refactor: unify the greptime artifacts building (#2015)

* refactor: unify the make targets of building images

* refactor: make Dockerfile more clean

1. Add dev-builder image to build greptime binary easily;
2. Add 'docker/ci/Dockerfile-centos' to release centos image;
3. Delete Dockerfile of aarch64 and just need to use one Dockerfile;

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

---------

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

.github/workflows/release.yml

@@ -399,7 +399,7 @@ jobs:
         uses: docker/build-push-action@v3
         if: success() || steps.unzip-arm64.conclusion == 'success' # Build and push all platform if unzip-arm64 succeeds
         with:
-          context: ./docker/ci/
+          context: .
           file: ./docker/ci/Dockerfile
           push: true
           platforms: linux/amd64,linux/arm64
@@ -411,7 +411,7 @@ jobs:
         uses: docker/build-push-action@v3
         if: success() || steps.download-arm64.conclusion == 'failure' # Only build and push amd64 platform if download-arm64 fails
         with:
-          context: ./docker/ci/
+          context: .
           file: ./docker/ci/Dockerfile
           push: true
           platforms: linux/amd64

Makefile

@@ -1,15 +1,67 @@
-IMAGE_REGISTRY ?= greptimedb
+# The arguments for building images.
+CARGO_PROFILE ?=
+FEATURES ?=
+TARGET_DIR ?=
+CARGO_BUILD_OPTS := --locked
+IMAGE_REGISTRY ?= docker.io
+IMAGE_NAMESPACE ?= greptime
 IMAGE_TAG ?= latest
+BUILDX_MULTI_PLATFORM_BUILD ?= false
+BUILDX_BUILDER_NAME ?= gtbuilder
+BASE_IMAGE ?= ubuntu
+RUST_TOOLCHAIN ?= $(shell cat rust-toolchain.toml | grep channel | cut -d'"' -f2)
+CARGO_REGISTRY_CACHE ?= ${HOME}/.cargo/registry
+
+# The arguments for running integration tests.
+ETCD_VERSION ?= v3.5.9
+ETCD_IMAGE ?= quay.io/coreos/etcd:${ETCD_VERSION}
+RETRY_COUNT ?= 3
+BUILD_JOBS ?= $(shell expr $$(nproc) / 2)
+ifeq ($(BUILD_JOBS), 0) # If the number of cores is less than 2, set the build jobs to 1.
+  BUILD_JOBS := 1
+endif
+
+ifdef CARGO_PROFILE
+	CARGO_BUILD_OPTS += --profile ${CARGO_PROFILE}
+endif
+
+ifdef FEATURES
+	CARGO_BUILD_OPTS += --features ${FEATURES}
+endif
+
+ifdef TARGET_DIR
+	CARGO_BUILD_OPTS += --target-dir ${TARGET_DIR}
+endif
+
+ifeq ($(BUILDX_MULTI_PLATFORM_BUILD), true)
+	BUILDX_MULTI_PLATFORM_BUILD_OPTS := --platform linux/amd64,linux/arm64 --push
+else
+	BUILDX_MULTI_PLATFORM_BUILD_OPTS := -o type=docker
+endif
 
 ##@ Build
 
 .PHONY: build
-build: ## Build debug version greptime.
+build: ## Build debug version greptime. If USE_DEV_BUILDER is true, the binary will be built in dev-builder.
-	cargo build
+ifeq ($(USE_DEV_BUILDER), true)
+	docker run --network=host \
+	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder:latest \
+	make build CARGO_PROFILE=${CARGO_PROFILE} FEATURES=${FEATURES} TARGET_DIR=${TARGET_DIR}
+else
+	cargo build ${CARGO_BUILD_OPTS}
+endif
 
 .PHONY: release
-release:  ## Build release version greptime.
+release:  ## Build release version greptime. If USE_DEV_BUILDER is true, the binary will be built in dev-builder.
-	cargo build --release
+ifeq ($(USE_DEV_BUILDER), true)
+	docker run --network=host \
+	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder:latest \
+	make release CARGO_PROFILE=${CARGO_PROFILE} FEATURES=${FEATURES} TARGET_DIR=${TARGET_DIR}
+else
+	cargo build --release ${CARGO_BUILD_OPTS}
+endif
 
 .PHONY: clean
 clean: ## Clean the project.
@@ -28,13 +80,38 @@ check-toml: ## Check all TOML files.
 	taplo format --check --option "indent_string=    "
 
 .PHONY: docker-image
-docker-image: ## Build docker image.
+docker-image: multi-platform-buildx ## Build docker image.
-	docker build --network host -f docker/Dockerfile -t ${IMAGE_REGISTRY}:${IMAGE_TAG} .
+	docker buildx build --builder ${BUILDX_BUILDER_NAME} \
+	  --build-arg="CARGO_PROFILE=${CARGO_PROFILE}" --build-arg="FEATURES=${FEATURES}" \
+	  -f docker/${BASE_IMAGE}/Dockerfile \
+	  -t ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/greptimedb:${IMAGE_TAG} ${BUILDX_MULTI_PLATFORM_BUILD_OPTS} .
+
+.PHONY: build-greptime-by-buildx
+build-greptime-by-buildx: multi-platform-buildx ## Build greptime binary by docker buildx. The binary will be copied to the current directory.
+	docker buildx build --builder ${BUILDX_BUILDER_NAME} \
+	  --target=builder \
+	  --build-arg="CARGO_PROFILE=${CARGO_PROFILE}" --build-arg="FEATURES=${FEATURES}" \
+	  -f docker/${BASE_IMAGE}/Dockerfile \
+	  -t ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/greptimedb-builder:${IMAGE_TAG} ${BUILDX_MULTI_PLATFORM_BUILD_OPTS} .
+
+	docker run --rm -v ${PWD}:/data \
+      --entrypoint cp ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/greptimedb-builder:${IMAGE_TAG} \
+      /out/target/${CARGO_PROFILE}/greptime /data/greptime
+
+.PHONY: dev-builder
+dev-builder: multi-platform-buildx ## Build dev-builder image.
+	docker buildx build --builder ${BUILDX_BUILDER_NAME} \
+	--build-arg="RUST_TOOLCHAIN=${RUST_TOOLCHAIN}" \
+	-f docker/dev-builder/Dockerfile \
+	-t ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder:${IMAGE_TAG} ${BUILDX_MULTI_PLATFORM_BUILD_OPTS} .
+
+.PHONY: multi-platform-buildx
+multi-platform-buildx: ## Create buildx multi-platform builder.
+	docker buildx inspect ${BUILDX_BUILDER_NAME} || docker buildx create --name ${BUILDX_BUILDER_NAME} --driver docker-container --bootstrap --use
 
 ##@ Test
-
 test: nextest ## Run unit and integration tests.
-	cargo nextest run --retries 3
+	cargo nextest run --retries ${RETRY_COUNT} --build-jobs=${BUILD_JOBS}
 
 .PHONY: nextest ## Install nextest tools.
 nextest:
@@ -56,6 +133,21 @@ clippy: ## Check clippy rules.
 fmt-check: ## Check code format.
 	cargo fmt --all -- --check
 
+.PHONY: start-etcd
+start-etcd: ## Start single node etcd for testing purpose.
+	docker run --rm -d --network=host -p 2379-2380:2379-2380 ${ETCD_IMAGE}
+
+.PHONY: stop-etcd
+stop-etcd: ## Stop single node etcd for testing purpose.
+	docker stop $$(docker ps -q --filter ancestor=${ETCD_IMAGE})
+
+.PHONY: run-it-in-container
+run-it-in-container: start-etcd ## Run integration tests in dev-builder.
+	docker run --network=host \
+	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry -v /tmp:/tmp \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder:latest \
+	make test sqlness-test BUILD_JOBS=${BUILD_JOBS}
+
 ##@ General
 
 # The help target prints out all targets with their descriptions organized
@@ -71,4 +163,4 @@ fmt-check: ## Check code format.
 
 .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($$0, 5) } ' $(MAKEFILE_LIST)
+	@awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n  make \033[36m<target>\033[0m\n"} /^[a-zA-Z_0-9-]+:.*?##/ { printf "  \033[36m%-30s\033[0m %s\n", $$1, $$2 } /^##@/ { printf "\n\033[1m%s\033[0m\n", substr($$0, 5) } ' $(MAKEFILE_LIST)

docker/aarch64/Dockerfile

@@ -1,60 +0,0 @@
-FROM ubuntu:22.04 as builder
-
-ENV LANG en_US.utf8
-WORKDIR /greptimedb
-
-# Install dependencies.
-RUN apt-get update && apt-get install -y \
-    libssl-dev \
-    protobuf-compiler \
-    curl \
-    git \
-    build-essential \
-    pkg-config \
-    wget
-
-# Install Rust.
-SHELL ["/bin/bash", "-c"]
-RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- --no-modify-path --default-toolchain none -y
-ENV PATH /root/.cargo/bin/:$PATH
-
-# Install cross platform toolchain
-RUN apt-get -y update && \
-    apt-get -y install g++-aarch64-linux-gnu gcc-aarch64-linux-gnu && \
-    apt-get install binutils-aarch64-linux-gnu
-
-COPY ./docker/aarch64/compile-python.sh ./docker/aarch64/
-RUN chmod +x ./docker/aarch64/compile-python.sh && \
-    ./docker/aarch64/compile-python.sh
-
-COPY ./rust-toolchain.toml .
-# Install rustup target for cross compiling.
-RUN rustup target add aarch64-unknown-linux-gnu
-COPY . .
-# Update dependency, using separate `RUN` to separate cache
-RUN cargo fetch
-
-# This three env var is set in script, so I set it manually in dockerfile.
-ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib/
-ENV LIBRARY_PATH=$LIBRARY_PATH:/usr/local/lib/
-ENV PY_INSTALL_PATH=/greptimedb/python_arm64_build
-
-# Set the environment variable for cross compiling and compile it
-# cross compiled python is `python3` in path, but pyo3 need `python` in path so alias it
-# Build the project in release mode.
-RUN export PYO3_CROSS_LIB_DIR=$PY_INSTALL_PATH/lib && \ 
-    alias python=python3 && \
-    cargo build --target aarch64-unknown-linux-gnu --release -F pyo3_backend
-
-# Exporting the binary to the clean image
-FROM ubuntu:22.04 as base
-
-RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get \
-    -y install ca-certificates \
-    curl
-
-WORKDIR /greptime
-COPY --from=builder /greptimedb/target/aarch64-unknown-linux-gnu/release/greptime /greptime/bin/
-ENV PATH /greptime/bin/:$PATH
-
-ENTRYPOINT ["greptime"]

docker/aarch64/compile-python.sh

@@ -1,87 +0,0 @@
-#!/usr/bin/env bash
-
-set -e
-
-# this script will download Python source code, compile it, and install it to /usr/local/lib
-# then use this python to compile cross-compiled python for aarch64
-ARCH=$1
-PYTHON_VERSION=3.10.10
-PYTHON_SOURCE_DIR=Python-${PYTHON_VERSION}
-PYTHON_INSTALL_PATH_AMD64=${PWD}/python-${PYTHON_VERSION}/amd64
-PYTHON_INSTALL_PATH_AARCH64=${PWD}/python-${PYTHON_VERSION}/aarch64
-
-function download_python_source_code() {
-  wget https://www.python.org/ftp/python/$PYTHON_VERSION/Python-$PYTHON_VERSION.tgz
-  tar -xvf Python-$PYTHON_VERSION.tgz
-}
-
-function compile_for_amd64_platform() {
-  mkdir -p "$PYTHON_INSTALL_PATH_AMD64"
-
-  echo "Compiling for amd64 platform..."
-
-  ./configure \
-    --prefix="$PYTHON_INSTALL_PATH_AMD64" \
-    --enable-shared \
-    ac_cv_pthread_is_default=no ac_cv_pthread=yes ac_cv_cxx_thread=yes \
-    ac_cv_have_long_long_format=yes \
-    --disable-ipv6 ac_cv_file__dev_ptmx=no ac_cv_file__dev_ptc=no
-
-  make
-  make install
-}
-
-# explain Python compile options here a bit:s
-# --enable-shared: enable building a shared Python library (default is no) but we do need it for calling from rust
-# CC, CXX, AR, LD, RANLIB: set the compiler, archiver, linker, and ranlib programs to use
-# build: the machine you are building on, host: the machine you will run the compiled program on
-# --with-system-ffi: build _ctypes module using an installed ffi library, see Doc/library/ctypes.rst, not used in here TODO: could remove
-# ac_cv_pthread_is_default=no ac_cv_pthread=yes ac_cv_cxx_thread=yes:
-# allow cross-compiled python to have -pthread set for CXX, see https://github.com/python/cpython/pull/22525
-# ac_cv_have_long_long_format=yes: target platform supports long long type
-# disable-ipv6: disable ipv6 support, we don't need it in here
-# ac_cv_file__dev_ptmx=no ac_cv_file__dev_ptc=no: disable pty support, we don't need it in here
-function compile_for_aarch64_platform() {
-  export LD_LIBRARY_PATH=$PYTHON_INSTALL_PATH_AMD64/lib:$LD_LIBRARY_PATH
-  export LIBRARY_PATH=$PYTHON_INSTALL_PATH_AMD64/lib:$LIBRARY_PATH
-  export PATH=$PYTHON_INSTALL_PATH_AMD64/bin:$PATH
-
-  mkdir -p "$PYTHON_INSTALL_PATH_AARCH64"
-
-  echo "Compiling for aarch64 platform..."
-  echo "LD_LIBRARY_PATH: $LD_LIBRARY_PATH"
-  echo "LIBRARY_PATH: $LIBRARY_PATH"
-  echo "PATH: $PATH"
-
-  ./configure --build=x86_64-linux-gnu --host=aarch64-linux-gnu \
-    --prefix="$PYTHON_INSTALL_PATH_AARCH64" --enable-optimizations \
-    CC=aarch64-linux-gnu-gcc \
-    CXX=aarch64-linux-gnu-g++ \
-    AR=aarch64-linux-gnu-ar \
-    LD=aarch64-linux-gnu-ld \
-    RANLIB=aarch64-linux-gnu-ranlib \
-    --enable-shared \
-    ac_cv_pthread_is_default=no ac_cv_pthread=yes ac_cv_cxx_thread=yes \
-    ac_cv_have_long_long_format=yes \
-    --disable-ipv6 ac_cv_file__dev_ptmx=no ac_cv_file__dev_ptc=no
-
-  make
-  make altinstall
-}
-
-# Main script starts here.
-download_python_source_code
-
-# Enter the python source code directory.
-cd $PYTHON_SOURCE_DIR || exit 1
-
-# Build local python first, then build cross-compiled python.
-compile_for_amd64_platform
-
-# Clean the build directory.
-make clean && make distclean
-
-# Cross compile python for aarch64.
-if [ "$ARCH" = "aarch64-unknown-linux-gnu" ]; then
-  compile_for_aarch64_platform
-fi

docker/centos/Dockerfile

@@ -1,4 +1,7 @@
-FROM centos:7
+FROM centos:7 as builder
+
+ARG CARGO_PROFILE
+ARG FEATURES
 
 ENV LANG en_US.utf8
 WORKDIR /greptimedb
@@ -21,4 +24,28 @@ SHELL ["/bin/bash", "-c"]
 RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- --no-modify-path --default-toolchain none -y
 ENV PATH /opt/rh/rh-python38/root/usr/bin:/usr/local/bin:/root/.cargo/bin/:$PATH
 
-CMD ["cargo", "build", "--release"]
+# Build the project in release mode.
+RUN --mount=target=.,rw \
+    --mount=type=cache,target=/usr/local/cargo/registry \
+    make build \
+    CARGO_PROFILE=${CARGO_PROFILE} \
+    FEATURES=${FEATURES} \
+    TARGET_DIR=/out/target
+
+# Export the binary to the clean image.
+FROM centos:7 as base
+
+ARG CARGO_PROFILE
+
+RUN yum install -y epel-release \
+    openssl \
+    openssl-devel  \
+    centos-release-scl  \
+    rh-python38  \
+    rh-python38-python-devel
+
+WORKDIR /greptime
+COPY --from=builder /out/target/${CARGO_PROFILE}/greptime /greptime/bin/
+ENV PATH /greptime/bin/:$PATH
+
+ENTRYPOINT ["greptime"]

docker/ci/Dockerfile

@@ -7,7 +7,7 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     python3-pip \
     curl
 
-COPY requirements.txt /etc/greptime/requirements.txt
+COPY ./docker/python/requirements.txt /etc/greptime/requirements.txt
 
 RUN python3 -m pip install -r /etc/greptime/requirements.txt
 

docker/ci/Dockerfile-centos

@@ -0,0 +1,16 @@
+FROM centos:7
+
+RUN yum install -y epel-release \
+    openssl \
+    openssl-devel  \
+    centos-release-scl  \
+    rh-python38  \
+    rh-python38-python-devel
+
+ARG TARGETARCH
+
+ADD $TARGETARCH/greptime /greptime/bin/
+
+ENV PATH /greptime/bin/:$PATH
+
+ENTRYPOINT ["greptime"]

docker/dev-builder/Dockerfile

@@ -0,0 +1,36 @@
+FROM ubuntu:22.04
+
+ENV LANG en_US.utf8
+WORKDIR /greptimedb
+
+# Install dependencies.
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
+    libssl-dev \
+    tzdata \
+    protobuf-compiler \
+    curl \
+    ca-certificates \
+    git \
+    build-essential \
+    pkg-config \
+    python3.10 \
+    python3.10-dev \
+    python3-pip
+
+RUN git config --global --add safe.directory /greptimedb
+
+# Install Python dependencies.
+COPY ./docker/python/requirements.txt /etc/greptime/requirements.txt
+RUN python3 -m pip install -r /etc/greptime/requirements.txt
+
+# Install Rust.
+SHELL ["/bin/bash", "-c"]
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- --no-modify-path --default-toolchain none -y
+ENV PATH /root/.cargo/bin/:$PATH
+
+# Install Rust toolchains.
+ARG RUST_TOOLCHAIN
+RUN rustup toolchain install ${RUST_TOOLCHAIN}
+
+# Install nextest.
+RUN cargo install cargo-nextest --locked

docker/ubuntu/Dockerfile

@@ -1,5 +1,8 @@
 FROM ubuntu:22.04 as builder
 
+ARG CARGO_PROFILE
+ARG FEATURES
+
 ENV LANG en_US.utf8
 WORKDIR /greptimedb
 
@@ -11,11 +14,9 @@ RUN apt-get update && apt-get install -y \
     git \
     build-essential \
     pkg-config \
-    python3 \
+    python3.10 \
-    python3-dev \
+    python3.10-dev \
-    python3-pip \
+    python3-pip
-    && pip3 install --upgrade pip \
-    && pip3 install pyarrow
 
 # Install Rust.
 SHELL ["/bin/bash", "-c"]
@@ -23,19 +24,32 @@ RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- --no-mo
 ENV PATH /root/.cargo/bin/:$PATH
 
 # Build the project in release mode.
-COPY . .
+RUN --mount=target=.,rw \
-RUN cargo build --release
+    --mount=type=cache,target=/usr/local/cargo/registry \
+    make build \
+    CARGO_PROFILE=${CARGO_PROFILE} \
+    FEATURES=${FEATURES} \
+    TARGET_DIR=/out/target
 
 # Export the binary to the clean image.
 # TODO(zyy17): Maybe should use the more secure container image.
 FROM ubuntu:22.04 as base
 
+ARG CARGO_PROFILE
+
 RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get \
     -y install ca-certificates \
+    python3.10 \
+    python3.10-dev \
+    python3-pip \
     curl
 
+COPY ./docker/python/requirements.txt /etc/greptime/requirements.txt
+
+RUN python3 -m pip install -r /etc/greptime/requirements.txt
+
 WORKDIR /greptime
-COPY --from=builder /greptimedb/target/release/greptime /greptime/bin/
+COPY --from=builder /out/target/${CARGO_PROFILE}/greptime /greptime/bin/
 ENV PATH /greptime/bin/:$PATH
 
 ENTRYPOINT ["greptime"]