Apache Cassandra™ 2.0

tpstats

Provides usage statistics of thread pools.

Synopsis

nodetool <options> tpstats
options are:
  • ( -h | --host ) <host name> | <ip address>
  • ( -p | --port ) <port number>
  • ( -pw | --password ) <password>
  • ( -u | --username ) <user name>

Synopsis Legend

  • Angle brackets (< >) means not literal, a variable
  • Italics mean optional
  • The pipe (|) symbol means OR or AND/OR
  • Ellipsis (...) means repeatable
  • Orange ( and ) means not literal, indicates scope

Description

Run the nodetool tpstats command on the local node. The nodetool tpstats command provides statistics about the number of active, pending, and completed tasks for each stage of Cassandra operations by thread pool. A high number of pending tasks for any pool can indicate performance problems, as described in http://wiki.apache.org/cassandra/Operations#Monitoring .

This table describes the indicators:

nodetool tpstats output
Name of statistic Task Related information
ReadStage Local reads  
RequestResponse Handle responses from other nodes  
MutationStage Local writes A high number of pending write requests indicates a problem handling them. Tune hardware or Cassandra configuration.
ReadRepairStage A digest query and update of replicas of a key  
ReplicateOnWriteStage Counter writes, replications after a local write  
GossipStage Handle gossip rounds every second  
AntiEntropyStage Repair consistency Nodetool repair
MigrationStage Make schema changes  
MemoryMeter Actual object memory including JVM overhead  
MemtablePostFlusher Flush the commit log and other operations after flushing the memtable  
FlushWriter Flush the memtable to disk, the status of the sort and write-to-disk operations Tune hardware or Cassandra configuration.
MiscStage Miscellaneous operations  
PendingRangeCalculator Calculate pending ranges per bootstraps and departed nodes Developer notes
commitlog_archiver Save the commit log  
InternalResponseStage Respond to non-client initiated messages, including bootstrapping and schema checking  
HintedHandoff Send missed mutations to other nodes  

Example

Run the command every two seconds.

nodetool -h labcluster tpstats

Example output is:

Pool Name                    Active   Pending      Completed   Blocked  All time blocked
ReadStage                         3         2       19570606         0                 0
RequestResponseStage              0         1       10552500         0                 0
MutationStage                     0         0       11554725         0                 0
ReadRepairStage                   0         0         124792         0                 0
ReplicateOnWriteStage             0         0              0         0                 0
GossipStage                       0         0           2713         0                 0
AntiEntropyStage                  0         0             50         0                 0
MigrationStage                    0         0              0         0                 0
MemoryMeter                       1         4            545         0                 0
MemtablePostFlusher               0         0            417         0                 0
FlushWriter                       0         0            371         0                 0
MiscStage                         0         0             25         0                 0
PendingRangeCalculator            0         0             16         0                 0   
commitlog_archiver                0         0              0         0                 0
InternalResponseStage             0         0              0         0                 0
HintedHandoff                     0         0             10         0                 0

Message type           Dropped
RANGE_SLICE                  0
READ_REPAIR                  0
BINARY                       0
READ                         0
MUTATION                     0
_TRACE                       0
REQUEST_RESPONSE             0
Show/hide