| This component profiles the
MetricsService 's load
metrics for the cpu, messaging, and persistence load of the
node aggegate, each agent, and core services (mts & metrics).
Note that these are the metrics service's thread-based CPU
metrics, as opposed to the operating system's
ProcLoadAvg "loadavg" metrics.
This is the same output as the "/metrics/agent/load" servlet
generated by
org.cougaar.core.qos.metrics.AgentLoadServlet .
Example output:
load_node_NodeA - #cpu_load, cpu_load_jips,
recv_message_count, sent_message_count,
recv_message_bytes, sent_message_bytes,
persist_size_in_bytes,
load_node_NodeA - 0.88, 1270.32,
146.00, 145.90,
132248.03, 186308.48,
10518528.00,
load_agent_AgentA - 0.31, 183.71,
145.15, 145.05,
132090.91, 186105.49,
0.00,
load_agent_NodeA - 0.01, 19.82,
0.80, 0.80,
117.49, 147.17,
0.00,
load_service_MTS - #cpu_load, cpu_load_jips,
load_service_MTS - 0.49, 533.80,
load_service_Metrics - 0.08, 99.95,
load_service_NodeRoot - 0.00, 1.54,
See Also: ProfilerCoordinator See Also: required coordinator component |