[SERVER-71765] Drop cluster parameter collections during automated restore Created: 01/Dec/22  Updated: 02/May/23

Status: Open
Project: Core Server
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Major - P3
Reporter: Varun Ravichandran Assignee: Backlog - Replication Team
Resolution: Unresolved Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
Assigned Teams:
Replication
Participants:

 Description   

When engaging the automated restore process, a target cluster is started from a snapshot of existing data, which may potentially include persisted non-default cluster parameter settings. Cloud has indicated that they do not wish for these settings to be automatically applied on the target cluster, preferring instead that the cluster start with all default cluster parameter values that they can then manually override via the setClusterParameter command. 

To this end, we should ensure that all cluster parameter namespace(s) are dropped on the config server AND all shards during the restoration process. On non-serverless clusters, this will involve dropping the config.clusterParameters collection. On serverless clusters, this will involve dropping all namespaces matching the form <tenant_id>_config.clusterParameters (e.g., every tenant's config database has its own cluster parameters collection).

After implementing this, we should test it by creating a snapshot that includes cluster parameter collections on the config server and all shards. After the target cluster has been initialized on the snapshot data and has completed the restoration process, we should assert that running getClusterParameter individually on each cluster node (mongods and mongoses) results in default values being returned rather than the values specified in the snapshot.


Generated at Thu Feb 08 06:19:54 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.