This section explains the procedure for setting a separate timeout value for each monitored object for service operation monitoring.
Point
Use this procedure if the timeout values for service operation monitoring need to be changed while service operation management is being performed.
Privileges required for execution
[Windows]
The privileges of a user belonging to the "Administrators" group are required to make these settings.
[UNIX]
System administrator (superuser) privileges are required to make these settings.
Storage location
The storage location of this file is as follows:
[Windows]
<Variable file directory>\control\template.dat |
[UNIX]
/etc/opt/FJSVssqc/template.dat |
The service operation monitoring function has two types of timeouts.
Collection timeout
The value of the collection timeout is the maximum amount of time allowed for the collection process (the process that operates during each collection interval). The collection timeout is set at 70 seconds by default.
When a collection timeout occurs, all the data collected during that collection interval becomes invalid and no performance information record is created.
Monitoring timeout
The value of the monitoring timeout is the maximum amount of time allowed for receiving a response to a request that was sent to a monitored object. The monitoring timeout is set at 10 seconds by default.
If a monitoring timeout occurs, "-1" is stored as the performance value of the monitored object that caused the timeout.
Note
"-1" is also stored in the performance information record when communications errors occur.
Approach for defining monitored objects
Performance information records are not created if collection timeouts occur. This means that definitions must be made so that collection timeouts do not occur, in order to perform successful monitoring.
If there are multiple monitored objects, the number of monitored objects and the values for monitoring timeouts and collection timeouts must satisfy the condition shown in the formula below, in order to take into account the possibility that timeouts will occur for all monitored objects.
Number of monitored objects * Monitoring timeout (10 seconds) < Collection timeout (70 seconds) |
Note: The maximum number of monitored objects is set at six by default.
The following items for the service operation monitoring function can be changed using the template file.
Collection interval: Either 1, 2, 5 or 10 (minutes) can be specified
Monitoring timeout value: Any value equal to or less than the collection interval
Collection timeout value: 5 seconds (or collection interval whichever is greater) + 30 seconds
Note
The larger the value specified for the monitoring timeout value, the less objects can be monitored, so take into account the number of monitored objects when specifying the monitoring timeout.
A warning message will be output if, as a result of the settings in "A.2 Response/Operation Information Collection Policy Creation Command", the number of monitored objects and the timeout values do not satisfy the condition shown in the formula above. However, the policy will be created even though the warning message is output.
If there are a lot of monitored objects, or if there is a problem with the settings for the monitored objects, the command may take a long time to return.
Information used
template.dat PING section CMDTIMEOUT parameter Default value: 70 seconds
template.dat PING section TIMEOUT parameter Default value: 10 seconds
ServiceConf.xml The number of monitored objects for each monitoring type
