Sprintcraft is moving to Kubernetes

Hey @team,

Hopefully you did not notice it (except my mattermost message and that you need to log in), but Sprintcraft is running on Kubernetes from now on. We successfully migrated it.

Some deployment process changes have happened during the migration, but all are documented or will be documented next sprint. During the migration I copied over all data as is, so no data diff should be between the old and new instance.

As usual, the old instance is not cleaned up after we are confident that Sprintcraft works as expected.

If you find any issues, please let me know.

PS: @Agrendalath thank you again for the massive MR reviews.


The historic worklogs cache has already been re-generated (run 1, run 2). Kudos for doing a smooth migration, @gabor!

1 Like

@gabor Kudos for the work, and thanks for the notice! :+1:

Looks like it was still working fine for me, and I could login without issue.

PS: Can we also make this thread public, since it shouldn’t need to disclose client info, and sprintcraft is an open source project?

@antoviaque :pray: Done. This is now a public thread.

FTR: There is a little bit of chance that it will run out of memory during new sprint creation. However, if that happens, the persistent celery task queue should keep us running. If I see the signs of such a behaviour, I’ll increase memory.

1 Like