Merge pull request #1584 from actions/robherley/upload-v4-improvements

Increase Artifact v4 upload speed

package-lock.json

@@ -3262,9 +3262,9 @@
       }
     },
     "node_modules/axios": {
-      "version": "1.4.0",
+      "version": "1.6.2",
-      "resolved": "https://registry.npmjs.org/axios/-/axios-1.4.0.tgz",
+      "resolved": "https://registry.npmjs.org/axios/-/axios-1.6.2.tgz",
-      "integrity": "sha512-S4XCWMEmzvo64T9GfvQDOXgYRDJ/wsSZc7Jvdgx5u1sd0JwsuPLqb3SYmusag+edF6ziyMensPVqLTSc1PiSEA==",
+      "integrity": "sha512-7i24Ri4pmDRfJTR7LDBhsOTtcm+9kjX5WiY1X3wIisx6G9So3pfMkEiU7emUBe46oceVImccTEM3k6C5dbVW8A==",
       "dev": true,
       "dependencies": {
         "follow-redirects": "^1.15.0",

packages/artifact/src/internal/shared/artifact-twirp-client.ts

@@ -52,8 +52,8 @@ class ArtifactHttpClient implements Rpc {
     contentType: 'application/json' | 'application/protobuf',
     data: object | Uint8Array
   ): Promise<object | Uint8Array> {
-    const url = `${this.baseUrl}/twirp/${service}/${method}`
+    const url = new URL(`/twirp/${service}/${method}`, this.baseUrl).href
-    debug(`Requesting ${url}`)
+    debug(`Requesting: ${url}`)
     const headers = {
       'Content-Type': contentType
     }

packages/artifact/src/internal/shared/config.ts

@@ -1,3 +1,5 @@
+import os from 'os'
+
 // Used for controlling the highWaterMark value of the zip that is being streamed
 // The same value is used as the chunk size that is use during upload to blob storage
 export function getUploadChunkSize(): number {
@@ -17,7 +19,8 @@ export function getResultsServiceUrl(): string {
   if (!resultsUrl) {
     throw new Error('Unable to get the ACTIONS_RESULTS_URL env variable')
   }
-  return resultsUrl
+
+  return new URL(resultsUrl).origin
 }
 
 export function isGhes(): boolean {
@@ -34,3 +37,17 @@ export function getGitHubWorkspaceDir(): string {
   }
   return ghWorkspaceDir
 }
+
+// Mimics behavior of azcopy: https://learn.microsoft.com/en-us/azure/storage/common/storage-use-azcopy-optimize
+// If your machine has fewer than 5 CPUs, then the value of this variable is set to 32.
+// Otherwise, the default value is equal to 16 multiplied by the number of CPUs. The maximum value of this variable is 300.
+export function getConcurrency(): number {
+  const numCPUs = os.cpus().length
+
+  if (numCPUs <= 4) {
+    return 32
+  }
+
+  const concurrency = 16 * numCPUs
+  return concurrency > 300 ? 300 : concurrency
+}

packages/artifact/src/internal/shared/interfaces.ts

@@ -38,6 +38,17 @@ export interface UploadOptions {
    * input of 0 assumes default retention setting.
    */
   retentionDays?: number
+  /**
+   * The level of compression for Zlib to be applied to the artifact archive.
+   * The value can range from 0 to 9:
+   * - 0: No compression
+   * - 1: Best speed
+   * - 6: Default compression (same as GNU Gzip)
+   * - 9: Best compression
+   * Higher levels will result in better compression, but will take longer to complete.
+   * For large files that are not easily compressed, a value of 0 is recommended for significantly faster uploads.
+   */
+  compressionLevel?: number
 }
 
 /*****************************************************************************

packages/artifact/src/internal/upload/blob-upload.ts

@@ -1,7 +1,7 @@
 import {BlobClient, BlockBlobUploadStreamOptions} from '@azure/storage-blob'
 import {TransferProgressEvent} from '@azure/core-http'
 import {ZipUploadStream} from './zip'
-import {getUploadChunkSize} from '../shared/config'
+import {getUploadChunkSize, getConcurrency} from '../shared/config'
 import * as core from '@actions/core'
 import * as crypto from 'crypto'
 import * as stream from 'stream'
@@ -29,13 +29,13 @@ export async function uploadZipToBlobStorage(
 ): Promise<BlobUploadResponse> {
   let uploadByteCount = 0
 
-  const maxBuffers = 5
+  const maxConcurrency = getConcurrency()
   const bufferSize = getUploadChunkSize()
   const blobClient = new BlobClient(authenticatedUploadURL)
   const blockBlobClient = blobClient.getBlockBlobClient()
 
   core.debug(
-    `Uploading artifact zip to blob storage with maxBuffers: ${maxBuffers}, bufferSize: ${bufferSize}`
+    `Uploading artifact zip to blob storage with maxConcurrency: ${maxConcurrency}, bufferSize: ${bufferSize}`
   )
 
   const uploadCallback = (progress: TransferProgressEvent): void => {
@@ -61,7 +61,7 @@ export async function uploadZipToBlobStorage(
     await blockBlobClient.uploadStream(
       uploadStream,
       bufferSize,
-      maxBuffers,
+      maxConcurrency,
       options
     )
 

packages/artifact/src/internal/upload/upload-artifact.ts

@@ -37,7 +37,10 @@ export async function uploadArtifact(
     }
   }
 
-  const zipUploadStream = await createZipUploadStream(zipSpecification)
+  const zipUploadStream = await createZipUploadStream(
+    zipSpecification,
+    options?.compressionLevel
+  )
 
   // get the IDs needed for the artifact creation
   const backendIds = getBackendIdsFromToken()

packages/artifact/src/internal/upload/zip.ts

@@ -5,6 +5,8 @@ import {createReadStream} from 'fs'
 import {UploadZipSpecification} from './upload-zip-specification'
 import {getUploadChunkSize} from '../shared/config'
 
+export const DEFAULT_COMPRESSION_LEVEL = 6
+
 // Custom stream transformer so we can set the highWaterMark property
 // See https://github.com/nodejs/node/issues/8855
 export class ZipUploadStream extends stream.Transform {
@@ -21,14 +23,16 @@ export class ZipUploadStream extends stream.Transform {
 }
 
 export async function createZipUploadStream(
-  uploadSpecification: UploadZipSpecification[]
+  uploadSpecification: UploadZipSpecification[],
+  compressionLevel: number = DEFAULT_COMPRESSION_LEVEL
 ): Promise<ZipUploadStream> {
+  core.debug(
+    `Creating Artifact archive with compressionLevel: ${compressionLevel}`
+  )
+
   const zip = archiver.create('zip', {
-    zlib: {level: 9} // Sets the compression level.
+    highWaterMark: getUploadChunkSize(),
-    // Available options are 0-9
+    zlib: {level: compressionLevel}
-    // 0 => no compression
-    // 1 => fastest with low compression
-    // 9 => highest compression ratio but the slowest
   })
 
   // register callbacks for various events during the zip lifecycle