Day 1: Caching Action

December 01, 2019

If you wish to cache dependencies or build outputs between runs, GitHub’s official Cache Action solves this.

Example Usage

This can cache absolutely anything you wish, regardless of platform or programming language.

For instance, if you are running Node tests in a CI/CD situation, if you cache the yarn cache, it will speed up the yarn up a bit. (It is reccomended that you don’t cache node modules as it could cause issues as you swap between node versions & architectures).

steps:
  - name: Get yarn cache
    id: yarn-cache
    run: echo "::set-output name=dir::$(yarn cache dir)"
  
  - name: Cache yarn cache
    uses: actions/cache@v1
    id: cache-yarn
    with:
      path: ${{ steps.yarn-cache.outputs.dir }}
      key: ${{ runner.os }}-yarn-${{ hashFiles('**/yarn.lock') }}
      restore-keys: |
        ${{ runner.os }}-yarn-
  
  - run: yarn
  - run: yarn test

It creates a hash based string using the contents of the lockfile so it will invalidate the cache if the lockfile changes. Then here, the yarn run will be extremely sped up as it doesn’t need to hit the internet to pull any dependencies down.

Limits

Each cache can have 400MB of cached content which is a gzipped tarball. The repo at large can have up to 2GB of caches. It will remove old caches either after the limit is hit or after a cache has been not accessed in a week.