DB2 Version 9.7 for Linux, UNIX, and Windows

agent_wait_time - Agent wait time monitor element

Time spent by an application queued to wait for an agent under concentrator configurations. The value is given in milliseconds.

Table 1. Table Function Monitoring Information
Table Function	Monitor Element Collection Level
MON_GET_CONNECTION table function - Get connection metrics	REQUEST METRICS BASE
MON_GET_CONNECTION_DETAILS table function - Get detailed connection metrics (reported in DETAILS XML document)	REQUEST METRICS BASE
MON_GET_SERVICE_SUBCLASS table function - Get service subclass metrics	REQUEST METRICS BASE
MON_GET_SERVICE_SUBCLASS_DETAILS table function - Get detailed service subclass metrics (reported in DETAILS XML document)	REQUEST METRICS BASE
MON_GET_UNIT_OF_WORK table function - Get unit of work metrics	REQUEST METRICS BASE
MON_GET_UNIT_OF_WORK_DETAILS table function - Get detailed unit of work metrics (reported in DETAILS XML document)	REQUEST METRICS BASE
MON_GET_WORKLOAD table function - Get workload metrics	REQUEST METRICS BASE
MON_GET_WORKLOAD_DETAILS table function - Get detailed workload metrics (reported in DETAILS XML document)	REQUEST METRICS BASE

Table 2. Event Monitoring Information
Event Type	Logical Data Grouping	Monitor Switch
Statistics	event_scstats (reported in the metrics document)	REQUEST METRICS BASE
Statistics	event_wlstats (reported in the metrics document)	REQUEST METRICS BASE
Unit of work	Reported in the in the system_metrics document.	-
Statistics	event_scmetrics*	REQUEST METRICS BASE
Statistics	event_wlmetrics*	REQUEST METRICS BASE
* When returned as part of this logical data group, this element reflects the change in value of this metric since the last statistics collection or database activation, whichever was more recent.

Usage

The agent_wait_time monitor element can be used to help evaluate how efficiently your system is running in a concentrator environment. A high agent wait relative to the total_request_time monitor element value indicates that requests are spending a lot of time queued waiting for agents, which may be indicative of one or more of the following:

The max_coordagents configuration parameter has been configured too small for your workload. You may need to increase the value of max_coordagents configuration parameter, or the ratio of max_coordagents configuration parameter to max_connections configuration parameter if you are running with both parameters set to AUTOMATIC, to ensure that enough coordinator agents are available to service your application requests in a timely manner.
Your workload is not committing frequently enough. For the concentrator to work efficiently, applications should issue commits relatively frequently to ensure that their agents can be freed up to serve requests on other applications. If your applications do not do frequent commits you may need to configure a proportionally higher number of coordinator agents to reduce the time spent waiting for agents to become available.