Helm is a package manager for Kubernetes, the project is a helm chart for Hive Metastore. The container is lifted from https://github.com/realtimedatalake/hive-metastore-docker.
DLink is the data lakehouse developing by deepexi.com. Hive-metastore is the sub project of DLink.
The following prerequisites are required for a successful and properly secured use of Helm.
- A Kubernetes cluster
- Deciding what security configurations to apply to your installation, if any
- Installing and configuring Helm.
Please go to https://helm.sh/docs/intro/quickstart for details.
- A database, e.g.postgresql or mysql
Create a database for hms, e.g. hms_db. Create a role/user and grant the read/write/login rights to the role/user, e.g. deep_hms.
There is example sql located in docker/script/create-user-db.postgres.sql, you can ask DBA to perform it.
We can get following values from dba if this step works right:
connectionURL: jdbc:postgresql://host.docker.internal:5432/hms_db
user: deep_hms
password: deep_hms
In this step, we need to setup tables and insert init data for hms. Hms provides commands for that:
schematool -initSchema -dbType postgres
Since setuping env to run above command is diffcult, we provide docker-compose.init.yml to make it easy. Please check the file in docker folder. Go to docker folder, trigger command:
docker compose -f docker-compose.init.yml up
Ensure the command completed with EXITED(0)
Clone this repo and go into the folder. Install helm chart with:
helm install hms-s3 . -f values-hms-example.yaml
'hms-s3' is the release name for the hms. '.' is the filepath of the chart.
Uninstall the chart by:
helm uninstall hms-s3