- Log in to the WSO2 Storage Server management console.
- Go to the Tools menu and click Cassandra Operations. The Cluster Nodes screen will open:
- Click View Operations associated with each cluster node in the cluster. The Node Operations screen opens:
It contains all the cluster-node-level, keyspace -level and column-family-level operations, such as backup, flush, scrub, shutting down RPC and Gossip servers etc.
The following topics explains each operation in detail:
General Cassandra Operations
Decommission: This operation removes the node from the cluster. If there is only one node in the cluster, it cannot be removed.
Drain: This operation flushes the entire node. This causes the node to stop accepting write requests until it is restarted.
Move: This operation changes node's currant token to a new one. This is important when load-balancing a cluster.
Perform GC (Garbage Collector): This operation activates the garbage collector on the node. It affects the overall latency of activities on the node.
Backup Operations
Take Snapshot: Enables the user to backup data in the node.
Clear Snapshot: Clears backups taken by the Take Snapshot operation and frees the memory that was utilized for them.
Enable Incremental Backup: Allows to fully backup the data in a node. Also, data will be continuously written to the backup as and when additional data is received by the node, even after the backup is taken.
Server Operations
Stop RPC: Allows to shutdown and restart the RPC server of the node.
Stop Gossip: Allows to shutdown and restart the Gossip server of the node.
Cache Operations
Invalidate Row Cache
Set Cache Capacity
Other Operations
Remove Token: Removes a token in the cluster or forcefully removes completion operations listed under this operation. User gets a live update of the currant token removal status in the management console.
Stop Compaction: Stops any listed, permitted compaction under this operation.
Rebuild: Allows to specify an available data center for re-building.
Set Stream Throughput: Set the stream throughput of the node.
Set Compaction Throughput: Sets the compaction throughput for the node.
Keyspace Operations
Click Keyspace Operations at the bottom of the Node Operations screen.
The Keyspace Operations screen opens with the list of operations associated with each keyspace as follows:
The operations are explained below.
Repair: Compares the replicas for each key responsible for the node and all replicas are updated to the newest available version.
Compact: Triggers major compaction in all column families of the keyspace and performs compaction of multiple SS tables into a single SS table.
Flush: This flushes the memtables (in memory) to SSTables (on disk), which also enables CommitLog segments to be deleted.
Cleanup: Prompts Cassandra to cleanup keys, which no longer belong to the node.
Scrub: This operation rebuilds sstables with correct bloom filters, with no data loss. The nodetool scrub creates a snapshot of data files before rebuilding.
UpgradeSSTables: While the Scrub operation rebuilds your sstables, it also discards data it deems broken and creates a snapshot of it, which has to be removed manually. To rebuild your sstables without going through this process, it is best to use UpgradeSSTables operation, which is particularly useful when upgrading your server or changing compression options.
Take Snapshot: Makes a backup of the the keyspace data.
Clear snapshot: Clears the backup of a keyspace.