Support Questions

Find answers, ask questions, and share your expertise

Falcon Mirror issue

avatar
Expert Contributor

I am facing checksum mismatch while copying data between two cluster.

Caused by: java.io.IOException: Check-sum mismatch. Also is there a way to ignore checksum through falcon UI? . Thanks in advance.

Check-sum mismatch between hdfs://test1:8020/raw/ech/incr/row_bn... and hdfs://test2:8020/raw/ech/.distcp.tmp.attempt_1471351807373_0.... Source and target differ in block-size.

1 ACCEPTED SOLUTION

avatar
Rising Star

@nyadav

It appears that block size is different in your two clusters . You can set flags -preserveBlockSize or -skipChecksum as below .

1. Suspend all Falcon jobs . 2. Modify template for falcon mirroring at /usr/hdp/current/falcon-server/data-mirroring/workflows/hdfs-replication-workflow.xml .Add the following argument at the end in argument list to preserve the block size . <arg>-preserveBlockSize</arg> <arg>true</arg> 3. Restart Falcon through Ambari. 4. Resubmit the job and verify if the HDFS mirror job is working fine now.

This will set this property for all mirror jobs.

View solution in original post

2 REPLIES 2

avatar
Rising Star

@nyadav

It appears that block size is different in your two clusters . You can set flags -preserveBlockSize or -skipChecksum as below .

1. Suspend all Falcon jobs . 2. Modify template for falcon mirroring at /usr/hdp/current/falcon-server/data-mirroring/workflows/hdfs-replication-workflow.xml .Add the following argument at the end in argument list to preserve the block size . <arg>-preserveBlockSize</arg> <arg>true</arg> 3. Restart Falcon through Ambari. 4. Resubmit the job and verify if the HDFS mirror job is working fine now.

This will set this property for all mirror jobs.

avatar
Expert Contributor

Thanks @prsingh, able to run the job after the change