Use MOC OpenStack for long term data storage [EPIC]

[MichaelClifford]

We have been requesting a location to host public data for some time now (https://github.com/operate-first/support/issues/23) and the consensuses decision has been to use MOC Openstack accounts for storage rather than something on one of the opf clusters as they are more ephemeral (https://github.com/open-infrastructure-labs/ops-issues/issues/33).

To that end, I have created a public bucket specifically for AI4CI project here: https://kzn-swift.massopen.cloud/swift/v1/ai4ci/
This bucket should allow for users to read data without restriction, but only users with credentials can write/delete data.

Here are the docs for interacting with the bucket: https://docs.massopen.cloud/en/latest/openstack/Object-Storage.html

Please reach out to me if you are working on this and I can provide credentials or add you to the openstack project managing the bucket.

We should migrate the existing ai4ci work such that all data is stored here, rather than on the opf-datacatalog bucket on smaug.

Acceptance Criteria:

• [ ] Confirm data can be publicly read, but credentials are required to write/delete. Create short notebook to demo interaction
• [ ] Confirm Trino/Superset connections can be maintained through this bucket. Give demo to team
• [ ] Move all existing relevant data from opf-datacatalog to ai4ci.
• [ ] Update all notebooks and pipelines to target this bucket.

[antter]

Working on this first point:

https://github.com/aicoe-aiops/ocp-ci-analysis/issues/443