ubi-hive

2021-12-14 update: The image here was converted to a standalone metastore with the postgres jdbc interface. The new standalone images will include metastore in their tag name.

Docker Image build of Hive from RHEL UBI base image. This includes mysql, postgresql jdbc interfaces.

Workflow

Basic workflow

1. Branch from main
2. Make changes
3. Increment build number file value
4. Commit and Push
5. Create PR merging into main
6. Get reviewed and approved

Build new version of Hive

1. Check available releases here. If current version is desired version, then stop unless you need a different hadoop image tag.
2. Checkout main and pull
3. Update Dockerfile : ENV METASTORE_VERSION setting it to the version required.
4. Execute get_metastore_version.sh to make sure that the output matches the new hive version.
5. Increment the value in image_build_num.txt using the make bump-image-tag command.
6. Run a test build by executing pr_check.sh.
7. If successful, remove LABEL quay.expires-after=3d from Dockerfile then commit changes and push branch.
8. Create a PR, this should execute a PR check script.
9. If successful, get approval and merge.

Utility Scripts

• get_metastore_version.sh : Get the current hive standalone-metastore version from the Dockerfile.
• get_image_tag.sh : Return the image tag made from the metastore version and the build number from image_build_num.txt
• docker-build-dev.sh : Executes a local test build of the docker image.

Integration Scripts

• pr_check.sh : PR check script (You should not need to modify this)
• build_deploy.sh : Build and deploy to Red Hat cloudservices quay org. (You should not need to modify this script)

Usage

This image supports arbitrary user ids by using the packaged /entrypoint.sh as container command to add UID and username.

Notice: the default username is hadoop, customize it by providing an environment variable named USER_NAME.