Skip to content

Commit

Permalink
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Loading
Loading status checks…
move bias fuzzer to scheduled.yml
Browse files Browse the repository at this point in the history
@assignUser
assignUser committed Mar 22, 2024

Verified

This commit was signed with the committer’s verified signature.
henrikvtcodes Henrik VT
SSH Key Fingerprint: rgiYvcVkgkMB8uoeJLyXfHj/7PsmAR6X1YuIioy8HTM
Verified
Learn about vigilant mode
1 parent 483b742 commit 1ea4657
Showing 2 changed files with 174 additions and 62 deletions.
58 changes: 2 additions & 56 deletions .github/workflows/linux-build.yml
View file
Original file line number Diff line number Diff line change
@@ -113,12 +113,6 @@ jobs:
shell: bash
working-directory: velox
steps:
- name: Get Function Signature Stash
uses: assignUser/stash/restore@v1
id: get-sig
with:
path: /tmp/signatures
key: function-signatures

- name: Get Ccache Stash
uses: assignUser/stash/restore@v1
@@ -130,55 +124,8 @@ jobs:
working-directory: ${{ github.workspace }}
run: |
mkdir -p '${{ env.CCACHE_DIR }}'
mkdir -p /tmp/signatures
- uses: actions/setup-python@v5
with:
python-version: '3.10'

- name: Checkout Main
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
uses: actions/checkout@v4
with:
# hardcode ref without broken pr
ref: '43a552715e4c0c005cb73cfb6dca0f996de76f3a'
path: velox_main
- name: Install Dependencies
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
working-directory: velox_main
run: |
source scripts/setup-ubuntu.sh
- name: Build PyVelox
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
env:
VELOX_DEPENDENCY_SOURCE: BUNDLED
MAKEFLAGS: "NUM_THREADS=8 MAX_HIGH_MEM_JOBS=4 MAX_LINK_JOBS=4"
working-directory: velox_main
run: |
python -m venv .venv
source .venv/bin/activate
make python-build
- name: Create Baseline Signatures
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
working-directory: velox_main
run: |
source .venv/bin/activate
python -m pip install deepdiff
python scripts/signature.py export --spark /tmp/signatures/spark_signatures_main.json
python scripts/signature.py export --presto /tmp/signatures/presto_signatures_main.json
- name: Save Function Signature Stash
uses: assignUser/stash/save@v1
with:
path: /tmp/signatures
key: function-signatures

- name: Checkout Contender
uses: actions/checkout@v4
- uses: actions/checkout@v4
with:
path: velox

@@ -195,8 +142,7 @@ jobs:
VELOX_DEPENDENCY_SOURCE: BUNDLED
MAKEFLAGS: "NUM_THREADS=8 MAX_HIGH_MEM_JOBS=4 MAX_LINK_JOBS=4"
run: |
make debug EXTRA_CMAKE_FLAGS="-DVELOX_ENABLE_ARROW=ON \
-DVELOX_BUILD_PYTHON_PACKAGE=ON"
make debug EXTRA_CMAKE_FLAGS="-DVELOX_ENABLE_ARROW=ON"
- name: CCache after
run: |
178 changes: 172 additions & 6 deletions .github/workflows/scheduled.yml
View file
Original file line number Diff line number Diff line change
@@ -70,21 +70,88 @@ env:
jobs:
compile:
name: Build
runs-on: 8-core
runs-on: 16-core
timeout-minutes: 120
env:
CCACHE_DIR: "${{ github.workspace }}/.ccache/"
CCACHE_BASEDIR: "${{ github.workspace }}"
LINUX_DISTRO: "ubuntu"
defaults:
run:
shell: bash
working-directory: velox
outputs:
bias_presto: ${{ steps.sig-check.outputs.presto }}
bias_spark: ${{ steps.sig-check.outputs.spark }}

steps:

- name: Get Function Signature Stash
uses: assignUser/stash/restore@v1
id: get-sig
with:
path: /tmp/signatures
key: function-signatures

- name: "Restore ccache"
uses: assignUser/stash/restore@v1
with:
path: "${{ env.CCACHE_DIR }}"
key: ccache-fuzzer

- name: "Checkout Repo"
- name: Ensure Stash Dirs Exists
working-directory: ${{ github.workspace }}
run: |
mkdir -p '${{ env.CCACHE_DIR }}'
mkdir -p /tmp/signatures
- uses: actions/setup-python@v5
with:
python-version: '3.10'

- name: Checkout Main
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
uses: actions/checkout@v4
with:
# hardcode ref without broken pr
ref: '43a552715e4c0c005cb73cfb6dca0f996de76f3a'
path: velox_main

- name: Install Dependencies
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
working-directory: velox_main
run: |
source scripts/setup-ubuntu.sh
- name: Build PyVelox
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
env:
VELOX_DEPENDENCY_SOURCE: BUNDLED
MAKEFLAGS: "NUM_THREADS=16 MAX_HIGH_MEM_JOBS=8 MAX_LINK_JOBS=8"
working-directory: velox_main
run: |
python -m venv .venv
source .venv/bin/activate
make python-build
- name: Create Baseline Signatures
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
working-directory: velox_main
run: |
source .venv/bin/activate
python -m pip install deepdiff
python scripts/signature.py export --spark /tmp/signatures/spark_signatures_main.json
python scripts/signature.py export --presto /tmp/signatures/presto_signatures_main.json
- name: Save Function Signature Stash
if: ${{ steps.get-sig.outputs.stash-hit != 'true' }}
uses: assignUser/stash/save@v1
with:
path: /tmp/signatures
key: function-signatures

- name: "Checkout Contender"
uses: actions/checkout@v4
with:
path: velox
@@ -93,14 +160,17 @@ jobs:

- name: "Install dependencies"
run: |
cd velox
source ./scripts/setup-ubuntu.sh
ccache -vsz
- name: Build
run: |
cd velox
make debug NUM_THREADS="${{ inputs.numThreads || 8 }}" MAX_HIGH_MEM_JOBS="${{ inputs.maxHighMemJobs || 8 }}" MAX_LINK_JOBS="${{ inputs.maxLinkJobs || 4 }}" EXTRA_CMAKE_FLAGS="-DVELOX_ENABLE_ARROW=ON ${{ inputs.extraCMakeFlags }}"
make debug NUM_THREADS="${{ inputs.numThreads || 16 }}" \
MAX_HIGH_MEM_JOBS="${{ inputs.maxHighMemJobs || 8 }}" \
MAX_LINK_JOBS="${{ inputs.maxLinkJobs || 8 }}" \
EXTRA_CMAKE_FLAGS="-DVELOX_ENABLE_ARROW=ON \
-DVELOX_BUILD_PYTHON_PACKAGE=ON \
${{ inputs.extraCMakeFlags }}"
- name: Ccache after
run: ccache -vs
@@ -111,6 +181,52 @@ jobs:
path: "${{ env.CCACHE_DIR }}"
key: ccache-fuzzer
retention-days: "${{ env.RETENTION }}"

- name: Build PyVelox
env:
VELOX_BUILD_DIR: "_build/debug"
run: |
python -m venv .venv
source .venv/bin/activate
python -m pip install -e .
- name: Create and test new function signatures
id: sig-check
run: |
source .venv/bin/activate
python -m pip install deepdiff
python scripts/signature.py export --spark /tmp/signatures/spark_signatures_contender.json
python scripts/signature.py export --presto /tmp/signatures/presto_signatures_contender.json
python scripts/signature.py bias /tmp/signatures/presto_signatures_main.json \
/tmp/signatures/presto_signatures_contender.json \
/tmp/signatures/presto_bias_functions 2>&1 > /tmp/signatures/presto-err-message \
|| echo "::notice ::Presto Signature check failed" \
| tee /tmp/signatures/presto-err-code
if [ -f "/tmp/signatures/presto_bias_functions" ]; then
echo "presto=true" >> $GITHUB_OUTPUT
else
echo "presto=false" >> $GITHUB_OUTPUT
fi
python scripts/signature.py bias /tmp/signatures/spark_signatures_main.json \
/tmp/signatures/spark_signatures_contender.json \
/tmp/signatures/spark_bias_functions \
|| echo "::notice ::Spark Signature check failed" \
| tee /tmp/signatures/spark-err-code
if [ -f "/tmp/signatures/spark_bias_functions" ]; then
echo "spark=true" >> $GITHUB_OUTPUT
else
echo "spark=false" >> $GITHUB_OUTPUT
fi
- name: Upload Signature Artifacts
uses: actions/upload-artifact@v4
with:
name: signatures
path: /tmp/signatures
retention-days: "${{ env.RETENTION }}"

- name: Upload presto fuzzer
uses: actions/upload-artifact@v4
@@ -266,7 +382,57 @@ jobs:
path: |
/tmp/spark_aggregate_fuzzer_repro
linux-spark-fuzzer-run:
spark-bias-fuzzer:
name: "Spark Fuzzer"
runs-on: ubuntu-latest
needs: compile
if: ${{ needs.compile.outputs.spark == 'true' }}
timeout-minutes: 120
steps:

- name: "Checkout Repo"
uses: actions/checkout@v4
with:
ref: "${{ inputs.ref }}"

- name: "Install dependencies"
run: source ./scripts/setup-ubuntu.sh

- name: Download spark expression fuzzer
uses: actions/download-artifact@v4
with:
name: spark_expression_fuzzer

- name: Download Signatures
uses: actions/download-artifact@v4
with:
name: signatures
path: /tmp/signatures

- name: "Run Spark Expression Fuzzer"
run: |
ls /tmp/signatures
mkdir -p /tmp/spark_fuzzer_repro/
chmod -R 777 /tmp/spark_fuzzer_repro
chmod +x spark_expression_fuzzer_test
./spark_expression_fuzzer_test \
--seed ${RANDOM} \
--duration_sec $DURATION \
--logtostderr=1 \
--minloglevel=0 \
--assign_function_tickets $(cat /tmp/signatures/spark_bias_functions) \
--repro_persist_path=/tmp/spark_fuzzer_repro \
&& echo -e "\n\nSpark Fuzzer run finished successfully."
- name: Archive Spark expression production artifacts
if: ${{ !cancelled() }}
uses: actions/upload-artifact@v4
with:
name: spark-fuzzer-failure-artifacts
path: |
/tmp/spark_bias_fuzzer_repro
spark-fuzzer:
name: "Spark Fuzzer"
runs-on: ubuntu-latest
needs: compile

0 comments on commit 1ea4657

Please sign in to comment.