-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: None
-
None
-
Storage Engines, Storage Engines - Server Integration
-
Fully Compatible
-
ALL
-
SESISailing - 2025-12-09
-
200
-
None
-
None
-
None
-
None
-
None
-
None
-
None
- db_disagg_storage_test has hit 10 minute timeout but logs are only 1 mins worth, followed by a core dump (bc we hit 10min)
- this means we are hanging somewhere in the test
- pushArtificialDeadline basically pushes another deadline onto a stack of deadlines. When the top deadline is met, the stack pops to the next deadline.
- Basically we are setting the deadline to be 999999999ms, then setting it again to be 5ms (the top of stack is now 5ms seeĀ here). But there is a race happening.
- When the writer Writes, it also advances the time by 10ms in order for the test to proceed and fulfil the deadline here. If the Write gets called before we check if the opCtx is dead here, then we will enter the sad path here, and sleep for the next deadline aka 99999999ms.
- We must ensure that the Write is called after we check if the opCtx is dead.