Hi Josef,

There is a PR made by Mark takes care of this behaviour:  
https://github.com/apache/nifi/pull/7095

I hope this will fit to your needs! If there is something is going unexpected, 
please reply!

Regards,
Bence

> On 2023. Mar 28., at 9:14, Simon Bence <[email protected]> wrote:
> 
> Hi Josef,
> 
> I will take a look as soon as I will have the chance and find you back.
> 
> Regards,
> Bence
> 
>> On 2023. Mar 28., at 8:44, [email protected] wrote:
>> 
>> Hi Bence
>>  
>> We still facing problems with the current NiFi v1.20.0 release and nested 
>> process groups. Would you mind having a look at the issue?
>>  
>> https://issues.apache.org/jira/browse/NIFI-11251
>>  
>> It seems that NiFi is not correctly stopping/starting the processors within 
>> the (nested?) PGs. As workaround we always stop the processors, do the NiFi 
>> registry update and start the processors again. Very annoying as we have 
>> HTTP listeners which stop answering in that period.
>>  
>> Cheers Josef
>>  
>>  
>> From: Zahner Josef, GSB-LR-TRW-LI <[email protected]>
>> Date: Tuesday, 13 December 2022 at 13:53
>> To: [email protected] <[email protected]>
>> Subject: Re: NiFi 1.19.1 Registry Nested Process Groups - Show Local Changes 
>> issue
>> 
>> Hi Bence
>>  
>> I mentioned yesterday another issue which we thought would be solved after 
>> the upgrade steps. Sadly it happened again and after an investigation on our 
>> side it seems to be fully reproducible, I’ve opened a Jira Bugticket 
>> https://issues.apache.org/jira/browse/NIFI-10973 with more details. It’s as 
>> well related to NiFi Registry nested flows…
>>  
>> Thanks for taking care of it :-).
>>  
>> Cheers Josef
>>  
>>  
>> From: Simon Bence <[email protected]>
>> Reply to: "[email protected]" <[email protected]>
>> Date: Monday, 12 December 2022 at 16:40
>> To: "[email protected]" <[email protected]>
>> Subject: Re: NiFi 1.19.1 Registry Nested Process Groups - Show Local Changes 
>> issue
>>  
>> Hi Josef, 
>>  
>> Thanks for the quick update! Meanwhile I tried to reproduce this situation 
>> but based on the known facts I was not able (Note: without any version 
>> change). I think it is connected to the version change indeed, but if you 
>> face with this ever again, please let me know!
>>  
>> Regards,
>> Bence
>> 
>> 
>> 
>> On 2022. Dec 12., at 16:33, <[email protected] 
>> <mailto:[email protected]>> <[email protected] 
>> <mailto:[email protected]>> wrote:
>>  
>> Ok quick update on this. We investigated the changes on the *.snapshot files 
>> on NiFi Registry. It seems that it’s mainly caused due to the fact that the 
>> NiFi Version changes from 1.18.0 to 1.19.1 and some other facts caused due 
>> to the upgrade. After commiting all the child PGs once the “Show Local 
>> Changes” seems to behave as expected.
>> We have seen another issue caused only on the first commit of the child PGs 
>> on v1.19.1, however I’ll not mention them as I’m guessing that this is 
>> highly related to the last commits with the older NiFi version.
>>  
>> Cheers Josef
>>  
>>  
>> From: "Zahner Josef, GSB-LR-TRW-LI" <[email protected] 
>> <mailto:[email protected]>>
>> Date: Monday, 12 December 2022 at 14:44
>> To: "[email protected] <mailto:[email protected]>" 
>> <[email protected] <mailto:[email protected]>>
>> Subject: NiFi 1.19.1 Registry Nested Process Groups - Show Local Changes 
>> issue
>>  
>> Hi guys
>>  
>> Sorry that I’ve to bother you again. We just upgraded to NiFi 1.19.1 due to 
>> the major issues with the NiFi Registry Client on NiFi 1.18.0.
>>  
>> The following issue still persists in the new version. Let’s say everything 
>> is up-to-date (parent and child PGs) from NiFi Registry perspective, now I 
>> do change eg. a label directly in the parent PG. I do see on that parent PG, 
>> that most likely every connection/service/processor has changed, even though 
>> I’ve changed only one single label size, please check the screenshot for 
>> NiFi Registry Client under “Show Local Changes”.  
>>  
>> <image001.png>
>>  
>> Is this a cosmetical issue or do we have to worry about that (we had massive 
>> NiFi registry issue on NiFi 1.18.0) ? Because in theory it makes no sense to 
>> that NiFi tells that everything has changed if only one label changes.
>>  
>> Thanks in advance
>> Josef
> 

Reply via email to