Artifacts Buildkite Plugin

A Buildkite plugin for uploading and downloading artifacts.

Uploading artifacts

This functionality duplicates the artifact_paths property in the pipeline yaml files; with the difference that it also allows downloading artifacts and that this plugin is executed before any command hook, so you can create dependencies on artifacts in your steps that are resolved before the actual step is executed. This is ideal for producing an artifact in one job and then downloading it in a different job before execution.

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "log/**/*.log"

You can specify multiple files/globs to upload as artifacts:

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: [ "log/**/*.log", "debug/*.error" ]

And even rename them before uploading them (can not use globs here though, sorry):

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: 
          - from: log1.log
            to: log2.log

User-defined ACL on uploaded files

When using AWS S3 or Google Cloud Storage as your artifact store, you can optionally define an object-level ACL for your uploaded artifacts. This allows you to have granular control over which artifacts are made public or private.

If not specified it will respect the relevant setting at the agent level.

eg: uploading a public file when using S3

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "coverage-report/**/*"
        s3-upload-acl: public-read

eg: uploading a private file when using GS

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "coverage-report/**/*"
        gs-upload-acl: private

Downloading artifacts

This downloads artifacts matching globs to the local filesystem. See downloading artifacts for more details.

steps:
  - command: ...
    plugins:
      - artifacts#v1.9.3:
          download: "log/**/*.log"

You can specify multiple files/patterns:

steps:
  - command: ...
    plugins:
      - artifacts#v1.9.3:
          download: [ "log/**/*.log", "debug/*.error" ]

Rename particular files after downloading them:

steps:
  - command: ...
    plugins:
      - artifacts#v1.9.3:
          download: 
            - from: log1.log
              to: log2.log

And even do so from different builds/steps:

steps:
  - command: ...
    plugins:
      - artifacts#v1.9.3:
          step: UUID-DEFAULT
          build: UUID-DEFAULT-2
          download: 
            - from: log1.log
              to: log2.log
              step: UUID-1
            - from: log3.log
              to: log4.log
              build: UUID-2

Configuration

upload (string, array of strings, {from,to}, array of {from,to})

A glob pattern, or array of glob patterns, for files to upload.

download (string, array of strings, {from,to}, array of {from,to[,step][,build]})

A glob pattern, or array of glob patterns, for files to download.

step (optional, string)

The job UUID or name to download the artifacts from unless specified otherwise in the download array specification.

build (optional, string)

The build UUID to download the artifact from unless specificed otherwise in the download array specification.

compressed (optional, string)

⚠️ Limitations:

• filename needs to end with .zip or .tgz and that will determine the compression executable to use
• path globs (*) are interpreted by agent's shell and (un)compressing program, meaning that * and ** will not work.

When uploading, the file or directory specified in the upload option will be compressed in a single file with this name and uploaded as a single artifact. The following example will get the directory matching log/my-folder, zip them up and upload a single artifact file named logs.zip:

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "log/my-folder"
        compressed: logs.zip

When downloading, this option states the actual name of the artifact to be downloaded in the download option will be extracted off of it. The following example will download the logs.tgz artifact and extract all files in it matching log/file.log:

steps:
  - command: ...
    plugins:
      - artifacts#v1.9.3:
          download: "log/file.log"
          compressed: logs.tgz

ignore-missing (optional, boolean)

If set to true, it will ignore errors caused when calling buildkite-agent artifact to prevent failures if you expect artifacts not to be present in some situations. When using the compressed property, it will ignore compressing the artifacts that are not present.

skip-on-status (optional, integer or array of integers, uploads only)

You can set this to the exit codes or array of exit codes of the command step (as defined by the BUILDKITE_COMMAND_EXIT_STATUS variable) that will cause the plugin to avoid trying to upload artifacts.

This should allow you to specify known failure conditions that you want to avoid uploading artifacts. For example, because you know logs will be huge or not useful.

Skip uploading if the main command failed with exit code 147:

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "log/*.log"
        skip-on-status: 147

Alternatively, skip artifact uploading on exit codes 1 and 5:

steps:
  - command: ...
    plugins:
    - artifacts#v1.9.3:
        upload: "log/*.log"
        skip-on-status:
          - 1
          - 5

Developing

To run testing, shellchecks and plugin linting use use bk run with the Buildkite CLI.

bk run

Or if you want to run just the tests, you can use the docker Plugin Tester:

docker run --rm -ti -v "${PWD}":/plugin buildkite/plugin-tester:latest

License

MIT (see LICENSE)