empiredan commented on code in PR #1877:
URL: 
https://github.com/apache/incubator-pegasus/pull/1877#discussion_r1481269644


##########
src/server/config.ini:
##########
@@ -236,20 +236,20 @@ stateful = true
   delay_for_fd_timeout_on_start = false
   empty_write_disabled = false
 
-  prepare_timeout_ms_for_secondaries = 3000
-  prepare_timeout_ms_for_potential_secondaries = 5000
+  prepare_timeout_ms_for_secondaries = 1000
+  prepare_timeout_ms_for_potential_secondaries = 3000

Review Comment:
   Based on the experiences of production environment, sometimes both of them 
would be increased to 6000/10000. Thus I think 1000/3000 would be a litter 
small, which would make partitions unstable.
   ```suggestion
     prepare_timeout_ms_for_secondaries = 3000
     prepare_timeout_ms_for_potential_secondaries = 5000
   ```



##########
src/server/config.ini:
##########
@@ -206,7 +206,7 @@
   # partition guardian is used to keep partitions healthy. 
   # e.g. if max_replicas_in_group = 3, one primary and two secondaries must be 
kept for each partition 
   partition_guardian_type = partition_guardian
-  replica_assign_delay_ms_for_dropouts = 600000
+  replica_assign_delay_ms_for_dropouts = 300000

Review Comment:
   I think it's better to keep replica_assign_delay_ms_for_dropouts, 
staleness_for_commit, group_check_interval_ms, checkpoint_interval_seconds, 
fd_lease_seconds, fd_grace_seconds and log_private_reserve_max_time_seconds 
just as what they have been in `config.ini` ? All of them have been long 
running on production environments. I'm afraid that once they are changed, 
maybe there would be some problems for the production environment. We can 
change the default values in code to be consistent with what they are in 
`config.ini`.



-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to