[
https://issues.apache.org/jira/browse/NIFI-2729?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15694653#comment-15694653
]
ASF GitHub Bot commented on NIFI-2729:
--------------------------------------
Github user ijokarumawak commented on the issue:
https://github.com/apache/nifi/pull/1270
Since it happens not every time, it can't guarantee this change eliminates
the possibility of timeout failure, but I was able to reproduce the same
behavior using JUnit rule to run the suspected test 500 times with 10 threads.
Here is the [sample
code](https://github.com/ijokarumawak/nifi/commit/ba5be6f3c89deda8197d59d789feb78cf29f277b#diff-d14afc9dcbb637a81d33f6303eab3909R35).
With 500 loops and 10 threads, I got following reproduce rate:
1. without any change: 10 failures out of 20 attempts: 50%
2. adding catch 504 (without fix 3): 0 failure out of 10 attempts: 0%
3. reducing the number of Jetty and LittleProxy threads (without fix 2): 0
failure out of 20 attempts: 0%
I hope the above fix.3 mitigate the timeout failure. Even if it does, fix.2
will catch the exception and let test pass through.
I would like to keep this as Unit test rather than make it as Integration
Test. If the change and above test result seem reasonable, please merge this
and let it run for a while to see if it actually works as expected.
Thanks!
> testSendSuccessWithProxy frequently timing out in travis-ci
> -----------------------------------------------------------
>
> Key: NIFI-2729
> URL: https://issues.apache.org/jira/browse/NIFI-2729
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 1.1.0
> Reporter: Andre
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
