Build Stats: Gather action stats & push it to BigQuery

This GitHub Action automates the process of gathering statistics from a specific run and pushing the collected data to a Google BigQuery table.

Usage

1. Create a service account key in Google Cloud Platform (GCP) with the necessary permissions to write to your BigQuery dataset.

2. Add the service account key as a secret in your GitHub repository settings. Name the secret GOOGLE_APPLICATION_CREDENTIALS.

3. Provide a BigQuery dataset, BigQuery table, and the project name in GCP, as inputs: BQ_DATASET, BQ_TABLE, GOOGLE_PROJECT_NAME.

4. Provide a GITHUB_TOKEN, as an input with repo access.

5. Preferably, this action should be the last one, scheduled after all other jobs. You can utilize the needs keyword to achieve this.

Getting Started

Example workflow:

build-stats:

  name: Build Stats
  runs-on: ubuntu-latest
  needs: [last_job]

  steps:
  - name: Get Github Build Stats for current run
    uses: SallyBlichG/github-build-stats@main
    with: 
      GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      GITHUB_ORG_ID: owner\organization id
      GOOGLE_PROJECT_NAME: google_project_name
      BQ_DATASET: bq_data_set
      BQ_TABLE: github_action_stats
      GOOGLE_APPLICATION_CREDENTIALS: ${{ secrets.GOOGLE_APPLICATION_CREDENTIALS }}

Inputs

All inputs are required!

GITHUB_TOKEN: GitHub token to access the build stats. Repo access should be good enough

GITHUB_ORG_ID: The organization \ owner id where your repository is located. It might be your username, company name, etc

GITHUB_REPOSITORY_NAME: The repository name that we with to get stats from its workflow. Default value is the current repository name, ${{ github.event.repository.name }}

GITHUB_RUN_ID: The run id that we wish to get stats from. Default value is the current run, ${{ github.run_id }}

BQ_DATASET: Big query dataset

BQ_TABLE: Big query table

GOOGLE_PROJECT_NAME: Google project name (GCP)

GOOGLE_APPLICATION_CREDENTIALS: Google credentials (GCP), should be the service account

Contributions

This project is using Valgrind for caching APT packages in GitHub Actions workflow. Valgrind is an instrumentation framework for building dynamic analysis tools. It can detect memory leaks and threading bugs, which are common in C and C++ programs.