'); $("#tabs").tabs({ selected: 1 }); $("#ic-homepage__ic-tips").append( quickTipHTML() ); unhideOneProductTip(); $("#ic-homepage__product-tips").wrapInner('
'); $("#ic-homepage__feed-tips").wrapInner('
'); });
IBM Tivoli Monitoring > Version 6.3 > User's Guides > Agent Builder User's Guide > The Tivoli Monitoring Agent Editor IBM Tivoli Monitoring, Version 6.3
Watchdog information
Use the Watchdog Information page to specify configuration information for the Agent Watchdog
For more information about the Agent Editor and Agent Information page, see The Tivoli Monitoring Agent Editor.
To open the Watchdog Information page (Figure 1), click Watchdog Information in the Agent Content section of the Agent Information page. You can also select or the Watchdog Information node in the Outline View.
Figure 1. Watchdog Information page
You can specify the following configuration information for the Agent Watchdog:
- Monitor this agent by default
Select this check box to put the agent under management by Agent Management Services when the agent is installed. The agent is monitored for unhealthy behavior or abnormal termination and is restarted by a watchdog.
- Check frequency (seconds)
How often the watchdog checks the agent process for unhealthy behavior or abnormal termination. The default is every 180 seconds.
- Maximum number of restarts
Number of times the Watchdog restarts the agent because of unhealthy behavior or abnormal termination in a 24-hour period before it alerts the administrator of the problem. The period starts at midnight each day. So, the first period from when the agent is started might be "short."
A restart occurs if the agent goes down for any reason or if the Watchdog must stop the agent because it is unresponsive or unhealthy. For example, the Watchdog stops and then starts the agent if the memory threshold is crossed. The default is four restarts in a 24-hour period, where the period is measured from midnight to 11:59 p.m. At midnight, the daily restart count for the agent returns to 0 automatically.
- Memory Threshold Information
Size of the agent process (in megabytes) to which the agent can grow before its watchdog deems it unhealthy. There is a separate value for Windows, Linux, and UNIX. If the agent process grows beyond the threshold, the watchdog stops the process and restarts it. There are no defaults for these properties. If no value is specified, the Watchdog does not monitor the process size. The metric uses the working set size on Windows, and the user memory on UNIX and Linux.
If the Watchdog stops the agent, and the maximum number of restarts is reached, the Watchdog sends an alert that the agent exceeded its restart count, and stops doing auto-restarts. The Watchdog still reports whether the agent is up or down assuming it is started in another manner such as through the Tivoli Enterprise Portal.
You must manually restart the agent using the AMS Start Agent Take Action command so the restart count does not get reset.
The count gets reset in one of the following ways (the Watchdog continues to work and report status, but does not do auto-restarts):
- The clock strikes midnight.
- The user uses the AMS Start Agent Take Action command, which has an input parameter called resetRestartCount. If you enter a value of 1 (meaning "true"or "yes"), the daily restart count resets back to 0.
See the following sections in the IBM Tivoli Monitoring Administrator's Guide:
- For Tivoli System Monitor Agents
Configure Agent Management Services on Tivoli System Monitor Agents
- For Tivoli Enterprise Monitoring Agents
Install Tivoli Agent Management Services
Parent topic:
The Tivoli Monitoring Agent Editor