Manage Maintenance Window configuration
Periodic maintenance is a mandatory part of the FPT MongoDB Enterprise service, not an optional feature. What you control is when it runs, and the maintenance window behaves the same here as on every other engine FPT Database Engine manages.
The full procedure is documented once in Change maintenance window.
Not available on a Sharded Cluster. Its detail page has no Maintenance tab — the tab set is Overview, Backup, Resource, Notification, and Action logs. A maintenance window is still set in the provisioning wizard, but there is no way to review or change it afterwards from the console.
What this means for your MongoDB clusters:
- The cycle is weekly, the window belongs to one cluster, and maintenance runs for one hour from the start time you set.
- If you never configure a window, the system uses a default of 23:00 every Saturday.
- You can only change the schedule while no maintenance event is already pending. Once one is scheduled, the window is locked until it completes.
- Notifications go out 7 days and 1 day before the event, by email only — Telegram does not support the Maintenance service.
- Critical patches may run as emergency maintenance outside your window. Choosing off-peak hours reduces impact but does not eliminate it.
Do not operate on the cluster while maintenance is running. Configuration changes, system modifications, and starting or stopping services during the window can degrade performance and stability, or produce unexpected errors. Resume once the event has finished.
Next steps
- Monitor & alert to set who receives maintenance emails, and to watch cluster health around the window
- Upgrade database version for the other planned event that restarts the cluster