ -1,12 +1,25 @@
-use api::v1::{ColumnDataType, ColumnDef, CreateExpr};
-use client::{admin::Admin, Client, Database};
+// 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::{ColumnDataType, ColumnDef, CreateTableExpr, TableId};
+use client::admin::Admin;
+use client::{Client, Database};
 use prost_09::Message;
-use substrait_proto::protobuf::{
-    plan_rel::RelType as PlanRelType,
-    read_rel::{NamedTable, ReadType},
-    rel::RelType,
-    PlanRel, ReadRel, Rel,
-};
+use substrait_proto::protobuf::plan_rel::RelType as PlanRelType;
+use substrait_proto::protobuf::read_rel::{NamedTable, ReadType};
+use substrait_proto::protobuf::rel::RelType;
+use substrait_proto::protobuf::{PlanRel, ReadRel, Rel};
 use tracing::{event, Level};
 
 fn main() {
@@ -20,35 +33,37 @@ 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(TableId { id: 1024 }),
+        region_ids: vec![0],
     };
 
     let admin = Admin::new("create table", client.clone());

src/client/examples/physical.rs

@@ -1,37 +0,0 @@
-use std::sync::Arc;
-
-use client::{Client, Database};
-use common_grpc::MockExecution;
-use datafusion::physical_plan::{
-    expressions::Column, projection::ProjectionExec, 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

@@ -1,3 +1,17 @@
+// 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};
 

src/client/src/admin.rs

@@ -1,12 +1,24 @@
+// 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;
-use crate::Client;
-use crate::Result;
+use crate::{error, Client, Result};
 
 #[derive(Clone, Debug)]
 pub struct Admin {
@@ -22,13 +34,13 @@ impl Admin {
         }
     }
 
-    pub async fn create(&self, expr: CreateExpr) -> Result<AdminResult> {
+    pub async fn create(&self, expr: CreateTableExpr) -> Result<AdminResult> {
         let header = ExprHeader {
             version: PROTOCOL_VERSION,
         };
         let expr = AdminExpr {
             header: Some(header),
-            expr: Some(admin_expr::Expr::Create(expr)),
+            expr: Some(admin_expr::Expr::CreateTable(expr)),
         };
         self.do_request(expr).await
     }
@@ -46,7 +58,19 @@ impl Admin {
             header: Some(header),
             expr: Some(admin_expr::Expr::Alter(expr)),
         };
-        Ok(self.do_requests(vec![expr]).await?.remove(0))
+        self.do_request(expr).await
+    }
+
+    pub async fn drop_table(&self, expr: DropTableExpr) -> Result<AdminResult> {
+        let header = ExprHeader {
+            version: PROTOCOL_VERSION,
+        };
+        let expr = AdminExpr {
+            header: Some(header),
+            expr: Some(admin_expr::Expr::DropTable(expr)),
+        };
+
+        self.do_request(expr).await
     }
 
     /// Invariants: the lengths of input vec (`Vec<AdminExpr>`) and output vec (`Vec<AdminResult>`) are equal.
@@ -70,6 +94,17 @@ impl Admin {
         );
         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> {

src/client/src/client.rs

@@ -1,17 +1,28 @@
+// 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 api::v1::greptime_client::GreptimeClient;
 use api::v1::*;
 use common_grpc::channel_manager::ChannelManager;
 use parking_lot::RwLock;
-use snafu::OptionExt;
-use snafu::ResultExt;
+use snafu::{OptionExt, ResultExt};
 use tonic::transport::Channel;
 
-use crate::error;
-use crate::load_balance::LoadBalance;
-use crate::load_balance::Loadbalancer;
-use crate::Result;
+use crate::load_balance::{LoadBalance, Loadbalancer};
+use crate::{error, Result};
 
 #[derive(Clone, Debug, Default)]
 pub struct Client {
@@ -128,8 +139,11 @@ impl Client {
             .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)?;
+        let mut client = self.make_client(&peer)?;
+        let result = client
+            .batch(req)
+            .await
+            .context(error::TonicStatusSnafu { addr: peer })?;
         Ok(result.into_inner())
     }
 

src/client/src/database.rs

@@ -1,31 +1,35 @@
+// 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 api::helper::ColumnDataTypeWrapper;
 use api::v1::codec::SelectResult as GrpcSelectResult;
+use api::v1::column::SemanticType;
 use api::v1::{
-    column::Values, object_expr, object_result, select_expr, Column, ColumnDataType,
-    DatabaseRequest, ExprHeader, InsertExpr, MutateResult as GrpcMutateResult, ObjectExpr,
-    ObjectResult as GrpcObjectResult, PhysicalPlan, SelectExpr,
+    object_expr, object_result, select_expr, DatabaseRequest, ExprHeader, InsertExpr,
+    MutateResult as GrpcMutateResult, ObjectExpr, ObjectResult as GrpcObjectResult, SelectExpr,
 };
-use common_base::BitVec;
 use common_error::status_code::StatusCode;
-use common_grpc::AsExcutionPlan;
-use common_grpc::DefaultAsPlanImpl;
+use common_grpc_expr::column_to_vector;
 use common_query::Output;
 use common_recordbatch::{RecordBatch, RecordBatches};
-use common_time::date::Date;
-use common_time::datetime::DateTime;
-use common_time::timestamp::Timestamp;
-use datafusion::physical_plan::ExecutionPlan;
 use datatypes::prelude::*;
 use datatypes::schema::{ColumnSchema, Schema};
 use snafu::{ensure, OptionExt, ResultExt};
 
-use crate::error;
-use crate::{
-    error::{ConvertSchemaSnafu, DatanodeSnafu, DecodeSelectSnafu, EncodePhysicalSnafu},
-    Client, Result,
-};
+use crate::error::{ColumnToVectorSnafu, ConvertSchemaSnafu, DatanodeSnafu, DecodeSelectSnafu};
+use crate::{error, Client, Result};
 
 pub const PROTOCOL_VERSION: u32 = 1;
 
@@ -85,24 +89,6 @@ impl Database {
         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)),
@@ -124,8 +110,6 @@ impl Database {
         obj_result.try_into()
     }
 
-    // TODO(jiachun) update/delete
-
     pub async fn object(&self, expr: ObjectExpr) -> Result<GrpcObjectResult> {
         let res = self.objects(vec![expr]).await?.pop().unwrap();
         Ok(res)
@@ -201,7 +185,9 @@ impl TryFrom<ObjectResult> for Output {
                 let vectors = select
                     .columns
                     .iter()
-                    .map(|column| column_to_vector(column, select.row_count))
+                    .map(|column| {
+                        column_to_vector(column, select.row_count).context(ColumnToVectorSnafu)
+                    })
                     .collect::<Result<Vec<VectorRef>>>()?;
 
                 let column_schemas = select
@@ -211,7 +197,12 @@ impl TryFrom<ObjectResult> for Output {
                     .map(|(column, vector)| {
                         let datatype = vector.data_type();
                         // nullable or not, does not affect the output
-                        ColumnSchema::new(&column.column_name, datatype, true)
+                        let mut column_schema =
+                            ColumnSchema::new(&column.column_name, datatype, true);
+                        if column.semantic_type == SemanticType::Timestamp as i32 {
+                            column_schema = column_schema.with_time_index(true);
+                        }
+                        column_schema
                     })
                     .collect::<Vec<ColumnSchema>>();
 
@@ -239,100 +230,11 @@ impl TryFrom<ObjectResult> for Output {
     }
 }
 
-fn column_to_vector(column: &Column, rows: u32) -> Result<VectorRef> {
-    let wrapper =
-        ColumnDataTypeWrapper::try_new(column.datatype).context(error::ColumnDataTypeSnafu)?;
-    let column_datatype = wrapper.datatype();
-
-    let rows = rows as usize;
-    let mut vector = VectorBuilder::with_capacity(wrapper.into(), rows);
-
-    if let Some(values) = &column.values {
-        let values = collect_column_values(column_datatype, values);
-        let mut values_iter = values.into_iter();
-
-        let null_mask = BitVec::from_slice(&column.null_mask);
-        let mut nulls_iter = null_mask.iter().by_vals().fuse();
-
-        for i in 0..rows {
-            if let Some(true) = nulls_iter.next() {
-                vector.push_null();
-            } else {
-                let value_ref = values_iter.next().context(error::InvalidColumnProtoSnafu {
-                    err_msg: format!(
-                        "value not found at position {} of column {}",
-                        i, &column.column_name
-                    ),
-                })?;
-                vector
-                    .try_push_ref(value_ref)
-                    .context(error::CreateVectorSnafu)?;
-            }
-        }
-    } else {
-        (0..rows).for_each(|_| vector.push_null());
-    }
-    Ok(vector.finish())
-}
-
-fn collect_column_values(column_datatype: ColumnDataType, values: &Values) -> Vec<ValueRef> {
-    macro_rules! collect_values {
-        ($value: expr, $mapper: expr) => {
-            $value.iter().map($mapper).collect::<Vec<ValueRef>>()
-        };
-    }
-
-    match column_datatype {
-        ColumnDataType::Boolean => collect_values!(values.bool_values, |v| ValueRef::from(*v)),
-        ColumnDataType::Int8 => collect_values!(values.i8_values, |v| ValueRef::from(*v as i8)),
-        ColumnDataType::Int16 => {
-            collect_values!(values.i16_values, |v| ValueRef::from(*v as i16))
-        }
-        ColumnDataType::Int32 => {
-            collect_values!(values.i32_values, |v| ValueRef::from(*v))
-        }
-        ColumnDataType::Int64 => {
-            collect_values!(values.i64_values, |v| ValueRef::from(*v as i64))
-        }
-        ColumnDataType::Uint8 => {
-            collect_values!(values.u8_values, |v| ValueRef::from(*v as u8))
-        }
-        ColumnDataType::Uint16 => {
-            collect_values!(values.u16_values, |v| ValueRef::from(*v as u16))
-        }
-        ColumnDataType::Uint32 => {
-            collect_values!(values.u32_values, |v| ValueRef::from(*v))
-        }
-        ColumnDataType::Uint64 => {
-            collect_values!(values.u64_values, |v| ValueRef::from(*v as u64))
-        }
-        ColumnDataType::Float32 => collect_values!(values.f32_values, |v| ValueRef::from(*v)),
-        ColumnDataType::Float64 => collect_values!(values.f64_values, |v| ValueRef::from(*v)),
-        ColumnDataType::Binary => {
-            collect_values!(values.binary_values, |v| ValueRef::from(v.as_slice()))
-        }
-        ColumnDataType::String => {
-            collect_values!(values.string_values, |v| ValueRef::from(v.as_str()))
-        }
-        ColumnDataType::Date => {
-            collect_values!(values.date_values, |v| ValueRef::Date(Date::new(*v)))
-        }
-        ColumnDataType::Datetime => {
-            collect_values!(values.datetime_values, |v| ValueRef::DateTime(
-                DateTime::new(*v)
-            ))
-        }
-        ColumnDataType::Timestamp => {
-            collect_values!(values.ts_millis_values, |v| ValueRef::Timestamp(
-                Timestamp::from_millis(*v)
-            ))
-        }
-    }
-}
-
 #[cfg(test)]
 mod tests {
-    use datanode::server::grpc::select::{null_mask, values};
+    use api::helper::ColumnDataTypeWrapper;
+    use api::v1::Column;
+    use common_grpc::select::{null_mask, values};
     use datatypes::vectors::{
         BinaryVector, BooleanVector, DateTimeVector, DateVector, Float32Vector, Float64Vector,
         Int16Vector, Int32Vector, Int64Vector, Int8Vector, StringVector, UInt16Vector,
@@ -416,12 +318,11 @@ 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

@@ -1,3 +1,17 @@
+// 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::any::Any;
 use std::sync::Arc;
 
@@ -25,8 +39,9 @@ pub enum Error {
     #[snafu(display("Missing result header"))]
     MissingHeader,
 
-    #[snafu(display("Tonic internal error, source: {}", source))]
+    #[snafu(display("Tonic internal error, addr: {}, source: {}", addr, source))]
     TonicStatus {
+        addr: String,
         source: tonic::Status,
         backtrace: Backtrace,
     },
@@ -47,24 +62,12 @@ pub enum Error {
     #[snafu(display("Mutate result has failure {}", failure))]
     MutateFailure { failure: u32, backtrace: Backtrace },
 
-    #[snafu(display("Invalid column proto: {}", err_msg))]
-    InvalidColumnProto {
-        err_msg: String,
-        backtrace: Backtrace,
-    },
-
     #[snafu(display("Column datatype error, source: {}", source))]
     ColumnDataType {
         #[snafu(backtrace)]
         source: api::error::Error,
     },
 
-    #[snafu(display("Failed to create vector, source: {}", source))]
-    CreateVector {
-        #[snafu(backtrace)]
-        source: datatypes::error::Error,
-    },
-
     #[snafu(display("Failed to create RecordBatches, source: {}", source))]
     CreateRecordBatches {
         #[snafu(backtrace)]
@@ -96,6 +99,12 @@ pub enum Error {
         #[snafu(backtrace)]
         source: common_grpc::error::Error,
     },
+
+    #[snafu(display("Failed to convert column to vector, source: {}", source))]
+    ColumnToVector {
+        #[snafu(backtrace)]
+        source: common_grpc_expr::error::Error,
+    },
 }
 
 pub type Result<T> = std::result::Result<T, Error>;
@@ -111,15 +120,13 @@ impl ErrorExt for Error {
             | Error::Datanode { .. }
             | Error::EncodePhysical { .. }
             | Error::MutateFailure { .. }
-            | Error::InvalidColumnProto { .. }
             | Error::ColumnDataType { .. }
             | Error::MissingField { .. } => StatusCode::Internal,
-            Error::ConvertSchema { source } | Error::CreateVector { source } => {
-                source.status_code()
-            }
+            Error::ConvertSchema { source } => source.status_code(),
             Error::CreateRecordBatches { source } => source.status_code(),
             Error::CreateChannel { source, .. } => source.status_code(),
             Error::IllegalGrpcClientState { .. } => StatusCode::Unexpected,
+            Error::ColumnToVector { source, .. } => source.status_code(),
         }
     }
 

src/client/src/lib.rs

@@ -1,3 +1,17 @@
+// 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.
+
 pub mod admin;
 mod client;
 mod database;
@@ -6,8 +20,6 @@ pub mod load_balance;
 
 pub use api;
 
-pub use self::{
-    client::Client,
-    database::{Database, ObjectResult, Select},
-    error::{Error, Result},
-};
+pub use self::client::Client;
+pub use self::database::{Database, ObjectResult, Select};
+pub use self::error::{Error, Result};

src/client/src/load_balance.rs

@@ -1,3 +1,17 @@
+// 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 enum_dispatch::enum_dispatch;
 use rand::seq::SliceRandom;
 

src/cmd/Cargo.toml

@@ -1,7 +1,8 @@
 [package]
 name = "cmd"
-version = "0.1.0"
-edition = "2021"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
 default-run = "greptime"
 
 [[bin]]
@@ -9,6 +10,7 @@ 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 = [
@@ -17,7 +19,10 @@ common-telemetry = { path = "../common/telemetry", features = [
 datanode = { path = "../datanode" }
 frontend = { path = "../frontend" }
 futures = "0.3"
+meta-client = { path = "../meta-client" }
 meta-srv = { path = "../meta-srv" }
+serde = "1.0"
+servers = { path = "../servers" }
 snafu = { version = "0.7", features = ["backtraces"] }
 tokio = { version = "1.18", features = ["full"] }
 toml = "0.5"
@@ -25,3 +30,6 @@ toml = "0.5"
 [dev-dependencies]
 serde = "1.0"
 tempdir = "0.3"
+
+[build-dependencies]
+build-data = "0.1.3"

src/cmd/build.rs

@@ -0,0 +1,29 @@
+// 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.
+
+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

@@ -1,15 +1,26 @@
+// 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::fmt;
 
 use clap::Parser;
-use cmd::datanode;
 use cmd::error::Result;
-use cmd::frontend;
-use cmd::metasrv;
-use common_telemetry::logging::error;
-use common_telemetry::logging::info;
+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,
@@ -33,6 +44,8 @@ enum SubCommand {
     Frontend(frontend::Command),
     #[clap(name = "metasrv")]
     Metasrv(metasrv::Command),
+    #[clap(name = "standalone")]
+    Standalone(standalone::Command),
 }
 
 impl SubCommand {
@@ -41,6 +54,7 @@ impl SubCommand {
             SubCommand::Datanode(cmd) => cmd.run().await,
             SubCommand::Frontend(cmd) => cmd.run().await,
             SubCommand::Metasrv(cmd) => cmd.run().await,
+            SubCommand::Standalone(cmd) => cmd.run().await,
         }
     }
 }
@@ -51,10 +65,24 @@ impl fmt::Display for SubCommand {
             SubCommand::Datanode(..) => write!(f, "greptime-datanode"),
             SubCommand::Frontend(..) => write!(f, "greptime-frontend"),
             SubCommand::Metasrv(..) => write!(f, "greptime-metasrv"),
+            SubCommand::Standalone(..) => write!(f, "greptime-standalone"),
         }
     }
 }
 
+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

@@ -1,6 +1,22 @@
+// 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 clap::Parser;
 use common_telemetry::logging;
-use datanode::datanode::{Datanode, DatanodeOptions, Mode};
+use datanode::datanode::{Datanode, DatanodeOptions, ObjectStoreConfig};
+use meta_client::MetaClientOpts;
+use servers::Mode;
 use snafu::ResultExt;
 
 use crate::error::{Error, MissingConfigSnafu, Result, StartDatanodeSnafu};
@@ -31,22 +47,22 @@ impl SubCommand {
     }
 }
 
-#[derive(Debug, Parser)]
+#[derive(Debug, Parser, Default)]
 struct StartCommand {
     #[clap(long)]
     node_id: Option<u64>,
     #[clap(long)]
-    http_addr: Option<String>,
-    #[clap(long)]
     rpc_addr: Option<String>,
     #[clap(long)]
     mysql_addr: Option<String>,
     #[clap(long)]
-    postgres_addr: Option<String>,
-    #[clap(long)]
     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 {
@@ -75,43 +91,41 @@ impl TryFrom<StartCommand> for DatanodeOptions {
             DatanodeOptions::default()
         };
 
-        if let Some(addr) = cmd.http_addr {
-            opts.http_addr = addr;
-        }
         if let Some(addr) = cmd.rpc_addr {
             opts.rpc_addr = addr;
         }
         if let Some(addr) = cmd.mysql_addr {
             opts.mysql_addr = addr;
         }
-        if let Some(addr) = cmd.postgres_addr {
-            opts.postgres_addr = addr;
+
+        if let Some(node_id) = cmd.node_id {
+            opts.node_id = Some(node_id);
         }
 
-        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;
-                opts.node_id = node_id;
-                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(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)
     }
@@ -119,45 +133,44 @@ impl TryFrom<StartCommand> for DatanodeOptions {
 
 #[cfg(test)]
 mod tests {
+    use std::assert_matches::assert_matches;
+
     use datanode::datanode::ObjectStoreConfig;
+    use servers::Mode;
 
     use super::*;
 
     #[test]
     fn test_read_from_config_file() {
         let cmd = StartCommand {
-            node_id: None,
-            http_addr: None,
-            rpc_addr: None,
-            mysql_addr: None,
-            postgres_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!("0.0.0.0:3000".to_string(), options.http_addr);
-        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!("0.0.0.0:5432".to_string(), options.postgres_addr);
-        assert_eq!(4, options.postgres_runtime_size);
+        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!(),
         };
     }
 
@@ -165,53 +178,54 @@ mod tests {
     fn test_try_from_cmd() {
         assert_eq!(
             Mode::Standalone,
-            DatanodeOptions::try_from(StartCommand {
-                node_id: None,
-                http_addr: None,
-                rpc_addr: None,
-                mysql_addr: None,
-                postgres_addr: None,
-                metasrv_addr: None,
-                config_file: None
-            })
-            .unwrap()
-            .mode
+            DatanodeOptions::try_from(StartCommand::default())
+                .unwrap()
+                .mode
         );
 
-        assert_eq!(
-            Mode::Distributed,
-            DatanodeOptions::try_from(StartCommand {
-                node_id: Some(42),
-                http_addr: None,
-                rpc_addr: None,
-                mysql_addr: None,
-                postgres_addr: None,
-                metasrv_addr: Some("127.0.0.1:3002".to_string()),
-                config_file: None
-            })
-            .unwrap()
-            .mode
-        );
-
-        assert!(DatanodeOptions::try_from(StartCommand {
-            node_id: None,
-            http_addr: None,
-            rpc_addr: None,
-            mysql_addr: None,
-            postgres_addr: None,
-            metasrv_addr: Some("127.0.0.1:3002".to_string()),
-            config_file: None,
-        })
-        .is_err());
-        assert!(DatanodeOptions::try_from(StartCommand {
+        let mode = DatanodeOptions::try_from(StartCommand {
             node_id: Some(42),
-            http_addr: None,
-            rpc_addr: None,
-            mysql_addr: None,
-            postgres_addr: None,
-            metasrv_addr: None,
-            config_file: None,
+            metasrv_addr: Some("127.0.0.1:3002".to_string()),
+            ..Default::default()
+        })
+        .unwrap()
+        .mode;
+        assert_matches!(mode, Mode::Distributed);
+
+        assert!(DatanodeOptions::try_from(StartCommand {
+            metasrv_addr: Some("127.0.0.1:3002".to_string()),
+            ..Default::default()
         })
         .is_err());
+
+        // Providing node_id but leave metasrv_addr absent is ok since metasrv_addr has default value
+        DatanodeOptions::try_from(StartCommand {
+            node_id: Some(42),
+            ..Default::default()
+        })
+        .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!(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

@@ -1,3 +1,17 @@
+// 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::any::Any;
 
 use common_error::prelude::*;
@@ -38,6 +52,15 @@ pub enum Error {
 
     #[snafu(display("Missing config, msg: {}", msg))]
     MissingConfig { msg: String, backtrace: Backtrace },
+
+    #[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>;
@@ -51,6 +74,8 @@ impl ErrorExt for Error {
             Error::ReadConfig { .. } | Error::ParseConfig { .. } | Error::MissingConfig { .. } => {
                 StatusCode::InvalidArguments
             }
+            Error::IllegalConfig { .. } => StatusCode::InvalidArguments,
+            Error::IllegalAuthConfig { .. } => StatusCode::InvalidArguments,
         }
     }
 
@@ -72,10 +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()