Only cache node_modules on main (#45544)

Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/45544

## This diff now does 5 things:
1. removes the old way we used `actions/setup-node` to manage the cache itself.
2. it creates a new `update-node-modules-cache` workflow, which is the only job that will update the node modules cache
3. it create a `yarn-install-with-cache` action that should be used install of directly calling `yarn install --non-interactive`.  This will load a cache against a hash of `package.json`.
4. updated the cache reaper to aggressively remove everything but the latest `npm-{{ hash('package.json') }}`.
5. removed a `cache-setup`, which couldn't be used (we're using artefacts now).

## Why are we doing this:
The various `node-cache-` keys for platforms and on various branches accounts for a very large proportion of the cache (10-20%).

We don't frequently change these dependencies, and even when we do running `yarn install` after loading the cache will resolve any issues.

Limiting the cache to `main` and aggressively pruning older cache entries will clean up a lot of "small win" caching.

Changelog: [Internal]

Reviewed By: cortinico

Differential Revision: D59917944

fbshipit-source-id: 4be6f1959e8fde642a4f208f7d19aceba2c3262f

.github/actions/build-android/action.yml

@@ -12,9 +12,8 @@ runs:
       run: git config --global --add safe.directory '*'
     - name: Setup node.js
       uses: ./.github/actions/setup-node
-    - name: Install dependencies
-      shell: bash
-      run: yarn install --non-interactive
+    - name: Install node dependencies
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Set React Native Version
       shell: bash
       run: node ./scripts/releases/set-rn-artifacts-version.js --build-type ${{ inputs.release-type }}

.github/actions/build-hermes-macos/action.yml

@@ -47,8 +47,7 @@ runs:
         fi
     - name: Yarn- Install Dependencies
       if: ${{ steps.check_if_apple_artifacts_are_there.outputs.ARTIFACTS_EXIST != 'true' }}
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Slice cache macosx
       if: ${{ steps.check_if_apple_artifacts_are_there.outputs.ARTIFACTS_EXIST != 'true' }}
       uses: actions/download-artifact@v4

.github/actions/build-npm-package/action.yml

@@ -103,8 +103,7 @@ runs:
     - name: Setup gradle
       uses: ./.github/actions/setup-gradle
     - name: Install dependencies
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Build packages
       shell: bash
       run: yarn build

.github/actions/cache-setup/action.yml

@@ -1,88 +0,0 @@
-name: cache-setup
-description: "Cache setup"
-inputs:
-  hermes-version:
-    description: "Hermes version"
-    required: true
-  react-native-version:
-    description: "React Native version"
-    required: true
-outputs:
-  cache-hit-hermes-tarball-release:
-    description: "Whether the hermes tarball release cache was hit"
-    value: ${{ steps.cache_hermes_tarball_release.outputs.cache-hit }}
-  cache-hit-hermes-tarball-debug:
-    description: "Whether the hermes tarball debug cache was hit"
-    value: ${{ steps.cache_hermes_tarball_debug.outputs.cache-hit }}
-  cache-hit-macos-bin-release:
-    description: "Whether the macos bin release cache was hit"
-    value: ${{ steps.cache_macos_bin_release.outputs.cache-hit }}
-  cache-hit-macos-bin-debug:
-    description: "Whether the macos bin debug cache was hit"
-    value: ${{ steps.cache_macos_bin_debug.outputs.cache-hit }}
-  cache-hit-dsym-release:
-    description: "Whether the dsym release cache was hit"
-    value: ${{ steps.cache_dsym_release.outputs.cache-hit }}
-  cache-hit-dsym-debug:
-    description: "Whether the dsym debug cache was hit"
-    value: ${{ steps.cache_dsym_debug.outputs.cache-hit }}
-runs:
-  using: composite
-  steps:
-    - name: Cache hermes tarball release
-      id: cache_hermes_tarball_release
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/hermes-runtime-darwin/hermes-ios-Release.tar.gz
-        key: v4-hermes-tarball-release-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}-${{ hashfiles('packages/react-native/sdks/hermes-engine/utils/build-apple-framework.sh') }}
-        enableCrossOsArchive: true
-    - name: Cache hermes tarball debug
-      id: cache_hermes_tarball_debug
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/hermes-runtime-darwin/hermes-ios-Debug.tar.gz
-        key: v4-hermes-tarball-debug-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}-${{ hashfiles('packages/react-native/sdks/hermes-engine/utils/build-apple-framework.sh') }}
-        enableCrossOsArchive: true
-    - name: Cache macos bin release
-      id: cache_macos_bin_release
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/osx-bin/Release
-        key: v2-hermes-release-macosx-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}
-        enableCrossOsArchive: true
-    - name: Cache macos bin debug
-      id: cache_macos_bin_debug
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/osx-bin/Debug
-        key: v2-hermes-debug-macosx-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}
-        enableCrossOsArchive: true
-    - name: Cache dsym release
-      id: cache_dsym_release
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/dSYM/Release
-        key: v2-hermes-release-dsym-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}
-        enableCrossOsArchive: true
-    - name: Cache dsym debug
-      id: cache_dsym_debug
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/dSYM/Debug
-        key: v2-hermes-debug-dsym-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}
-        enableCrossOsArchive: true
-    - name: HermesC Apple
-      id: hermesc_apple
-      uses: actions/cache@v4
-      with:
-        path: /tmp/hermes/hermesc-apple
-        key: v2-hermesc-apple-${{ inputs.hermes-version }}-${{ inputs.react-native-version }}
-        enableCrossOsArchive: true
-    - name: Cache hermes workspace
-      uses: actions/cache@v4
-      with:
-        path: |
-          /tmp/hermes/download/
-          /tmp/hermes/hermes/
-        key: v1-hermes-${{ inputs.hermes-version }}
-        enableCrossOsArchive: true

.github/actions/create-release/action.yml

@@ -15,8 +15,7 @@ runs:
   using: composite
   steps:
     - name: Yarn install
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Configure Git
       shell: bash
       run: |

.github/actions/prepare-hermes-workspace/action.yml

@@ -69,8 +69,7 @@ runs:
 
     - name: Yarn- Install Dependencies
       if: ${{ steps.meaningful-cache.outputs.HERMES_CACHED != 'true' }}
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
 
     - name: Download Hermes tarball
       if: ${{ steps.meaningful-cache.outputs.HERMES_CACHED != 'true' }}

.github/actions/setup-node/action.yml

@@ -5,10 +5,6 @@ inputs:
     description: 'The node.js version to use'
     required: false
     default: '18'
-  cache:
-    description: 'The package manager to use for caching dependencies'
-    required: false
-    default: 'yarn'
 runs:
   using: "composite"
   steps:
@@ -16,4 +12,3 @@ runs:
       uses: actions/setup-node@v4
       with:
         node-version: ${{ inputs.node-version }}
-        cache: ${{ inputs.cache }}

.github/actions/test-ios-helloworld/action.yml

@@ -41,8 +41,7 @@ runs:
       shell: bash
       run: ls -lR "$HERMES_WS_DIR"
     - name: Run yarn
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Setup ruby
       uses: ruby/setup-ruby@v1.170.0
       with:

.github/actions/test-ios-rntester/action.yml

@@ -37,8 +37,7 @@ runs:
     - name: Setup node.js
       uses: ./.github/actions/setup-node
     - name: Run yarn
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Download Hermes
       uses: actions/download-artifact@v4
       with:

.github/actions/test-js/action.yml

@@ -13,8 +13,7 @@ runs:
       with:
         node-version: ${{ inputs.node-version }}
     - name: Yarn install
-      shell: bash
-      run: yarn install --non-interactive
+      uses: ./.github/actions/yarn-install-with-cache
     - name: Run Tests - JavaScript Tests
       shell: bash
       run: node ./scripts/run-ci-javascript-tests.js --maxWorkers 2

.github/actions/yarn-install-with-cache/action.yml

@@ -0,0 +1,28 @@
+name: yarn-install-with-cache
+inputs:
+  update-cache:
+    description: Update the cache, only do this if you are update-node-modules-cache.yml
+    default: "false"
+description: Only update node_modules if on main
+runs:
+  using: composite
+  steps:
+    - name: Load node_modules from cache
+      # Restore for all branches, but save for 'main'.
+      uses: actions/cache/restore@v4
+      with:
+        path: node_modules/
+        key: node-modules-${{ hashFiles('package.json') }}
+    - name: Install dependencies
+      shell: bash
+      run: yarn install --non-interactive
+    - name: Save node_modules to the cache
+      if: github.ref == 'refs/heads/main' && inputs.update-cache == 'true'
+      uses: actions/cache/save@v4
+      with:
+        path: node_modules/
+        # We're assuming that variations on branches will slightly vary from main,
+        # so it's always important to run yarn install --non-interactive after this
+        # cache is restored.
+        key: node-modules-v1-${{ hashFiles('package.json') }}
+        enableCrossOsArchive: true

.github/workflows/publish-bumped-packages.yml

@@ -17,7 +17,7 @@ jobs:
       - name: Setup node.js
         uses: ./.github/actions/setup-node
       - name: Run Yarn Install
-        run: yarn install
+        uses: ./.github/actions/yarn-install-with-cache
       - name: Build packages
         run: yarn build
       - name: Set NPM auth token

.github/workflows/test-all.yml

@@ -291,9 +291,8 @@ jobs:
           path: /tmp/maven-local
       - name: Setup gradle
         uses: ./.github/actions/setup-gradle
-      - name: Run yarn
-        shell: bash
-        run: yarn install --non-interactive
+      - name: Run yarn install
+        uses: ./.github/actions/yarn-install-with-cache
       - name: Prepare the Helloworld application
         shell: bash
         run: node ./scripts/e2e/init-project-e2e.js --useHelloWorld --pathToLocalReactNative "$GITHUB_WORKSPACE/build/$(cat build/react-native-package-version)"

.github/workflows/update-node-modules-cache.yml

@@ -0,0 +1,16 @@
+name: Update node modules cache
+
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - main
+
+jobs:
+  update_node_modules_cache:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Install yarn dependencies and update cache
+        uses: ./.github/actions/yarn-install-with-cache
+        with:
+          update-cache: "true"

scripts/clean-gha-cache.js

@@ -8,6 +8,8 @@
 const {execSync} = require('node:child_process');
 
 const CACHE_LIMIT = (10 * 1024 ** 3) * 0.9;
+// Doing this to capture node-modules- and the pre-existing node-cache-<platform>-yarn-<sha> entries
+const NODE_CACHE_KEY = 'node-';
 
 function cleanData(rawStr) {
   const now = new Date();
@@ -15,8 +17,10 @@ function cleanData(rawStr) {
   return json
     .map(raw => ({
       ...raw,
+      createdAt: now - new Date(raw.createdAt),
       msSinceLastAccessed: now - new Date(raw.lastAccessedAt),
     }))
+    // Order: oldest last access time first
     .sort((a, b) => b.msSinceLastAccessed - a.msSinceLastAccessed);
 }
 
@@ -25,26 +29,51 @@ const hrs = ms => (ms / (1000 * 60 * 60)).toFixed(1) + 'hrs';
 
 const dryRun = process.argv.indexOf('--dry') !== -1;
 
+function cacheToString(entry) {
+  return `[${hrs(entry.msSinceLastAccessed).padStart(7)}] ${mb(entry.sizeInBytes).padStart(7)} -> ${entry.key}`;
+}
+
 function main() {
   const cacheUsage = cleanData(execSync(
-    'gh cache list --sort last_accessed_at --json id,key,lastAccessedAt,sizeInBytes --limit 1000',
+    'gh cache list --sort last_accessed_at --json id,key,createdAt,lastAccessedAt,sizeInBytes --limit 1000',
     'utf8'
   ));
 
   let total = 0;
   let remove = [];
   let cleaned = 0;
+
+  // Be aggressive with node cache entries. Ignore entries < 1MB and only keep most
+  // recently created entry.
+  const nodeCacheUsage = cacheUsage
+    .filter(({key, sizeInBytes}) =>
+      // I've observed some noisy entries, ignore anything that isn't > 1MB
+      key.startsWith(NODE_CACHE_KEY) && sizeInBytes > 1024 * 1024
+    )
+    .sort((a, b) => b.createdAt - a.createdAt);
+  // Leave the latest entry only
+  const keeping = nodeCacheUsage.pop();
+
+  console.log('TASK: clean up old node_modules cache entries.', keeping ? `\nkeeping ${cacheToString(keeping)}` : ' Skipping, no cache entries.');
+  for (const entry of nodeCacheUsage) {
+      console.warn(`removing ${cacheToString(entry)}`);
+      cleaned += entry.sizeInBytes;
+      remove.push(entry.id);
+  }
+
+  // Cleanup everything else
+  console.log('TASK: clean up everything else that takes us over our threshold: ' + mb(CACHE_LIMIT));
   for (let i = cacheUsage.length - 1; i > 0; i--) {
-    const {id, key, msSinceLastAccessed, sizeInBytes} = cacheUsage[i];
-    total += sizeInBytes;
+    const cache = cacheUsage[i];
+    total += cache.sizeInBytes;
 
     // Are we in the danger zone?
     if (total > CACHE_LIMIT) {
-      console.warn(`[${hrs(msSinceLastAccessed).padStart(7)}] ${mb(sizeInBytes).padStart(7)} -> ${key}`);
-      cleaned += sizeInBytes;
-      remove.push(id);
+      console.warn(cacheToString(cacheUsage[i]));
+      cleaned += cache.sizeInBytes;
+      remove.push(cache.id);
     } else {
-      console.warn(`skip ${cacheUsage.length - i} ${mb(sizeInBytes)} ${hrs(msSinceLastAccessed)}`);
+      console.warn(`skip ${cacheUsage.length - i} ${mb(cache.sizeInBytes)} ${hrs(cache.msSinceLastAccessed)}`);
     }
   }
   console.warn(`Identifed ${remove.length} cache keys for removal, reducing cache from ${mb(total)} -> ${mb(total - cleaned)}`);
@@ -52,9 +81,9 @@ function main() {
   const cleanup = remove.map(id => `gh cache delete ${id} --repo facebook/react-native`);
   for (const cmd of cleanup) {
     if (dryRun) {
-      console.warn(`DRY: ${cmd}`);
+      console.warn(`Skip: ${cmd}`);
     } else {
-      console.warn(`${cmd} -> ${execSync(cmd, 'utf8').toString()}`);
+      console.warn(`${cmd} 🪓 ${execSync(cmd, 'utf8').toString()}`);
     }
   }
 }