[GitHub] [ozone] guihecheng commented on pull request #2352: HDDS-5363. Datanode shutdown due to too many bad volumes in CI

[GitBox <gi...@apache.org>]

guihecheng commented on pull request #2352:
URL: https://github.com/apache/ozone/pull/2352#issuecomment-865063204


   Ah, nice catch! Thanks very much @adoroszlai , sorry, this is my careless mistake.
   In HDDS-5268, I tried to move the periodic volume checker into `StorageVolumeChecker`, and all volumeSets can register themself for periodic check, so I used the `synchronized` to protect register and check. But I didn't realize that the object lock was already used in the `ResultHandler` to protect the `healthyVolumes` and `failedVolumes` in another thread. So there is a dead lock.
   After your commit, the `healthyVolumes` and `failedVolumes` are all concurrent sets, so locks are separate.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org