-
Type: Task
-
Resolution: Done
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
Replication
-
Fully Compatible
-
v8.0
-
Repl 2024-04-01, Repl 2024-04-15, Repl 2024-04-29, Repl 2024-05-13, Repl 2024-05-27, Repl 2024-06-10, Repl 2024-06-24, Repl 2024-07-08
The mongo layer implicitly has an assumption that it always initiates all checkpoints in the storage engine. However, there are several situations where WT might internally take a checkpoint (the most recent one being background compaction) without the mongo layer being aware, which means that this assumption is not correct.
We should investigate whether there are areas that are relying on this assumption, especially if they could affect correctness of the system.
- is related to
-
WT-12333 Provide mechanism to (re)set whether checkpoints should default to stable or unstable
- Backlog
-
SERVER-84706 Investigate if setting the oldest timestamp greater than the stable timestamp can be avoided
- Closed
- related to
-
SERVER-91841 Repaired nodes rejoining replica set can hit invariant
- Closed