I remanufactured a small amount of data, tested and modified the MapReduce
configuration, but the same error still occurred。
Fortunately, I found some additional error logs on the yarn cluster. This
error log is the same as that mentioned in issue 3970. After modifying
mapreduce.tar.gz to delete kafka-clients-0.8.2.1.jar, more conflicts will
occur. The error logs are as follows
------------------------------------------------------------------------------------
2019-10-24 11:07:41,977 FATAL [Thread-13]
org.apache.hadoop.yarn.YarnUncaughtExceptionHandler: Thread
Thread[Thread-13,5,main] threw an Error. Shutting down now...
java.lang.NoSuchMethodError:
com.google.common.hash.Hasher.putString(Ljava/lang/CharSequence;)Lcom/google/common/hash/Hasher;
at
org.apache.kylin.engine.mr.steps.FactDistinctColumnsMapper$CuboidStatCalculator.putRowKeyToHLLNew(FactDistinctColumnsMapper.java:385)
at
org.apache.kylin.engine.mr.steps.FactDistinctColumnsMapper$CuboidStatCalculator.run(FactDistinctColumnsMapper.java:419)
at java.lang.Thread.run(Thread.java:748)
------------------------------------------------------------------------------------
This issue 3970 doesn't mention how to solve the new conflicts brought by
it. Can you tell me how to solve the bug brought by guava?
