The sample threshold monitoring definition file can be used to enable threshold monitoring to be performed with respect to the server performance information items shown in the following table.
Windows server performance information
If the monitored object is running Windows, the sample definition file can be used to monitor the following server performance information.
No. | Monitoring item | Evaluation method |
|---|---|---|
1 | CPU usage [%] | If the usage rate stays above 80% (48 seconds' usage in one minute), then there is a bottleneck at the CPU and performance problems may occur (or may have occurred already). |
2 | Physical disk idle time [sec] | If the disk busy rate for a physical disk stays above 60% (that is, if the physical disk is idle for less than 24 seconds in one minute), then the disk load is causing a bottleneck and performance problems may occur (or may have occurred already). |
3 | Free disk space[%] | Business activities may stop if the amount of free disk space becomes too low. |
4 | Free memory [bytes] | If the amount of free memory intermittently trends around 4 MB, then there is a memory bottleneck and performance problems may occur (or may have occurred already). |
Solaris server performance information
If the monitored object is running Solaris, the sample definition file can be used to monitor the following server performance information.
No. | Monitoring item | Evaluation method |
|---|---|---|
1 | CPU usage [%] | If the usage rate stays above 90% (54 seconds' usage in one minute), then there is a bottleneck at the CPU and performance problems may occur (or may have occurred already). |
2 | Physical disk busy time [sec] | If the disk busy rate for a physical disk stays above 60% (that is, if the physical disk is busy for more than 32 seconds in one minute, then the disk load is causing a bottleneck and a performance problem may occur (or may have occurred already). |
3 | Free disk space [%] | Business activities may stop if the amount of free disk space becomes too low. |
4 | Free memory [bytes] | If the amount of free memory intermittently trends around the value of the "lotsfree" kernel parameter (*1), then there is a memory bottleneck and performance problems may occur (or may have occurred already). |
*1: The value that has been set for the "lotsfree" kernel parameter can be checked using the kstat command. By default, the value of this parameter is either 1/64 of physical memory or 512 KB, whichever is larger. Refer to the Solaris manuals for more information.
Linux server performance information
If the monitored object is running Linux, the sample definition file can be used to monitor the following server performance information.
No. | Monitoring item | Evaluation method |
|---|---|---|
1 | CPU usage [%] | If the usage rate stays above 90% (54 seconds' usage in one minute), then there is a bottleneck at the CPU and performance problems may occur (or may have occurred already). |
2 | Physical disk busy time [sec] | If the disk busy rate for a physical disk stays above 80% (that is, if the physical disk is busy for more than 48 seconds in one minute), then the disk load is causing a bottleneck and a performance problem may occur (or may have occurred already). |
3 | Free disk space [%] | Business activities may stop if the amount of free disk space becomes too low. |
4 | Available memory space [bytes] | If the amount of free memory tends to intermittently show a low value, then memory shortages may be a bottleneck and a performance problem may occur (or may have occurred already). The threshold value for the amount of free memory varies according to factors such as the version of Linux, the edition and the amount of memory on board. It should be adjusted to suit the operating conditions. The free memory threshold is set at 5,120 KB in the sample file. |
Storage directory
The sample file is stored in the following directory:
[Windows]
<Installation directory >\sample\alertconfig.txt |
[UNIX]
/opt/FJSVssqc/sample/alertconfig.txt |
Point
To use the sample definition file on a server where threshold monitoring is being performed, back up the existing threshold definition file (alertconfig.txt) and then overwrite it with the sample definition file.
