2

I'm running a DSE cluster in AWS: m2.4xlarge instances running Datastax Enterprise 4.6.1, with Cassandra 2.0.12.200 and Opscenter 5.1.0.

When we try to do a backup of a keyspace, we get this: 

Snapshot of keyspaces [XXXXXXX] on node XXX.XXX.XXX.XXX failed: javax.management.RuntimeMBeanException: java.lang.RuntimeException: Tried to hard link to file that does not exist /raid0/cassandra/data/XXXXXX/XXXXXX/XXXXXXXXXXXX-jb-1-Index.db

Any ideas?

Undo
  • 25,519
  • 37
  • 106
  • 129
sebamontini
  • 360
  • 2
  • 10

2 Answers2

2

This is likely the following known issue:

https://issues.apache.org/jira/browse/CASSANDRA-6433

The workaround for this is a rolling restart and it is fixed in c* 2.1. It seems to be caused when you drop a keyspace and re-create it again.

phact
  • 7,305
  • 23
  • 27
0

I had a similar issue with a dropped keyspace and compaction.

Run a "nodetool repair" and check the "system" and "opscenter" keyspaces for traces of the deleted keyspace. You might need to manually remove stale rows.

BeepBoop
  • 1,282
  • 10
  • 12