zbentley opened a new issue, #15278:
URL: https://github.com/apache/pulsar/issues/15278

   **Describe the bug**
   Sometimes, when terminating a partitioned topic via the management API, 
termination blocks "forever" (until I time out the HTTP request). However, no 
matter when I time out that request, a subsequent request to terminate the 
topic always succeeds.
   
   This happens regardless of whether I'm connected to a broker or a proxy, and 
doesn't seem to coincide with any stacktraces in the broker logs, though I 
didn't check for "distant" log entries from the moment of termination.
   
   **To Reproduce**
   This does not reliably reproduce. I've seen it happen most frequently when 
attempting to terminate partitioned topics that were created very shortly 
before doing this (seconds or less beforehand).
   
   1. Create a partitioned topic.
   2. Immediately try to terminate it via the management API.
   3. Observe that sometimes termination times out with an HTTP timeout.
   4. When a timeout occurs, immediately re-attempt termination (send the exact 
same HTTP request).
   5. Observe that subsequent requests succeed.
   
   **Expected behavior**
   Termination requests shouldn't block forever, and should return 
success/failure when the broker has terminated a topic. 
   
   The fact that an immediate followup request succeeds indicates to me that 
the broker has successfully terminated the topic before that point, but the 
handling code for the first (timing out) request doesn't know about it.
   
   Broker 2.9.1, MacOS and Linux, standalone and SN platform brokers reproduced.
   


-- 
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]

Reply via email to