[ 
https://issues.apache.org/jira/browse/RATIS-1771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17678137#comment-17678137
 ] 

Kaijie Chen commented on RATIS-1771:
------------------------------------

{quote}Why change the behavior? Since we already have timeout, it may not be 
useful to support abortion. Also, supporting abortion is easy to lead to race 
conditions. We should keep our code simple.{quote}

I'm planning to change yield leadership to use TransferLeadership in RATIS-1770.
So sometimes TransferLeadership from user may be blocked by the yield 
leadership.
I'm OK to keep it as is.

> Abort pending request on new TransferLeadership
> -----------------------------------------------
>
>                 Key: RATIS-1771
>                 URL: https://issues.apache.org/jira/browse/RATIS-1771
>             Project: Ratis
>          Issue Type: Sub-task
>            Reporter: Kaijie Chen
>            Priority: Major
>
> Currently the TransferLeadership will fail if there is a pending request with 
> different transferee.
> This Jira is planning to change this behavior. The pending request will be 
> aborted if the new request specifies a different transferee.
> It is noteworthy that the client should NOT retry the aborted request.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to