halfpipe-ml-deploy

Deploy xquery apps to MarkLogic with ml-deploy.

There are 2 commands for deploying:

• /ml-deploy/deploy-local-zip deploy a local zip file. e.g. one created in a build task
• /ml-deploy/deploy-ml-modules deploy a version of the shared ml modules library from artifactory

All arguments are passed as environment variables.

APP_VERSION is optional. By default it is set to the current git revision. So apps should reach them like this: http://{MARKLOGIC_HOST}:7655/{APP_NAME}/{GIT_REVISION}/some/endpoint.xqy

Setting USE_BUILD_VERSION: true will use the environment variable BUILD_VERSION instead of GIT_REVISION.

It is possible to deploy to more than one MARKLOGIC_HOST by providing a comma separated list of hostnames.

Halfpipe examples

There are halfpipe task types for the 2 commands:

deploying a local zip created in build task

tasks:
- type: deploy-ml-zip
  deploy_zip: target/xquery.zip
  app_name: example-app       # optional. defaults to pipeline name
  app_version: v1             # optional. defaults to GIT_REVISION
  use_build_version: true     # optional. Use BUILD_VERSION instead of GIT_REVISION. defaults to false
  targets:
  - marklogic.host
  ...

deploying a version of the shared ml modules library from artifactory to 2 MarkLogic hosts

tasks:
- type: deploy-ml-modules
  name: Deploy XQuery to Dev and QA
  ml_modules_version: "2.1428"
  app_name: halfpipe-example  # optional. defaults to pipeline name
  app_version: v1             # optional. defaults to GIT_REVISION
  use_build_version: true     # optional. Use BUILD_VERSION instead of GIT_REVISION. defaults to false
  targets:
  - marklogic.dev.host
  - marklogic.qa.host