[
https://issues.apache.org/jira/browse/TEZ-1770?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14208384#comment-14208384
]
Rajesh Balamohan commented on TEZ-1770:
---------------------------------------
Not completely sure, if we need to maintain extraRetryForConnectFailure with
the additional logic. We end up retrying with sleep, when we get
ConnectException (i.e, when NM is getting restarted).
{code}
long connectStartTime = System.currentTimeMillis();
{code}
connectStartTime can be directly initied to System.currentTimeMillis.
> Handle ConnectExceptions correctly when establishing connections to an NM
> which may be down
> -------------------------------------------------------------------------------------------
>
> Key: TEZ-1770
> URL: https://issues.apache.org/jira/browse/TEZ-1770
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Siddharth Seth
> Assignee: Siddharth Seth
> Priority: Critical
> Attachments: TEZ-1770.1.txt
>
>
> Thanks [~sidharta-s] for reporting this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
