[jira] [Commented] (NIFI-4825) Processor is not shutting down normally

["Koji Kawamura (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NIFI-4825?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16492304#comment-16492304 ] 

Koji Kawamura commented on NIFI-4825:
-------------------------------------

Hi [~ktj1312], it's been a while since you reported this JIRA, but do you still have the issue? I think the reason why you could not clear the queued FlowFiles was that ReplaceText was still running, or hanging at an unexpected place. The ReplaceText processor in your screenshot shows 2 at the right top corner, that means there were 2 threads running. One is processing, and holding the 100 input FlowFiles, and the other is trying to stop the processor.

I recommend to take a thread dump by executing following command:
{code}
${NIFI_HOME}/bin/nifi.sh dump
{code}
The thread dump will be written to the nifi-bootstrap.log file. It will be useful to investigate the issue further. Please take the dump when you encountered this situation again.

> Processor is not shutting down normally
> ---------------------------------------
>
>                 Key: NIFI-4825
>                 URL: https://issues.apache.org/jira/browse/NIFI-4825
>             Project: Apache NiFi
>          Issue Type: Bug
>          Components: Core Framework, Core UI
>    Affects Versions: 1.4.0, 1.5.0
>         Environment: Clustering with 3 Node and Seperated 3 zookeeper
>            Reporter: kim teak jin
>            Priority: Critical
>         Attachments: error_template (1).xml, image-2018-01-29-13-55-07-904.png, image-2018-01-29-13-55-56-745.png
>
>
> Hi dear Nifi users and Developers.
> If Json type time, created by default, is inserted, the problem did not recognize the type as ISODate in MongoDB and is using ReplaceText processor to resolve this.
> Now the input to MongoDB is working well, but there is another problem.
> So was v1.4, and I currently have updated to v1.5.
> The same phenomenon is occurring later but when the processor is not operational anymore, the input queue becomes full approximately one day after it starts on the ReplaceText processor.
> So I stopped the processor and tried to erase the queue and the processor in front of it, but that didn't work.
> The following picture is a message that happens when I try to clear the queue.
> !image-2018-01-29-13-55-56-745.png!
> Now I think this is very important to me and if it continues to happen, I think I will have to take a different solution.
> This is because it occurs continuously even if the queue can be deleted normally after stopping and re-enabled successfully.
> So, temporarily, I copy the processor and know that it is not the right solution though I am doing first aid such as photo.
> !image-2018-01-29-13-55-07-904.png!
> Currently, the left processor shows up as an inactive halt, but buttons such as setting window and starting to stop do not appear.
>  I am attaching my template here, so please do not hesitate to deal with the issue quickly.
>  
> We look forward to a speedy handling of the issues involved.
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)