Hi Jonathan, Can you please file a JIRA?
Ismael On Sat, Jul 14, 2018 at 3:44 AM Jonathan Santilli < jonathansanti...@gmail.com> wrote: > Hello, hope you all are Ok, > > I would like to know if this is the expected behavior: > > ERROR Error while creating ephemeral at /brokers/ids/BROKER_ID, node > already exists and owner '*216186131422332301*' does not match current > session '*288330817911521280*' (kafka.zk.KafkaZkClient$CheckedEphemeral) > > INFO Result of znode creation at /brokers/ids/BROKER_ID is: NODEEXISTS > (kafka.zk.KafkaZkClient) > > ERROR Uncaught exception in scheduled task 'isr-expiration' > (kafka.utils.KafkaScheduler) > > > After that, the logs keep showing constantly the following: > > > INFO [Partition TOPIC_NAME-PARTITION-ID broker=BROKER_ID] Shrinking ISR > from 1,3,2 to 1 (kafka.cluster.Partition) > INFO [Partition TOPIC_NAME-PARTITION-ID broker=BROKER_ID Cached zkVersion > [0] not equal to that in zookeeper, skip updating ISR > (kafka.cluster.Partition) > INFO [Partition __consumer_offsets-PARTITION-ID broker=BROKER_ID] Shrinking > ISR from 1,2,3 to 1 (kafka.cluster.Partition) > INFO [Partition __consumer_offsets-PARTITION-ID broker=BROKER_ID] Cached > zkVersion [139] not equal to that in zookeeper, skip updating ISR > (kafka.cluster.Partition) > > > The only way to recover was restarting the Broker. > > I will really appreciate any clue about this error, > > > Cheers! > -- > Santilli Jonathan >
