Cassandra Troubleshooting

CASSANDRA0001: Discover Operation Fails

Issue

The discover operation fails and displays the following message in the CommCell Console:

Failed to find Cluster nodes
    

Resolution

Check the messages in the following log files that are located on the Gateway Node (you can get Gateway Node value on the Instance Properties dialog box):

• cvPriam.log

• cvd.log

Issue

The discover operation fails because of a timeout.

Resolution

By default, the Commvault software waits 10 seconds for the Priam server to start. Set the time to a longer interval. For more information, see Setting the Time to Wait for the Priam Server to Start.

Issue

The discover operation fails because of a setup error.

Resolution

Make sure that you have the correct priam.jar file (under the /opt/commvault/Base directory) linked.

On the command line, type the following command:

python2.7 cvCassandra_install.py <commvault_Base_Folder> 3 <Commvault_log_files_directory>
    

If the problem still persists after you run the command, then contact your Account Representative and have the following files:

• The CommServe computer log files

• The Cassandra client log files

• The Cassandra system.log file and the debug.log file.

CASSANDRA0002: Backup Operation Are Slow

Issue

The backup performance is slow.

Resolution

Increase the number of data readers (Subclient dialog box, Storage Device tab, Number of Data Readers box) to a value that causes the streams to be equally distributed. For example, set the value to a number that is a multiple of the number of nodes in the cluster.

CASSANDRA0003: A Restore Job Completes with Errors

Issue

A restore job completes with errors because because the software cannot refresh the keyspace at the end of the restore. This situation can happen because of one of the following issues:

• There is a keyspace refresh error

• There is a sstableloader error.

Resolution

• If you perform an out-of-place restore, then make sure that the source and destination sstable are compatible.

• The interface that Cassandra uses must be in the /etc/host file on the Cassandra nodes.

• Nodes that are in different datacenters must be connected. If you have cloud deployments, to prevent the software from streaming restored data on one datacenter to other datacenters, set the additional setting nCassIgnoreNodesInOtherDC to 1.

For instructions about adding an additional settings from the CommCell Console, see Adding or Modifying Additional Settings from the CommCell Console.

• For additional information, see the following files:

  • The CassandraCtrl.log file and the cvPriam.log file - These files are in the /var/log/commvault/Log_Files directory.

  • The Cassandra system.log file and the debug.log file