[
https://issues.apache.org/jira/browse/OAK-11069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17882337#comment-17882337
]
Axel Hanikel commented on OAK-11069:
------------------------------------
To avoid any misunderstanding due to the title of this issue: The fix makes
sure the thread does not die _silently_, but it still dies on anything other
than a StorageException. It is leaving a message in the log, however.
A slight improvement to the situation is that _any_ kind of StorageException
now causes the operation to be retried, not only StorageExceptions of a certain
type.
The idea is here not to keep retrying if we get anything other than a
StorageException because we don't expect any other exception here, and retrying
could actually make things worse than just stopping further writes. Once we
know more about what's happening, we can come up with a more precise fix. If it
turns out we only get StorageExceptions as expected, we're good already.
> AzureRepositoryLock: Thread should not die silently on exceptions other than
> StorageException
> ---------------------------------------------------------------------------------------------
>
> Key: OAK-11069
> URL: https://issues.apache.org/jira/browse/OAK-11069
> Project: Jackrabbit Oak
> Issue Type: Bug
> Components: segment-azure
> Affects Versions: 1.68.0
> Reporter: Axel Hanikel
> Assignee: Miroslav Smiljanic
> Priority: Major
> Fix For: 1.70.0
>
>
> AzureRepositoryLock should handle other types of exceptions in refreshLease.
> https://github.com/apache/jackrabbit-oak/blob/53b51f29e40cd0f0f9e30999751b63de64038737/oak-segment-azure/src/main/java/org/apache/jackrabbit/oak/segment/azure/AzureRepositoryLock.java#L140-L140
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
