richardstartin edited a comment on issue #7476: URL: https://github.com/apache/pinot/issues/7476#issuecomment-926717352
Thanks for the information. This is quite a complex issue. What that stack trace means is that some call to `Unsafe.get` or `Unsafe.put` led to a SIGBUS error, which was caught asynchronously, meaning the frame in the error may or may not be directly related to the underlying cause. The underlying cause will be related to memory mapping, possibly an OOM writing to `tmpfs`. There's a good explanation to refer to [here](https://stackoverflow.com/questions/45536049/java-lang-internalerror-a-fault-occurred-in-a-recent-unsafe-memory-access-opera). We need to reproduce this, but if it happens again, you can figure out how much `tmpfs` is used by running `df -h` and looking for the row starting with `tmpfs`. If it happens again and `Use%` is high , then `tmpfs` OOM is likely to be the cause. If the risk of this happening again is too high, I recommend using more RAM until we have reproduced the issue and improved memory efficiency. To help reproduce the issue, it would be good to get a redacted copy of your table config so we can see which indexes etc. you are using. This will narrow the search for the culprit. Also, knowing `-Xmx` and `-XX:MaxDirectMemorySize` values would be useful. -- 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: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
