build: Lock GitHub runners' OS (#1765)

Jesse Claven · web-flow · commit ab1f3f746ddf · 2024-05-30T13:50:32.000Z
* build: Lock GitHub runners' OS This was motivated by our macOS jobs failing [2] because colima is missing. It looks like this is because the latest versions of the macOS runner no longer have colima installed by default [1]. colima is now explicitly installed. [1] actions/runner-images#6216 [2] `/Users/runner/work/_temp/f19ffbff-27a9-4fc7-80b6-97791d2de141.sh: line 9: colima: command not found` * build: Lock Colima * build: Move macOS Docker installation to script * build: Move macOS libomp activation to script * build: Use latest Colima The > 0.6.0 releases actually fix the issue we have linked [1][2][3]. [1] abiosoft/colima#577 [2] https://github.com/jesse-c/MLServer/blob/c3acd60995a72141027eff506e4fd330fe824179/hack/install-docker-macos.sh#L18-L20 [3] > Switch to new user-v2 network. Fixes abiosoft/colima#648, abiosoft/colima#603, abiosoft/colima#577, abiosoft/colima#779, abiosoft/colima#137, abiosoft/colima#740.
diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
@@ -15,7 +15,7 @@ jobs:
       matrix:
         python-version: ["3.9", "3.10"]
         target: ["dataplane", "model-repository"]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-22.04
     steps:
       - uses: actions/checkout@v4
       - name: Set up Python ${{ matrix.python-version }}
@@ -36,8 +36,10 @@ jobs:
   lint:
     strategy:
       matrix:
-        python-version: ["3.9", "3.10"]
-    runs-on: ubuntu-latest
+        python-version:
+          - "3.9"
+          - "3.10"
+    runs-on: ubuntu-22.04
     steps:
       - uses: actions/checkout@v4
       - name: Set up Python ${{ matrix.python-version }}
@@ -57,33 +59,30 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        os: [ubuntu-latest, macos-latest]
+        os:
+          - ubuntu-22.04
+          - macos-13
         # NOTE: There's no pre-built `grpcio` wheel for Python 3.11 yet
         # https://github.com/grpc/grpc/issues/32454
-        python-version: ["3.9", "3.10"]
+        python-version:
+          - "3.9"
+          - "3.10"
         is-pr:
           - ${{ github.event_name == 'pull_request' }}
         exclude:
           # MacOS tests take a lot of time, so we will run them only on merge
           # From https://github.com/orgs/community/discussions/26253
           - is-pr: true
-            os: macos-latest
+            os: macos-13
     runs-on: ${{ matrix.os }}
     steps:
-      - name: Setup docker (missing on MacOS)
-        if: runner.os == 'macos'
-        run: |
-          # From https://github.com/actions/runner-images/issues/17#issuecomment-1537238473
-          # From https://github.com/abiosoft/colima/discussions/273#discussioncomment-4959736
-          # NOTE: Use the `slirp` driver rather than the default to avoid
-          # random connection errors.
-          # From https://github.com/abiosoft/colima/issues/577
-          brew install docker docker-buildx
-          mkdir -p $HOME/.docker/cli-plugins
-          ln -sfn $(which docker-buildx) $HOME/.docker/cli-plugins/docker-buildx
-          colima start --memory 5 --network-driver slirp
-          sudo ln -sf $HOME/.colima/default/docker.sock /var/run/docker.sock
       - uses: actions/checkout@v4
+      - name: Setup Docker (missing on MacOS)
+        if: runner.os == 'macos'
+        run: ./hack/install-docker-macos.sh
+      - name: Install OpenMP (in MacOS)
+        if: runner.os == 'macOS'
+        run: brew install libomp
       - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
@@ -97,16 +96,28 @@ jobs:
         with:
           virtualenvs-create: false
       - name: Install Dependencies
-        run: poetry install --sync --only dev
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          poetry install --sync --only dev
       - name: Test
-        run: tox -e mlserver
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          tox -e mlserver
 
   runtimes:
     strategy:
       fail-fast: false
       matrix:
-        os: [ubuntu-latest, macos-latest]
-        python-version: ["3.9", "3.10"]
+        os:
+          - ubuntu-22.04
+          - macos-13
+        python-version:
+          - "3.9"
+          - "3.10"
         tox-environment:
           - sklearn
           - xgboost
@@ -122,10 +133,13 @@ jobs:
           # MacOS tests take a lot of time, so we will run them only on merge
           # From https://github.com/orgs/community/discussions/26253
           - is-pr: true
-            os: macos-latest
+            os: macos-13
     runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v4
+      - name: Install OpenMP (in MacOS)
+        if: runner.os == 'macOS'
+        run: brew install libomp
       - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
@@ -139,18 +153,30 @@ jobs:
         with:
           virtualenvs-create: false
       - name: Install Dependencies
-        run: poetry install --sync --only dev
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          poetry install --sync --only dev
       - name: Test
-        run: tox -c ./runtimes/${{ matrix.tox-environment }}
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          tox -c ./runtimes/${{ matrix.tox-environment }}
 
   # Ensure that having all the runtimes installed together works
   all-runtimes:
     if: github.event_name == 'push'
     strategy:
       fail-fast: false
       matrix:
-        os: [ubuntu-latest, macos-latest]
-        python-version: ["3.9", "3.10"]
+        os:
+          - ubuntu-22.04
+          - macos-13
+        python-version:
+          - "3.9"
+          - "3.10"
     runs-on: ${{ matrix.os }}
     steps:
       - name: Maximize build space
@@ -164,30 +190,13 @@ jobs:
           remove-docker-images: 'true'
           overprovision-lvm: 'true'
           swap-size-mb: 1024
-      - name: Setup docker (missing on MacOS)
+      - uses: actions/checkout@v4
+      - name: Setup Docker (missing on MacOS)
         if: runner.os == 'macOS'
-        run: |
-          # From https://github.com/actions/runner-images/issues/17#issuecomment-1537238473
-          # From https://github.com/abiosoft/colima/discussions/273#discussioncomment-4959736
-          # NOTE: Use the `slirp` driver rather than the default to avoid
-          # random connection errors.
-          # From https://github.com/abiosoft/colima/issues/577
-          brew install docker docker-buildx
-          mkdir -p $HOME/.docker/cli-plugins
-          ln -sfn $(which docker-buildx) $HOME/.docker/cli-plugins/docker-buildx
-          colima start --memory 5 --network-driver slirp
-          sudo ln -sf $HOME/.colima/default/docker.sock /var/run/docker.sock
-      - name: Downgrade OpenMP (in MacOS)
+        run: ./hack/install-docker-macos.sh
+      - name: Install OpenMP (in MacOS)
         if: runner.os == 'macOS'
-        run: |
-          # Recent versions of OpenMP cause segfaults in MacOS when training
-          # LightGBM / XGBoost models (but only when Torch is present)
-          # https://github.com/microsoft/LightGBM/issues/4229
-          # https://github.com/autogluon/autogluon/issues/1442
-          wget https://raw.githubusercontent.com/Homebrew/homebrew-core/fb8323f2b170bd4ae97e1bac9bf3e2983af3fdb0/Formula/libomp.rb
-          brew unlink libomp
-          brew install libomp.rb
-      - uses: actions/checkout@v4
+        run: brew install libomp
       - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
@@ -201,6 +210,14 @@ jobs:
         with:
           virtualenvs-create: false
       - name: Install Dependencies
-        run: poetry install --sync --only dev
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          poetry install --sync --only dev
       - name: Test
-        run: tox -e all-runtimes
+        run: |
+          if [[ ${{ runner.os == 'macOS' }} ]]; then
+            source ./hack/activate-libomp-macos.sh
+          fi
+          tox -e all-runtimes
diff --git a/hack/activate-libomp-macos.sh b/hack/activate-libomp-macos.sh
@@ -0,0 +1,6 @@
+#!/usr/bin/env bash
+
+set -Eeuo pipefail
+
+export LDFLAGS="-L/usr/local/opt/libomp/lib"
+export CPPFLAGS="-I/usr/local/opt/libomp/include"
diff --git a/hack/install-docker-macos.sh b/hack/install-docker-macos.sh
@@ -0,0 +1,24 @@
+#!/usr/bin/env bash
+
+set -Eeuo pipefail
+
+# Install Docker
+# From https://github.com/actions/runner-images/issues/17#issuecomment-1537238473
+# From https://github.com/abiosoft/colima/discussions/273#discussioncomment-4959736
+brew install docker docker-buildx
+mkdir -p $HOME/.docker/cli-plugins
+ln -sfn $(which docker-buildx) $HOME/.docker/cli-plugins/docker-buildx
+
+# Install Lima
+brew install lima
+# Don't explicitly start Lima here (e.g. with `limactl start`).
+# Let Colina do that, otherwise, # Colima seems to hang
+# waiting for the SSH requirement to be met.
+
+# Install Colima
+sudo mkdir -p /usr/local/bin
+sudo curl -L -o /usr/local/bin/colima https://github.com/abiosoft/colima/releases/download/v0.6.9/colima-Darwin-x86_64 && sudo chmod +x /usr/local/bin/colima
+colima start --memory 5 --runtime docker
+
+# Link Colima and Docker
+sudo ln -sf $HOME/.colima/default/docker.sock /var/run/docker.sock
