Github user markap14 commented on the issue:
https://github.com/apache/nifi/pull/930
@joewitt I've not reviewed the entire PR here, but did review the commit
for the changes to class loading, and I believe these changes to be correct. I
made one in-line comment but other than that I
Github user bbende commented on the issue:
https://github.com/apache/nifi/pull/930
Using the latest updates I was able to verify the 0.10 processors against
the 0.10 broker using PLAINTEXT, SSL, and SASL, and also tested the 0.9.0
processors against an 0.10 broker.
I'm a +1
Github user joewitt commented on the issue:
https://github.com/apache/nifi/pull/930
Problem identified. Filed as NIFI-2660
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
Github user joewitt commented on the issue:
https://github.com/apache/nifi/pull/930
Update on weird behavior with subclassing. Turns out I can run either 0.9
or 0.10. Whichever runs first wins. There is clearly something going on with
classloading. Looking into Kafka source code
Github user bbende commented on the issue:
https://github.com/apache/nifi/pull/930
Quick update... I was finally able to get a Kerberos environment setup and
did a successful test with SASL, so that covers the 0.9.0 client against the
0.9.0 broker, with PLAINTEXT, SSL, and SASL.
Github user bbende commented on the issue:
https://github.com/apache/nifi/pull/930
I've been reviewing this... so far I have tested ConsumeKafka and
PublishKafka (0.9.0) against an Apache Kafka 0.9.0 broker, with PLAINTEXT and
also with SSL, and both appear to be working well. Trying
