greatsharp opened a new pull request, #362:
URL: https://github.com/apache/kvrocks-controller/pull/362

   This enhancement can improve the failover efficiency when a lots of master 
node probe failed concurrently.
   
   Since there is only one can UpdateCluster successfully each time, we should 
reset the failure count after UpdateCluster successfully, and can do failover 
later again, there is no need to wait more than 15 
seconds(ping_interval_seconds*max_ping_count).
   
   For example, if 3 master node doing failover concurrently, it will cost more 
than 45 seconds before, but now it is 18 seconds (12s + 3s + 3s).
   
   BTW, we should enlarge the minIdleConns to each node, this can save time in 
acquiring new connections when doing probe and sync cluster topology to all 
nodes, 10 is enough.


-- 
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