[
https://issues.apache.org/jira/browse/NIFI-14708?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18005093#comment-18005093
]
Cemre Mengu commented on NIFI-14708:
------------------------------------
We found out that our NiFi instance was starving on threads. Increasing the
"Maximum Timer Driven Thread Count" in settings worked perfectly.
We never increased this value in 1.x though it could handle everything with
just 10 default threads. One thing we noticed is that the new ConsumeKafka
processors occupy 1 thread by default. This was not the case for 1.x. When idle
(i.e. no data coming from topics) ConsumeKafka processors did not use any
threads.
Was there a change in behavior with ConsumeKafka ?
> Run Once Is Very Slow
> ---------------------
>
> Key: NIFI-14708
> URL: https://issues.apache.org/jira/browse/NIFI-14708
> Project: Apache NiFi
> Issue Type: Bug
> Components: Core Framework, Core UI
> Affects Versions: 2.4.0
> Reporter: Cemre Mengu
> Priority: Minor
> Attachments: Recording 2025-07-03 143959.mp4
>
>
> RunOnce generally on any processor takes about 20 seconds. I have attached a
> video showing me doing it from GenerateFlowFile processor.
> This was not the case for 1.x versions. I am not sure if it is related but
> starting & stopping processors also takes a lot of time now (around 5-6
> seconds).
> Any known issues or workarounds about this? Although I consider it a minor
> issue since it doesn't prevent us from using the system, it is really
> annoying when dealing with a lot of processors.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
