Database Reference
In-Depth Information
EndpointDetails(host:10.10.21.206, datacenter:us-east,
rack:1a)])
TokenRange(start_token:2795178789021647703,
end_token:2799131500537456455, endpoints:[10.10.21.228,
10.10.21.206], rpc_endpoints:[10.10.21.228, 10.10.21.206],
endpoint_details:[EndpointDetails(host:10.10.21.228,
datacenter:us-east, rack:1a),
EndpointDetails(host:10.10.21.206, datacenter:us-east,
rack:1a)])
[-- snip -- ]
tpstats
The term
tpstats
stands for "thread pool statistics". This is an important measure of
how much pressure the node is bearing at a given instant. You can see task counts that are
running, pending, completed, and blocked in various stages. The rule of thumb is that
more pending tasks indicate a performance bottleneck. The following command-line out-
put shows the result of the usage of
tpstats
:
$ bin/nodetool tpstats
Pool Name Active Pending
Completed Blocked All time blocked
CounterMutationStage 0 0
0 0 0
ReadStage 0 0
13 0 0
RequestResponseStage 0 0
1298937 0 0
MutationStage 0 0
2542081 0 0
ReadRepairStage 0 0
0 0 0
GossipStage 0 0
7715 0 0
CacheCleanupExecutor 0 0
0 0 0
AntiEntropyStage 0 0
