Rolling Maintenance Procedure
Parent: Atlas Maintenance Windows · Published reference · snapshot 2026-10-15
↓ Facts as markdownall context files
Depth-first rabbithole dossier for Rolling Maintenance Procedure; source-anchored research pack.
These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.
How it works
- [low] Stepping down the primary is preferable to directly shutting it down because stepping down expedites the failover procedure. Sources: https://www.mongodb.com/docs/manual/tutorial/upgrade-revision.md [source]
- [low] When restarting `mongod` as a standalone instance for maintenance, set the `disableLogicalSessionCacheRefresh` parameter to `true` in the `setParameter` section, and be sure to remove this parameter when restarting the instance as a replica set member. Sources: https://www.mongodb.com/docs/manual/tutorial/perform-maintenance-on-replica-set-members.md [source]
Facts and statements
- [medium] Atlas deploys major version upgrades and executes maintenance or infrastructure operations, such as applying security patches or scaling, using a rolling update strategy to minimize downtime and preserve cluster availability. Sources: https://www.mongodb.com/docs/atlas/architecture/current/operational-readiness-checklist.md, https://www.mongodb.com/docs/atlas/reference/faq/deployment.md [source]
- [low] Atlas uses a rolling-migration strategy for moving nodes from an original region to a new region, rolling the migration cluster by cluster starting from the secondary nodes. Sources: https://www.mongodb.com/docs/atlas/reference/faq/deployment.md [source]
- [low] To prevent a member from being elected primary again for five minutes during maintenance, specify a 300-second waiting period. Sources: https://www.mongodb.com/docs/manual/tutorial/perform-maintenance-on-replica-set-members.md [source]
- [low] Maintenance on a secondary involves stopping it, restarting it as a standalone on a different port, performing maintenance operations, and then restarting `mongod` as a member of the replica set. Sources: https://www.mongodb.com/docs/manual/tutorial/perform-maintenance-on-replica-set-members.md [source]
- [low] Maintenance on the primary should be performed last, and after the primary steps down, the replica set will elect a new primary. Sources: https://www.mongodb.com/docs/manual/tutorial/perform-maintenance-on-replica-set-members.md [source]
- [low] After performing all maintenance tasks on a standalone instance, restart the `mongod` instance as a replica set member with its original configuration to undo the configuration changes made for standalone operation. Sources: https://www.mongodb.com/docs/manual/tutorial/perform-maintenance-on-replica-set-members.md [source]
Related concepts
- Maintenance — is a part of Rolling Maintenance Procedure
- Rolling — is a part of Rolling Maintenance Procedure
- Procedure — is a part of Rolling Maintenance Procedure
Children
- No children recorded.