The following list identifies some areas that may be pursued when investigating a problem that is believed to be processor occupancy (load) related.
Multiply the number in the Total Calls field by 20 for the identified 3-minute time interval (the time when the problem occurred).
ADDITIONAL INFORMATION:
If the product exceeds the advertised BHCC of the configuration, it is the load for this time interval causing the problem. If the product does not exceed the BHCC for the configuration, this load is not the problem.
Compare the number in the Total Atmpts field with the Total Calls field for the identified 3-minute time interval (the time when the problem occurred).
ADDITIONAL INFORMATION:
If the number of attempts is significantly greater than the number of calls, a significant percent of the occupancy is due to processing off-hook and on-hook stimuli that do not result in a completed call.
Examine the hardware error log for an excessive amount of maintenance activity (for example, a high number of errors).
Refer to the list measurements communications-links report to determine if any of the links are receiving an abnormal amount of traffic.
Check with the users to determine if a certain feature(s) is used heavily during the identified time interval.
Refer the problem to maintenance personnel with the suggestion they check the software error log.
Result
After a serious performance problem is detected and corrected, use the clear measurements occupancy busiest-intervals command and clear the log of busiest entries. You can then focus on current performance problems.
Note:
You must only use the clear measurements occupancy busiest-intervals command to clear data from resolved problems.
