Chris, I have seen that when the diskspace kafka relies on is full. We've seen a number of interesting exceptions recently in testing various configurations. But recommend checking that.
Also, what version of Kafka broker are you using? With Apache NiFi 0.5.x we moved to the kafka client 0.9. In doing that we messed up support for 0.8. So...with the upcoming release we will move back to the 0.8 client and thus it works great with Kafka 0.8 and 0.9 brokers albeit without the new SSL and Kerberos support they added in their 0.9 work. We have a JIRA item to go after that for our next feature bearing release. Thanks Joe On Wed, Mar 16, 2016 at 11:01 AM, McDermott, Chris Kevin (MSDU - STaTS/StorefrontRemote) <[email protected]> wrote: > I say strange because the timeout (63ms) is so very short. The communication > timeout I’ve set is 30 sec. Has anyone overseen this? > > 2016-03-16 14:41:38,227 ERROR [Timer-Driven Process Thread-8] > o.apache.nifi.processors.kafka.PutKafka > PutKafka[id=852c8d42-a2fa-3478-b06b-84ceb6\ > 6f8b0b] Failed to send > StandardFlowFileRecord[uuid=a0074162-0066-49e7-918b-cea1cfc5a955,claim=StandardContentClaim > [resourceClaim=StandardResour\ > ceClaim[id=1458079089737-67, container=default, section=67], offset=377796, > length=743],offset=0,name=2349680613178720,size=743] to Kafka; routi\ > ng to 'failure'; last failure reason reported was > org.apache.kafka.common.errors.TimeoutException: Failed to update metadata > after 63 ms.;: org.\ > apache.kafka.common.errors.TimeoutException: Failed to update metadata after > 63 ms.
