Scrape and Index Workflow Runs #1385
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
name: Scrape and Index Workflow Runs | |
on: | |
schedule: | |
- cron: "30 * * * *" | |
workflow_dispatch: | |
inputs: | |
name: | |
required: true | |
type: string | |
repository: | |
required: true | |
type: string | |
default: "cilium/cilium" | |
branch: | |
required: true | |
type: string | |
default: main | |
until: | |
required: true | |
type: string | |
since: | |
required: true | |
type: string | |
events: | |
required: false | |
type: string | |
default: push | |
env: | |
TARGET_INDEX: runs-oss | |
defaults: | |
run: | |
shell: bash | |
jobs: | |
setup-matrix: | |
runs-on: ubuntu-22.04 | |
outputs: | |
matrix: ${{ steps.set-matrix.outputs.matrix }} | |
steps: | |
- if: ${{ github.event_name == 'workflow_dispatch' }} | |
name: Prepare workflow_dispatch inputs | |
env: | |
INPUTS_JSON: ${{ toJSON(inputs) }} | |
run: | | |
echo $INPUTS_JSON | jq -c '{ "include": [.] }' > /tmp/matrix.json | |
- if: ${{ github.event_name == 'schedule' }} | |
name: Clone scheduled inputs | |
uses: actions/checkout@v4 | |
with: | |
sparse-checkout: | | |
.github/scheduled.json | |
- if: ${{ github.event_name == 'schedule' }} | |
name: Prepare scheduled inputs | |
run: | | |
until=$(date '+%Y-%m-%dT%H') | |
since=$(date -d '-3 hours' '+%Y-%m-%dT%H') | |
m=$(jq -c '.include |= map(. + {"until": "'$until'", "since": "'$since'"})' < ./.github/scheduled.json) | |
echo $m > /tmp/matrix.json | |
- if: ${{ github.event_name != 'schedule' && github.event_name != 'workflow_dispatch' }} | |
run: | | |
exit 1 | |
- id: set-matrix | |
run: | | |
cat /tmp/matrix.json | |
echo "matrix=$(cat /tmp/matrix.json)" >> $GITHUB_OUTPUT | |
scrape-and-index: | |
needs: setup-matrix | |
strategy: | |
# GitHub API doesn't like concurrent requests. | |
# Additionally, since our token is limited in the | |
# number of requests we can make, just do one | |
# repository at a time. If it takes longer that's ok | |
# since it reduces the chance we run into a limit. | |
max-parallel: 1 | |
# Scrape jobs are independent, therefore one job's failure | |
# shouldn't impact another's. | |
fail-fast: false | |
matrix: ${{ fromJSON(needs.setup-matrix.outputs.matrix) }} | |
runs-on: ubuntu-22.04 | |
steps: | |
- uses: actions/checkout@v4 | |
- name: Echo Matrix | |
run: | | |
echo '${{ toJSON(matrix) }}' | |
- name: Scrape Workflow Runs | |
env: | |
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} | |
run: | | |
go run . workflow runs \ | |
--repository ${{ matrix.repository }} \ | |
--branch ${{ matrix.branch }} \ | |
--until ${{ matrix.until }} \ | |
--since ${{ matrix.since }} \ | |
--events ${{ matrix.events }} \ | |
--index ${{ env.TARGET_INDEX }} \ | |
--verbose > results.json | |
- name: Bail if no results were found | |
id: bail | |
run: | | |
ls -lah results.json | |
if ! [ -s results.json ]; then | |
echo "No results, exiting early" | |
echo "bail=bail" >> $GITHUB_OUTPUT | |
fi | |
- name: Create splits | |
if: ${{ steps.bail.outputs.bail != 'bail' }} | |
run: | | |
mkdir splits | |
cd splits | |
split -l 200 ../results.json | |
ls -lah | |
cd .. | |
- name: Perform bulk requests | |
if: ${{ steps.bail.outputs.bail != 'bail' }} | |
working-directory: ./splits | |
run: | | |
set -x | |
for fn in ./*; do | |
curl -XPOST \ | |
--data-binary @$fn \ | |
--output $fn-response.json \ | |
-H "Content-Type: application/json" \ | |
--verbose \ | |
-u ${{ secrets.OPENSEARCH_USER }}:${{ secrets.OPENSEARCH_PASS }} \ | |
${{ vars.OPENSEARCH_URL }}/_bulk | |
done | |
- name: Upload results | |
if: ${{ steps.bail.outputs.bail != 'bail' }} | |
uses: actions/upload-artifact@v4 | |
with: | |
name: ${{ matrix.name }} | |
path: "splits/**" | |
- name: Fail if any items failed to index | |
if: ${{ steps.bail.outputs.bail != 'bail' }} | |
working-directory: ./splits | |
run: | | |
cat *.json | jq --exit-status '.errors | not' | |
cat *.json | jq --exit-status '.error | not' | |