DB Time & Average Active Session

DB Time is total time spent by user processes either actively working or actively waiting in a database call.

  •   DB Time is the time spent by foreground sessions, not background sessoins
  •   DB Time is the sum of time spent by all forground sessions, therefore it could be greater than the clock time (Elapsed Time) of a AWR snapshot window
  •   DB Time includes the time spent in database call, therefore it doens't include session idle time
  •   The primary components of DB Time are CPU time, IO time and non-idle wait time.

Average Active Session (AAS) is a metric used to describe the rate of DB Time change over time. There are two methods to calculate AAS.

  •   'cumulative AAS': for an AWR snapshot window, AAS = DB Time / Elapsed Time
  •   'sampled AAS': for a given time window within ASH, AAS = Session counts collected by sampling * Sampling interval time / Elapsed Time
 Case Study Summary

The case scenarios include:

 Case 1 : Use DB Time to detect performance bottleneck or degradation

Excessive DB Time is a sympton of either database performance bottleneck or performance degradation. The database performance issue can usually be identified by checking the time frame when DB Time is higher than normal.

DBspeed : 'DB Time S' = 'DB time in seconds'

 Case 2 : Use DB Time to check load balance among RAC instances

DB Time increases as database workload increases. For a RAC database, the variation of DB Time across all database instances can tell whether the goal of load balance is achieved or not.

 Case 3 : Understand the gap between AWR AAS and ASH AAS

There could be times when there is gap between AWR AAS and ASH AAS.

One of the most common causes is the accumulation delay of DB Time. For a long running operations, it contributes to the DB Time when it completes. This is called accumulation delay, which makes AWR AAS more senstive to end time of performance issue. On the contrary, Session Count of ASH jumps up whenever sessions get stuck, which makes ASH AAS more sensitive to the begin time of performance issue.

The gap between AWR AAS and ASH AAS could also be caused by other variables such as the AWR snapshot interval time, ASH sampling interval time, etc.

Copyright 2011 Actrace. All Rights Reserved.
Home | Product | Case Study | Support | Download | Contact Us