-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: None
When running schema_abort (on my branch for WT-11176) with tiered I came across a seed that somehow loses the opts->tiered_flush_next_us. Therefore, the checkpoint thread never runs with flush and therefore, the parent never starts its timer to kill the child. I finally killed my test 15 minutes after. Running with the same seeds reproduced it again.
(I have no reason to think the problem is related to my changes.)
The command is:
test_schema_abort -m -PT -h WT_TEST-m -s 0 -T 6 -t 25 -PSD231042,E7537542
It appears that either the opts->tiered_flush_next_us is cleared and never set to start or the call to testutil_tiered_flush_complete to initialize isn't done.