Details
-
Bug
-
Resolution: Unresolved
-
Major
-
Morpheus
-
None
-
Untriaged
-
0
-
Unknown
Description
In a customer setup, it has been observed during recovery that one index is missing disk snapshots while all other indexes were able to recover from disk snapshot.
Since all indexes of a stream start together, other indexes with valid disk snapshots have also been rolled back to zero. Instead of rolling back all indexes to zero, indexer can make two groups: One group of indexes with valid restartTs and another group of indexes with zero restartTs and build these two groups separately.
Attachments
Issue Links
- relates to
-
MB-47544 Reset Index on storage corruption
- Open