OSD Clusters Checker

A friendly helper which scans for OSD clusters and sees if anyone of them is worth cleaning up or not to save some bucks :)

Execution

Prerequisities

The following credentials are required to run the OSD Clusters Checker:

• OCM_TOKEN: OCM Token which allows you to read your organization's OSD clusters via OCM API.
• SLACK_TOKEN: Slack Token which is authorized to post message on your Slack handle.

Running the Clusters Checker

• Have Go 1.17 on your host.
• Build the clusters-checker:

make build

• Run it:

export OCM_TOKEN="foo"
export SLACK_TOKEN="bar"

bin/clusters-checker scan --envs-and-org-ids production:your-org-id-in-production --envs-and-org-ids staging:your-org-id-in-stagubg --envs-and-org-ids integration:your-org-id-in-integration  --slack-channel-id channel-id

Cleanup

• Run the following:

make clean

Release and Deployment

The deployment of this tool is GitOps-driven over the internal Red Hat's app-interface pipelines.

Those pipelines recognise a change under ./deploy/cronjob-template.yaml and accordingly, re-deploy the cronjob.

Hence, to cut a new release against this whole process, raise a PR to update the CronJob's image with the latest image pushed to Quay registry corresponding to quay.io/mtsre/mtsre-clusters-checker.

Contributions

Feel free to suggest anything by raising an Issue on this repository, getting a clearance from the maintainers and proceeding to work on its associated PR.

Future?

• Carve this into a CLI which can be used by any team in the future to perform the same kind of checks for the Openshift clusters belonging to their org.
• Add support for skip lists to specify the clusters-checker to ignore certain clusters while performing the scan.