[Feature flag] Transition Sidekiq duplicate jobs to RedisSharedState

Summary

This issue is to rollout the transition of Sidekiq duplicate jobs data to SharedState on production, that is currently behind the use_primary_and_secondary_stores_for_duplicate_jobs and use_primary_store_as_default_for_duplicate_jobs feature flags.

Owners

Team: groupscalability
Most appropriate slack channel to reach out to: #scalability-423-sidekiq-zonal-clusters
Best individual to reach out to: @alejandro
PM: @lmcandrew

Stakeholders

Expectations

What are we expecting to happen?

We enable use_primary_and_secondary_stores_for_duplicate_jobs. Data is written to both the Sidekiq store and the SharedState store. Some ReadFromPrimaryError and PipelinedDiffError errors are expected and normal at this stage.
ReadFromPrimaryError and PipelinedDiffError errors disappear after a short period of time. We should wait at least 6 hours, the default TTL for keys related to this feature before proceeding (note: scheduled jobs may have a different TTL).
We enable use_primary_store_as_default_for_duplicate_jobs. SharedState becomes the main store. There are no ReadFromPrimaryError nor PipelinedDiffError errors.
We disable use_primary_and_secondary_stores_for_duplicate_jobs. The Sidekiq store stops being used completely.
Transition is complete. An MR to remove the MultiStore and make SharedState the only store should follow.

When is the feature viable?

None

What might happen if this goes wrong?

MultiStore could have a breaking bug, causing widespread errors. Disabling both flags bypasses MultiStore and returns to the old status quo
If ReadFromPrimaryError and PipelinedDiffError persist after 6 hours but functionality is unaffected, duplicate jobs data may not be persisted correctly to SharedState. Thus we wouldn't be able to continue with the transition, we should disable both flags and investigate.

What can we monitor to detect problems with this?

Sidekiq: Overview - https://dashboards.gitlab.net/d/sidekiq-main/sidekiq-overview?orgId=1. Note especially Service Apdex, Service Error Ratio, and Queue lengths

What can we check for monitoring production after rollouts?

Messages with instance_name: DuplicateJobs:
- gstg: https://nonprod-log.gitlab.net/goto/82cd25b0-ecae-11ec-b771-57a829f2c394
- gprd: https://log.gprd.gitlab.net/goto/67c69b40-fea9-11ec-b86b-d963a1a6788e
Jobs with job_status: deduplicated https://log.gprd.gitlab.net/goto/9d6ad1f0-e584-11ec-aade-19e9974a7229. Be on the lookout for abrupt increases or decreases in volume after changing the feature flags. Unrestrained growth or a drop to zero should immediately trigger a rollback.

Rollout Steps

Rollout on non-production environments

Specific rollout on production

Ensure that the feature MRs have been deployed to both production and canary.
- /chatops run auto_deploy status <merge-commit-of-your-feature>
If you're using project-actor, you must enable the feature on these entries:
- /chatops run feature set --project=gitlab-org/gitlab,gitlab-org/gitlab-foss,gitlab-com/www-gitlab-com <feature-flag-name> true
If you're using group-actor, you must enable the feature on these entries:
- /chatops run feature set --group=gitlab-org,gitlab-com <feature-flag-name> true
If you're using user-actor, you must enable the feature on these entries:
- /chatops run feature set --user=<your-username> <feature-flag-name> true
Verify that the feature works on the specific entries. Posting the QA result in this issue is preferable.

Preparation before global rollout

Check if the feature flag change needs to be accompanied with a change management issue. Cross link the issue here if it does.
Ensure that you or a representative in development can be available for at least 2 hours after feature flag updates in production. If a different developer will be covering, or an exception is needed, please inform the oncall SRE by using the @sre-oncall Slack alias.
Ensure that documentation has been updated (More info).
Announce on the feature issue an estimated time this will be enabled on GitLab.com.
Ensure that any breaking changes have been announced following the release post process to ensure GitLab customers are aware.
Notify #support_gitlab-com and your team channel (more guidance when this is necessary in the dev docs).

Global rollout on production

For visibility, all /chatops commands that target production should be executed in the #production slack channel and cross-posted (with the command results) to the responsible team's slack channel (#g_TEAM_NAME).

Ensure that the feature MRs have been deployed to non-production environments.
- /chatops run auto_deploy status [id of your commit]
Enable use_primary_and_secondary_stores_for_duplicate_jobs globally on non-production environments.
- /chatops run feature set use_primary_and_secondary_stores_for_duplicate_jobs true
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, and gitlab_redis_multi_store_read_fallback_total. After at most 6 hours, they should both be zero. Also monitor gitlab_redis_multi_store_method_missing_total. It should always be zero. Link to Thanos
Enable use_primary_store_as_default_for_duplicate_jobs globally on non-production environments.
- /chatops run feature set use_primary_store_as_default_for_duplicate_jobs true
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, gitlab_redis_multi_store_read_fallback_total, and gitlab_redis_multi_store_method_missing_total. They should all be zero. Link to Thanos
Disable use_primary_and_secondary_stores_for_duplicate_jobs globally on non-production environments, so that SharedState is used as the default store.
- /chatops run feature set use_primary_and_secondary_stores_for_duplicate_jobs false
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, gitlab_redis_multi_store_read_fallback_total, and gitlab_redis_multi_store_method_missing_total. They should all be zero. Link to Thanos
Announce on the feature issue that the feature has been globally enabled.
Wait for at least one day for the verification term.

(Optional) Release the feature with the feature flag

If you're still unsure whether the feature is deemed stable but want to release it in the current milestone, you can change the default state of the feature flag to be enabled. To do so, follow these steps:

WARNING: This approach has the downside that it makes it difficult for us to clean up the flag. For example, on-premise users could disable the feature on their GitLab instance. But when you remove the flag at some point, they suddenly see the feature as enabled and they can't roll it back to the previous behavior. To avoid this potential breaking change, use this approach only for urgent matters.

Release the feature

After the feature has been deemed stable, the clean up should be done as soon as possible to permanently enable the feature and reduce complexity in the codebase.

You can either create a follow-up issue for Feature Flag Cleanup or use the checklist below in this same issue.

Rollback Steps

Enable use_primary_and_secondary_stores_for_duplicate_jobs, so that we go back to using both stores.
- /chatops run feature set use_primary_and_secondary_stores_for_duplicate_jobs true
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, and gitlab_redis_multi_store_read_fallback_total. After at most 6 hours, they should both be zero. Also monitor gitlab_redis_multi_store_method_missing_total. It should always be zero. Link to Thanos
Disble use_primary_store_as_default_for_duplicate_jobs to go back to using the Queues store as the primary
- /chatops run feature set use_primary_store_as_default_for_duplicate_jobs false
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, gitlab_redis_multi_store_read_fallback_total, and gitlab_redis_multi_store_method_missing_total. They should all be zero. Link to Thanos
Disable use_primary_and_secondary_stores_for_duplicate_jobs so that Queues is the only store used
- /chatops run feature set use_primary_and_secondary_stores_for_duplicate_jobs false
Verify that the feature works as expected.
- Monitor the Prometheus metrics gitlab_redis_multi_store_pipelined_diff_error_total, gitlab_redis_multi_store_read_fallback_total, and gitlab_redis_multi_store_method_missing_total. They should all be zero. Link to Thanos

Edited Aug 03, 2022 by Alejandro Rodríguez