Github user jvwing commented on the issue:
https://github.com/apache/nifi/pull/239
Thanks, @mans2singh, I will review the changes. I don't think the AppVeyor
build is reliable, there don't appear to be any recent successful builds.
For GetKinesisStream exceptions, I think your approach of logging without
creating a FlowFile is correct. I'm not sure we could count on there being
anything useful to put in a FlowFile other than the error message, such that
the failed FlowFile can be meaningfully processed other than logging. I did
not find any other Get* processor with a failure route.
Have you experienced errors at that stage of Kinesis client processing? I
don't have enough experience with Kinesis Streams to know what bad things are
going to happen there, so I don't have clear failure scenarios to think
through. By checkpointing to the last successfully processed record, I believe
the 'bad' record will be retried by the KCL. This seems reasonable, but if
NiFi continues to fail on the same record, won't we end up in an infinite retry
loop?
---
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
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
