SSTables not cleaned up after failed restore

If a restore fails and not all SSTables are copied, OpsCenter does not clean up the remaining tables.

If a restore that bypasses SSTableloader fails, some SSTables might be copied over, whereas other SSTables are not. In this scenario, OpsCenter does not clean up SSTables that were copied or revert any changes in the Cassandra data directory. This behavior could cause the keyspace or table to be unusable, and possibly result in duplicate, inconsistent, and even lost data.

Note: This scenario only applies to restores that do not use SSTableloader. If SSTableloader is used with the restore, the copied files are staged in a temporary directory that OpsCenter cleans up.

Retry the restore and view the Restore Report to track the progress and status of the restore. The Restore Report dialog can be closed at any time without impacting the restore process. Reopen the report by clicking on the In Progress restore in the Activity tab. If the restore completes successfully, all SSTables are copied over.