Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
1.70.0
-
None
Description
This is a follow-up to OAK-11069: Apparently the refresher thread still dies without a trace. I'm therefore assuming some java.lang.Error occurs, which is hidden behind the executor. An approach could be to avoid the executor in favour of creating the thread ourselves, and catching all Throwables to make sure we handle them. We leave fatal errors such as OOME to pop up in other parts of the code.
Attachments
Issue Links
- split from
-
OAK-11069 AzureRepositoryLock: Thread should not die silently on exceptions other than StorageException
- Closed
- links to