You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pulsar.apache.org by GitBox <gi...@apache.org> on 2018/08/30 17:33:09 UTC

[GitHub] merlimat opened a new pull request #2487: Ensure standalone service comes back quickly after ungraceful restarts

merlimat opened a new pull request #2487: Ensure standalone service comes back quickly after ungraceful restarts
URL: https://github.com/apache/incubator-pulsar/pull/2487
 
 
   ### Motivation
   
   After an ungraceful restart of Pulsar standalone, it might take few iterations of restart to get the service back up. 
   
   This is due to the previous ZK sessions not being "expired". 
   
   The main problem is that by having broker & zk combined the 2 behaviors clash:
     * Broker try to register z-node and finds it there from other session (and restart) — which is correct for a regular broker
     * ZK server doesn’t have the time to actually expire the session because broker exits immediately
   
   
   ### Modifications
   
   In standalone mode, clear the registration z-nodes from the previous processes.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services