Details
-
Improvement
-
Resolution: Done
-
Major - P3
-
None
-
None
Description
In regards to: http://mms.mongodb.com/help/backup/tutorial/enable-backup-for-sharded-cluster/
At present our explanation for what a Cluster Checkpoint is is:
A cluster checkpoint is a valid point in time when a sharded cluster can be restored. You must pause the sharded cluster balancer to create each checkpoint.
We should expand on this to explain/clarify the following:
- We state "You must pause the sharded cluster balancer" - The "you" implies to me that we expect the user to do this. We should reword to be clear on the source of the pause.
- Cluster checkpoints are light weight - requiring oplog token insertion on mongod and configdb instances. We should clarify this as being light weight also means negligible backup-charge cost for the cloud and negligible disk space for on-prem.
- Our default for checkpoints is disabled. We should explain why - meaning what are the downsides of enabling (like a short disruption of the balancing process for example).
- There has been confusion as to whether checkpoints are required in order to have a restorable sharded backup. We should state clearly that they are there to allow for additional points in time that can be used for restore.
Attachments
Issue Links
- depends on
-
DOCS-3302 MMS OnPrem Backup: Document failover procedures if a Daemon fails
-
- Closed
-