-
Type: Task
-
Resolution: Unresolved
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Storage Execution
-
14
When creating a time-series collection, first we create the buckets collection and then we create the view non-atomically. This allows for cases where the former can succeed but not the latter (for instance replication rollback). We should consider using an internal multi-document transaction here so that these two steps are performed atomically.
- fixes
-
SERVER-90862 Creating a collection or a view should fail if a bucket namespace exists without its view
- Backlog
- related to
-
SERVER-67736 Create time-series view and bucket collection in one storage transaction
- Backlog