hangout 版本 1.3.1 kafka 0.10
kafka 故障后，hangout 报错
I0418 01:33:27.948472 1 bulk_http.go:144] bulk 15457 docs with execution_id 45326
I0418 01:33:27.948537 1 bulk_http.go:154] try to bulk with host (http://10.191.68.23:9200/_bulk)
I0418 01:33:28.980086 1 bulk_http.go:161] bulk done with execution_id 45326 1.032 15457 14977.712
I0418 01:37:27.950759 1 brokers.go:131] got 25 brokers
E0418 01:37:29.849768 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:43008->10.191.54.224:9092: i/o timeout
E0418 01:37:29.849808 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:37:29.949877 1 broker.go:177] broker se.net:9092 is dead, (re)open it after sleep 200ms
E0418 01:38:00.651407 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:48402->10.191.54.224:9092: i/o timeout
E0418 01:38:00.651467 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:38:00.751551 1 broker.go:177] broker nease.net:9092 is dead, (re)open it after sleep 200ms
E0418 01:38:31.452661 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:48446->10.191.54.224:9092: i/o timeout
E0418 01:38:31.452716 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:38:31.552793 1 broker.go:177] broker .net:9092 is dead, (re)open it after sleep 200ms
E0418 01:39:02.285946 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:48506->10.191.54.224:9092: i/o timeout
E0418 01:39:02.285993 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:39:02.386042 1 broker.go:177] broker ease.net:9092 is dead, (re)open it after sleep 200ms
E0418 01:39:33.138383 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:48588->10.191.54.224:9092: i/o timeout
E0418 01:39:33.138415 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:39:33.238536 1 broker.go:177] broker i.nease.net:9092 is dead, (re)open it after sleep 200ms
E0418 01:40:03.945554 1 simple_consumer.go:371] fetch error:read tcp 10.191.66.203:48686->10.191.54.224:9092: i/o timeout
E0418 01:40:03.945592 1 fetch_response.go:284] could not read enough bytes(4) to get fetchresponse length. read 0 bytes
I0418 01:40:04.045787 1 broker.go:177] broker .i.nease.net:9092 is dead, (re)open it after sleep 200ms
省略重复
E0418 04:37:00.279222 1 brokers.go:269] get metadata of [suez_g66_logtail] from canal-sssez.i.nease.net:9092 error: There is no leader for this topic-partition as we are in the middle of a leadership election.
E0418 04:37:00.323861 1 brokers.go:269] get metadata of [suez_g66_logtail] from canal-kafssssase.net:9092 error: There is no leader for this topic-partition as we are in the middle of a leadership election.
E0418 04:37:00.387194 1 brokers.go:269] get metadata of [suez_g66_logtail] from canal-sss9-sssease.net:9092 error: There is no leader for this topic-partition as we are in the middle of a leadership election.
省略重复
E0418 06:28:36.474490 1 brokers.go:269] get metadata of [suez_g66_logtail] from canal-kafka14-suez.i.nease.net:9092 error: There is no leader for this topic-partition as we are in the middle of a leadership election.
E0418 06:28:36.675459 1 brokers.go:269] get metadata of [suez_g66_logtail] from canal-kafka10-suez.i.nease.net:9092 error: There is no leader for this topic-partition as we are in the middle of a leadership election.
kafka 故障恢复后，hangout 一直不消费，刷这个，强制重启后开始消费
I0418 06:28:36.876990 1 group_consumer.go:235] sync logstash_g66_logtail_11 error: The coordinator is not aware of this member.
I0418 06:28:36.877011 1 group_consumer.go:169] try to join group logstash_g66_logtail_11
I0418 06:28:36.877333 1 group_consumer.go:190] join logstash_g66_logtail_11 error: The coordinator is not aware of this member.
省略重复
I0419 00:00:00.001397 1 group_consumer.go:190] join logstash_g66_logtail_11 error: The coordinator is not aware of this member.
I0419 00:00:00.001403 1 group_consumer.go:169] try to join group logstash_g66_logtail_11
补充下时间点，kafka 是18日凌晨1点30全挂了，早上8点恢复，但是这个不消费是 19日22点发现并重启的，数据分区没有rotate，但是从最新读了，而且后面补数据重新指定offset 是可以消费的。