[GitHub] [accumulo] ddanielr commented on issue #3157: Accumulo will fail to start if startup.tserver.avail.min.count is too large

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

ddanielr commented on issue #3157:
URL: https://github.com/apache/accumulo/issues/3157#issuecomment-1376052248

   This while loop handles the waiting logic for startup.
   https://github.com/apache/accumulo/blob/1.10/server/master/src/main/java/org/apache/accumulo/master/Master.java#L1516-L1530
   
   Looking at the comparison, it breaks out if `minTserverCount` is less than amount of current servers.
   https://github.com/apache/accumulo/blob/1.10/server/master/src/main/java/org/apache/accumulo/master/Master.java#L1514
   
   Or if there are no more retries available from `tserverRetry`.
   https://github.com/apache/accumulo/blob/1.10/server/master/src/main/java/org/apache/accumulo/master/Master.java#L1506-L1509
    
   I'm not very familiar with the Retry class, but it doesn't look like the `maxRetries` value is set as part object creation, just `maxWait`. However, the `infiniteRetries()` method is called which sets `maxRetries` to `-1` and I would think, always allows retries to be possible. 
   
   https://github.com/apache/accumulo/blob/18c4f1ab7167cbdf8123b6bbc5f57b728650ad21/fate/src/main/java/org/apache/accumulo/fate/util/Retry.java#L49
   
   If this is indeed the problem, it could be fixed by either changing the while loop to be duration based, or calculating the correct amount of retries for the time interval and setting the `MaxRetries` param of `tserverRetry`.


-- 
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