[
https://issues.apache.org/jira/browse/NIFI-1330?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15069886#comment-15069886
]
Oleg Zhurakousky commented on NIFI-1330:
----------------------------------------
Just to clarify and after discussing it with [~joewitt] "_timeouts should
continually be updated based on activity_". In this case we have 3 activities
_send -> confirm -> complete_. This means that timeout should be updated after
each successful activity (hey, we're talking). But it doesn't appear to be at
the moment. Basically the way current timeout is applied is all 3 must happen
within a single timeout.
> Site-to-Site timeout can occur in the middle of transaction
> -----------------------------------------------------------
>
> Key: NIFI-1330
> URL: https://issues.apache.org/jira/browse/NIFI-1330
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 0.4.1
> Reporter: Oleg Zhurakousky
>
> In the current state the timeout can occur in the middle of transaction. For
> example, timeout can occur between successful completion of _send(..)_ and
> _confirm(..)_. At this point confirm fails and flow file dropped on the
> receiving end even though it was successfully transferred. Can be a problem
> when sending large data.
> Also, if both _send(..)_ and _confirm(..)_ succeed the timeout may occur
> before or during the _complete(..)_ call which is more of a nuisance since
> other then ERROR log message (2015-12-23 09:12:08,253 ERROR [Site-to-Site
> Worker Thread-0] o.a.nifi.remote.SocketRemoteSiteListener Unable to
> communicate with remote instance . . .) everything succeeds.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
