Submit a ticketCall us

WebinarUpcoming Webinar: How Help Desk and Remote Support Pays for Itself

Learn how help desk software can simplify ticketing management, allow you to track hardware and software assets, and accelerate the speed of IT support and service delivery. Gain insights on how remote support tools allow your IT team to maximize their efficiency and ticket resolution by expediting desktop troubleshooting, ultimately helping keep end-users happy and productive.

Register here.

Home > Success Center > Database Performance Analyzer (DPA) > DPA Documentation > DPA 11.1 Administrator Guide > Alerts > Alert categories

Alert categories

Created by Anthony.Rinaldi_ret, last modified by Anthony.Rinaldi_ret on Sep 15, 2016

Views: 711 Votes: 0 Revisions: 6

You can use the out-of-the-box alerts in DPA, or customize them to meet your needs. There are four preconfigured DPA alert categories.

Wait Time

Wait Time alerts are based on the amount of time users or applications waited on the database. For example, the Average SQL Wait Time alert fires when a SQL statement causes wait time that exceeds a threshold. These alerts are critical because they ensure that you are only alerted when users and applications are affected.


These alerts are based on the alarm thresholds of resource metrics, such as CPU utilization and memory usage. For example, the Single Resource Metric alert notifies you when the values of the specified resource metric exceed the thresholds.


Administrative alerts are typical alerts surrounding the health of the database system. For example, the Database Parameter Changes alert notifies you when any database parameter has changed. The Database Availability alert notifies you when the database instance is no longer accessible.


Custom alerts are user-specified queries that are run against the monitored database or the DPA repository. The query returns a number (or set of numbers) that may trigger an alert depending on user-defined threshold settings.

For example, you can enter a query to detect the number of canceled orders in the last 10 minutes:

select count(*) from orders where status='CANCELLED' and date > sysdate - 10/1440;

You can write a SQL query to set alerts on SQL Server health metrics.

You can see which type of orders have high cancellation rates:

select order_type, count(*) from orders where status='CANCELLED' and date > sysdate - 10/1440 group by order_type;

You may want to use complex logic to determine if an order has truly been canceled. In this case, you can write a function or procedure on the monitored database instance.

  • Oracle: select mycustomfunction(param1) from dual
  • SQL Server: select mycustomfunction(param1)
  • Sybase: mycustomprocedure(param1)
  • DB2: select dbo.mycustomfunction(param1) from SYSIBM.SYSDUMMY1
  • MySQL: mycustomprocedure(param1)
  • Azure SQL: select mycustomfunction(param1)

These functions or procedures are written in the native programming language of the database.

Last modified