17.00 - Intervals - Teradata Workload Management

Teradata Vantageā„¢ - Workload Management User Guide

Advanced SQL Engine
Teradata Workload Management
June 2020
User Guide
You can set the following intervals on the Other tab of the General view:
Event Interval
How often TASM evaluates all of the events defined in the state matrix. The default is 60 seconds.
Flex Throttle Action Interval
How often TASM evaluates system resource availability. The default is 60 seconds.
Dashboard Interval
How often TASM collects data about workloads, including the following metrics:
  • Arrivals
  • Completions
  • Delays
  • Exceptions
  • Average response time
  • CPU use
  • I/O use
  • The number of requests that meet their SLGs
This data is displayed real-time in the Viewpoint Workload Monitor portlet. It is also logged in the long-term repository, TDWMSummaryLog. Teradata recommends setting this interval to 60 seconds to match the default refresh rate of the Viewpoint Workload Monitor portlet.
Logging Interval
How often TASM flushes log files from memory to disk. The following log tables are refreshed at this interval:
  • TDWMSummaryLog
  • TDWMExceptionLog
  • TDWMEventLog
  • TDWMEventHistoryLog
If log memory cache fills up before the logging interval ends, Vantage flushes the log file to disk immediately. The Logging Interval specifies the maximum time between memory flushes.
Exception Interval
How often TASM checks for exceptions. The default is 60 seconds, which is a reasonable interval for detecting a long-running step.

Both the Dashboard Interval and the Logging Interval must be multiples of the Event Interval. Also, the Logging Interval must be a multiple of the Dashboard Interval. This ensures a smooth roll-up of summary data to the dashboard data and the logging data.

The Exception Interval affects only exception checks done during a request, when the request duration exceeds the exception checking interval. The Exception Interval does not affect the exception checks TASM does at the end of each request step.
Tip: Teradata recommends that the Exception Interval be less than or equal to the session rate, which is the rate for updating session-level statistics in memory. The session rate is controlled by the PM/API request SET SESSION RATE.

If using skew difference, beware of an issue in detection accuracy: Whenever multiple applications (besides TASM asynchronous exception checking) issue MONITOR SESSION commands, the internal collection cache is flushed and accumulations restart at the shortest interval being used. If Teradata Viewpoint is set to refresh (submit MONITOR SESSION) every 30 seconds and the exception interval is set to 60 seconds, Viewpoint will receive new data and reset the cache every 30 seconds. When TASM issues the MONITOR SESSION command, it will contain not 60 seconds, but just 30 seconds of accumulated data.

Tip: In order to keep the MONITOR SESSION data TASM uses as complete as possible, Teradata recommends that other PMPC applications that use MONITOR SESSION do so at an interval greater than the exception interval. That way, the other PMPC applications will always get the PMPC statistics TASM collects, maintaining the accuracy of TASM in computing skew difference values.
An additional timer, DBQLFlushRate, defined in the DBS Control utility General fields, tells the system how often to flush all DBQL tables (including step, object, and SQL tables). The Logging Interval causes DBQLogTbl to be flushed at an additional rate.