Throughput - Service Classes

Purpose

Show how much work is happening in each WLM service class.

One of the main purposes of WLM service classes is to apply different priorities and or policies to different types of work. For example, limiting the number of large queries allowed to run concurrently. Service classes are also used to give higher priority to business critical applications.

Validating that service classes are serving their intended purpose requires monitoring the performance of activities that are running in them.

Each row in this view shows a service subclass. You can use filtering or sorting to group the service subclasses by service superclass.

Metrics Shown

Service Superclass Name

Source: mon_get_service_subclass.service_superclass_name

Service Subclass Name

Source: mon_get_service_subclass.service_subclass_name

CPU Time

Source: mon_get_service_subclass.total_cpu_time

Activities Completed / m

The number of activities completed per minute is one measure of how much work is being accomplished by connections associated with this service subclass. Activities typically correspond to a query, although there a few additional things that count as an activity including LOAD and stored procedures calls.

Source: mon_get_service_subclass.act_completed_total

Activities Aborted / m

Source: mon_get_service_subclass.act_aborted_total

Activities Rejected / m

Source: mon_get_service_subclass.act_rejected_total

App Rqsts Completed / m

Source: mon_get_service_subclass.app_rqsts_completed_total

Avg App Rqst Time (ms)

Activities Queued / m

Source: mon_get_service_subclass.wlm_queue_assignments_total

Rows Read / m

Source: mon_get_service_subclass.rows_read