[KYLIN-1440] Kylin streaming stop refreshed when one replica of partition goes down

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: v1.4.0
• Component/s: NRT Streaming
• Assignee: Zhong Yanghong
• Reporter: qianqiaoneng
• Created: 23/Feb/16 17:52
• Updated: 09/Aug/17 03:59
• Resolved: 23/Feb/16 23:22

Description

java.lang.NullPointerException: unable to find leadBroker with config:org.apache.kylin.streaming.KafkaClusterConfig@3c1 partitionId:0
at com.google.common.base.Preconditions.checkNotNull(Preconditions.java:208)
at org.apache.kylin.streaming.StreamingUtil.getFirstAndLastOffset(StreamingUtil.java:99)
at org.apache.kylin.streaming.StreamingUtil.findClosestOffsetWithDataTimestamp(StreamingUtil.java:88)
at org.apache.kylin.job.streaming.StreamingBootstrap$4.run(StreamingBootstrap.java:287)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)

[root@kylin-dev7-eaz-704018 kafka_2.10-0.8.2.0]# bin/kafka-topics.sh --describe --zookeeper
nous.seo_session
Topic:nous.seo_session PartitionCount:3 ReplicationFactor:3Configs:retention.ms=604800000
Topic: nous.seo_session Partition: 0 Leader: 8Replicas: 1,8,9 Isr: 8,9
Topic: nous.seo_session Partition: 1 Leader: 2Replicas: 2,9,10 Isr: 2,10,9
Topic: nous.seo_session Partition: 2 Leader: 4Replicas: 4,10,11 Isr: 4,10,11

It should be a normal situation and Kafka can still work normally. However Something in Kylin streaming went wrong when dealing with such cases.