[GitHub] [accumulo] milleruntime commented on issue #1790: Flaky test: FateStarvationIT

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

milleruntime commented on issue #1790:
URL: https://github.com/apache/accumulo/issues/1790#issuecomment-1287367989

   Assuming that ZK couldn't be reached for whatever reason, it looks like the `top()` method in ZooStore has code to retry. But there is an additional try/catch within the retry that will bypass the retry loop and throw a RuntimeException. So if connection is lost at the right moment, it won't retry. Did you see any retries with this messsage: `        log.debug("zookeeper error reading " + txpath + ": " + ex, ex);
   `?
   
   It seems like we should just drop the try/catch around this call:
   https://github.com/apache/accumulo/blob/2dbdc085390794e7925740a81439c66dc3dfbb7c/core/src/main/java/org/apache/accumulo/core/fate/ZooStore.java#L305
   
   I don't know why we need nested try/catch statements. It doesn't make sense why we would want to retry when making a ZK call to `getData()` but not `getChildren()`.


-- 
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: notifications-unsubscribe@accumulo.apache.org

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