sodonnel opened a new pull request, #4453:
URL: https://github.com/apache/ozone/pull/4453

   ## What changes were proposed in this pull request?
   
   Right now, "hdds.scm.replication.command.deadline.factor" is a fraction. For 
long durations such as 60 minutes, the difference between SCM deadline and 
Datanode deadline will be 60 - 60 * 0.9, which is 6 minutes. This is a 
significant difference, so perhaps this configuration should be a duration 
instead, like 30 seconds.
   
   Currently the APIs provided by RM expose the DN deadline as a parameter. We 
could remove this and just let the RM decide a deadline for commands in the DN.
   
   This PR removes the parameter hdds.scm.replication.command.deadline.factor 
and also passing the datanodeDeadline via the various methods. Instead we now 
have hdds.scm.replication.event.timeout.datanode.offset, defaulting to 30s, 
which is the amount of time we subtract from event.timeout (SCM deadline) to 
get the datanode deadline.
   
   ## What is the link to the Apache JIRA
   
   https://issues.apache.org/jira/browse/HDDS-8230
   
   ## How was this patch tested?
   
   Existing test cover this.


-- 
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