Organization Isolation Phase 1.5 is envisioned as an intermediate "waiting" phase between Isolation [Phase 1](https://gitlab.com/groups/gitlab-org/-/epics/11837) and [Phase 2](https://gitlab.com/groups/gitlab-org/-/epics/11838). ~"group::tenant scale" does not anticipate any "active" work during this phase. Rather, we will be waiting on other teams to complete their work, or run automations and wait for automation runs to finish before we can proceed to Phase 2. In Phase 1: - We kicked off automation runs to setup `sharding` keys and desired sharding keys. - Some tables could not be classified via automation, and teams have to do this manually or create issues for working on them. We are tracking this in https://gitlab.com/gitlab-org/gitlab/-/issues/455137. - We also kicked off some BG migrations to backfill sharding keys, but there are a lot of tables this needs to be run on and this will span multiple milestones. ## Goal for Organization Isolation Phase 1.5 - [x] We will kick off and wait for the work in https://gitlab.com/gitlab-org/gitlab/-/issues/455137 to finish. - [ ] Wait for all the issues created from within https://gitlab.com/gitlab-org/gitlab/-/issues/455137 to be completed by respective teams, so that all table classifications are complete. - [ ] Wait for all BG jobs that are backfilling sharding keys to be complete, across all tables. The end of Phase 1.5 will be marked when: - All tables will be classified as cell-local or clusterwide. - All tables will have sharding key present, defined and backfill completed. In essence, Phase 1.5 is just a "waiting phase" for mainly migrations to run and get completed, before we can move to Phase 2. See [this dashboard](https://cells-progress-tracker-gitlab-org-tenant-scale-g-f4ad96bf01d25f.gitlab.io/sharding_keys) to track progress. The only active work ~"group::tenant scale" may have to do is: - Fine-tuning our Housekeeper keeps to support the automations needed to run backfill migrations in an optimized manner. - Running the keeps themselves to generate MRs that other teams will review and merge. - Provide knowledge and support to other teams so that they are able to classify their tables, manually.