kevinrr888 commented on PR #5735:
URL: https://github.com/apache/accumulo/pull/5735#issuecomment-3058316888
> How did you find these changes needed to be made? Were you seeing failures
with this test before this change?
Figured out the first problem since I saw a prop being changed that was
recently set as fixed, so I knew it wouldn't be picked up. Was easy for me to
identify since I was the one who made the fixed property changes.
The second problem was a lot harder to identify. Essentially, restarting the
TServers would result in things hanging. The `online` operation after
restarting never completed. I found other places where we restart the TServers,
but it was very hard to identify any difference which would cause this test to
hang but other tests to work fine. Eventually narrowed it down to something
done in one test which wasn't done in this test:
From CleanWALIT:
`// use raw local file system so walogs sync and flush will work`
`hadoopCoreSite.set("fs.file.impl", RawLocalFileSystem.class.getName());`
I saw WAL related problems in the logs, so thought adding this to the config
for TotalQueuedIT might work, and it did.
Then, noticed `cfg.useMiniDFS();` wasn't doing anything. As this is also
related to DFS, replaced `hadoopCoreSite.set("fs.file.impl",
RawLocalFileSystem.class.getName());` with `cfg.useMiniDFS(true);` and
everything was working.
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
