Hi, In addition to the above suggestion, I think you can also check the broker logs clearly. You should see the reason. For example, here's what I saw in my logs:
[2021-02-26 16:38:45,777] INFO [GroupCoordinator 0]: Member consumer-groupId-1-7e24136f-8d7e-4303-a150-32dae06c18c0 in group groupId has failed, removing it from the group (kafka.coordinator.group.GroupCoordinator) [2021-02-26 16:38:45,779] INFO [GroupCoordinator 0]: Preparing to rebalance group groupId in state PreparingRebalance with old generation 2 (__consumer_offsets-18) (*reason: removing member consumer-groupId-1-7e24136f-8d7e-4303-a150-32dae06c18c0 on heartbeat expiration*) (kafka.coordinator.group.GroupCoordinator) So, you should know that it's because heartbeat expired in my case. Thanks. Luke On Mon, Jul 5, 2021 at 9:56 AM Shilin Wu <s...@confluent.io.invalid> wrote: > There might be many causes, but from broker's point of view: This consumer > is dead. > > Consumers are considered dead when one of the following happended: > > 1. didn't send out heartbeat for session.timeout.ms(defaults to 10 > seconds) period of time. (Consumers sent heart beat ever > heartbeat.interval.ms). Did you configure this reasonably? Note that > heartbeat runs in a background thread, so your app logic should not affect > it. > > 2. didn't poll for max.poll.interval.ms (defaults to 5 minutes) period of > time. Sometimes although the consumer is sending heartbeat (by the bg > thread), the main thread may be dead. Is your application polled too much > that couldn't be processed in 5 minutes so next pool is delayed by more > than 5 minutes? > > > > [image: Confluent] <https://www.confluent.io> > Wu Shilin > Solution Architect > +6581007012 > Follow us: [image: Blog] > < > https://www.confluent.io/blog?utm_source=footer&utm_medium=email&utm_campaign=ch.email-signature_type.community_content.blog > >[image: > Twitter] <https://twitter.com/ConfluentInc>[image: LinkedIn] > <https://www.linkedin.com/company/confluent/>[image: Slack] > <https://slackpass.io/confluentcommunity>[image: YouTube] > <https://youtube.com/confluent> > [image: Kafka Summit] <https://www.kafka-summit.org/> > > > On Sun, Jul 4, 2021 at 6:38 PM sunil chaudhari < > sunilmchaudhar...@gmail.com> > wrote: > > > Hi, > > There is something called as heartbet consumer thread. > > This threads running on consumer keeps sending heartbeats at regular > > interval as per the setting heartbeat.interval.ms. It keeps on telling > > broker that I am very much alive. > > > > > > > https://docs.confluent.io/platform/current/installation/configuration/consumer-configs.html#consumerconfigs_heartbeat.interval.ms > > > > There is another setting session.timeout.ms. If the time duration > between > > 2 > > heartbeats exceeds the time limit set in this setting then broker removes > > the client(consumer) from the group. > > > > Please read above settings in documentation. > > > > I hope you can co-relate this with your problem. > > > > Regards, > > Sunil. > > > > > > > > > > On Sun, 4 Jul 2021 at 1:36 PM, Kafka Life <lifekafka...@gmail.com> > wrote: > > > > > Dear kafka Experts > > > > > > Could one of you please help to explain what this below log in broker > > > instance mean..and what scenarios it would occur when there is no > change > > > done . > > > > > > INFO [GroupCoordinator 9610]: Member > > > > > > > > > webhooks-retry-app-840d3107-833f-4908-90bc-ea8c394c07c3-StreamThread-2-consumer-f87c3b85-5aa1-40f5-a42f-58927421b89e > > > in group webhooks-retry-app has failed, removing it from the group > > > (kafka.coordinator.group.GroupCoordinator) > > > > > > > > > INFO [GroupCoordinator 9611]: Member > > > cm.consumer.9-d65d39d3-703f-408b-bf4b-fbf087321d8c in group > > > cm_group_apac_sy_cu_01 has failed, removing it from the group > > > (kafka.coordinator.group.GroupCoordinator) > > > > > > > > > Please help to explain . > > > > > > thanks > > > > > >
