[08/16/2022 - 00:00 UTC] - GSTG - Disable post failover maintenance script in the Patroni 2004

Production Change

Change Summary

We need to disable post failover/switchover maintenance in the Patroni 2004 cluster, because the vacumdb -analyze-only processes that starts 30 minutes after the promotion of a node will conflict with the GIN reindex CONCURRENTLY processes that are required in the OS Upgrade as pointed out at db-migration!299 (comment 1059259353)

The GIN reindexing can take several hours to execute.

The post-failover-maintenance.sh currently is only necessary after a DB MVU (Major Version Upgrade) to analyze and collect tables statistics that are zeroed during a MVU.

This MR needs to be applied the CR [08/17/2022 - 00:00 UTC] - Patroni Clusters OS Upgrade - GSTG - Final - #7577 (closed)

Change Details

  1. Services Impacted - ServicePatroni ServicePatroniCI
  2. Change Technician - @rhenchen.gitlab
  3. Change Reviewer - @bshah11
  4. Time tracking - 1 hour
  5. Downtime Component - yes

Detailed steps for the change

Change Steps - steps to take to execute the change

Estimated Time to Complete (mins) - 1 hour

  1. Set label changein-progress /label ~change::in-progress
  2. Get green light from @sre-oncall and @release-managers
  3. Take node of the the current patroni leader nodes: 
    ssh patroni-ci-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl list"
ssh patroni-main-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl list"
ssh patroni-ci-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl list"
ssh patroni-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl list"
  4. Disable Chef Client on both gstg patroni-2004 clusters 
    knife ssh "roles:gstg-base-db-patroni-2004" "sudo chef-client-disable"
knife ssh "roles:gstg-base-db-patroni-ci-2004" "sudo chef-client-disable"
knife ssh "roles:gstg-base-db-patroni" "sudo chef-client-disable"
knife ssh "roles:gstg-base-db-patroni-ci" "sudo chef-client-disable"
  5. Merge MR - https://gitlab.com/gitlab-com/gl-infra/chef-repo/-/merge_requests/2188
  6. Re-enable and Execute Chef Client in just one Reader node, to evaluate if Patroni service will restart with the modification 
    ssh <reader_fqdn> "sudo chef-client-enable; sudo chef-client;"
  7. Che the logs to Evaluate if Chef restarted the Patroni service with the modification 
    ssh <reader_fqdn> "tail -n 10000 /var/log/gitlab/patroni/patroni.log"
ssh <reader_fqdn> "tail -n 10000 /var/log/gitlab/postgresql/postgresql.log"
  8. Check if the on_role_change setting have changed on the evaluation node: 
    ssh <reader_fqdn> "sudo cat /var/opt/gitlab/patroni/patroni.yml | grep on_role_change"
  9. Re-enable and Execute Chef Client in all nodes of both gstg patroni-2004 clusters 
    knife ssh "roles:gstg-base-db-patroni-2004" "sudo chef-client-enable; sudo chef-client;"
knife ssh "roles:gstg-base-db-patroni-ci-2004" "sudo chef-client-enable; sudo chef-client;"
knife ssh "roles:gstg-base-db-patroni" "sudo chef-client-enable; sudo chef-client;"
knife ssh "roles:gstg-base-db-patroni-ci" "sudo chef-client-enable; sudo chef-client;"
  10. Check if the on_role_change setting have changed for each node in the 2004 patroni clusters in gstg: 
    ssh_cluster_regex.sh "patroni.*gstg" "sudo cat /var/opt/gitlab/patroni/patroni.yml | grep on_role_change"
  11. Check Patroni logs and postgresql logs on the existing leaders to see if there was any restart/promotion
  12. Force reload of the Patroni clusters config files (no downtime) 
    ssh patroni-main-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload gstg-patroni-main-pg12-2004"
ssh patroni-ci-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload gstg-patroni-ci-2004"
ssh patroni-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload pg12-ha-cluster-stg"
ssh patroni-ci-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload gstg-patroni-ci"
  13. Force a switchover of the Patroni clusters to validate if the change had effect Downtime = up to 5 minutes 
    ssh patroni-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl switchover pg12-ha-cluster-stg"
ssh patroni-ci-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl switchover gstg-patroni-ci"
  14. Check Patroni logs on the new cluster Leaders
  15. Set label changecomplete /label ~change::complete

Rollback

Rollback steps - steps to be taken in the event of a need to rollback this change

Estimated Time to Complete (mins) - 15 minutes

  1. Revert MR - https://gitlab.com/gitlab-com/gl-infra/chef-repo/-/merge_requests/2188
  2. Execute Chef Client on both gstg patroni-2004 clusters 
    knife ssh "roles:gstg-base-db-patroni-2004" "sudo chef-client"
knife ssh "roles:gstg-base-db-patroni-ci-2004" "sudo chef-client"
  3. Check if the on_role_change setting have changed for each node in the 2004 patroni clusters in gstg: 
    ssh_cluster_regex.sh "patroni.*2004.*gstg" "sudo cat /var/opt/gitlab/patroni/patroni.yml | grep on_role_change"
  4. Force reload of the Patroni clusters config files (no downtime) 
    ssh patroni-ci-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload gstg-patroni-main-pg12-2004"
ssh patroni-ci-2004-01-db-gstg.c.gitlab-staging-1.internal "sudo gitlab-patronictl reload gstg-patroni-ci-2004"
  5. Set label changeaborted /label ~change::aborted

Monitoring

  • Chek if there was a node restart after the chef-client execution

    • Location: 
      tail -n 10000 /var/log/gitlab/patroni/patroni.log
tail -n 10000 /var/log/gitlab/postgresql/postgresql.log
    • If there was any node restart after the chef-client execution the change in Production needs to be made node by node

  • Chek if the post-failover-maintenance.sh is being executed in the new Writer node after the switchover

    • Location: 
      ps -ef "/var/opt/gitlab/patroni/scripts/post-failover-maintenance.sh"
    • If the script is being executed after a promotion, the change didn't worked, rollback the change

Key metrics to observe

There are no metrics that can point the sucess/failure of this change, the evaluation needs to be done through log analyzis

Change Reviewer checklist

C4 C3 C2 C1:

  • Check if the following applies:
    • The scheduled day and time of execution of the change is appropriate.
    • The change plan is technically accurate.
    • The change plan includes estimated timing values based on previous testing.
    • The change plan includes a viable rollback plan.
    • The specified metrics/monitoring dashboards provide sufficient visibility for the change.

C2 C1:

  • Check if the following applies:
    • The complexity of the plan is appropriate for the corresponding risk of the change. (i.e. the plan contains clear details).
    • The change plan includes success measures for all steps/milestones during the execution.
    • The change adequately minimizes risk within the environment/service.
    • The performance implications of executing the change are well-understood and documented.
    • The specified metrics/monitoring dashboards provide sufficient visibility for the change.
      • If not, is it possible (or necessary) to make changes to observability platforms for added visibility?
    • The change has a primary and secondary SRE with knowledge of the details available during the change window.
    • The labels blocks deployments and/or blocks feature-flags are applied as necessary

Change Technician checklist

  • Check if all items below are complete:
    • The change plan is technically accurate.
    • This Change Issue is linked to the appropriate Issue and/or Epic
    • Change has been tested in staging and results noted in a comment on this issue.
    • A dry-run has been conducted and results noted in a comment on this issue.
    • For C1 and C2 change issues, the change event is added to the GitLab Production calendar.
    • For C1 and C2 change issues, the SRE on-call has been informed prior to change being rolled out. (In #production channel, mention @sre-oncall and this issue and await their acknowledgement.)
    • For C1 and C2 change issues, the SRE on-call provided approval with the eoc_approved label on the issue.
    • For C1 and C2 change issues, the Infrastructure Manager provided approval with the manager_approved label on the issue.
    • Release managers have been informed (If needed! Cases include DB change) prior to change being rolled out. (In #production channel, mention @release-managers and this issue and await their acknowledgment.)
    • There are currently no active incidents that are severity1 or severity2
    • If the change involves doing maintenance on a database host, an appropriate silence targeting the host(s) should be added for the duration of the change.
Edited by Rafael Henchen