If you are performing your own usage data analysis, (perhaps through an exit routine you provide for IEAURP), you may need to analyze two records. If your customer does not specify interval synchronization and are collecting data at the task level, you may receive multiple type 89 records for the same reporting interval. Without synchronization, task data collection, which is invoked by SMF address space interval processing, might not coincide with type 89 record generation (although both may have the same interval value, such as 30 minutes).
Suppose task data collection is scheduled for 4:45 and 5:15, and type 89 record generation is scheduled for 5:00 and 5:30. Both are on a 30 minute interval, but not synchronized. It is possible that the 5:30 reporting interval will include usage data from both the 4:00 - 5:00 and the 5:00 - 6:00 usage data intervals. Because the usage reporting program presents product usage in hourly buckets, SMF generates two records for the 5:30 interval. The first represents usage for the 4:00 - 5:00 usage data interval and the second reflects usage for the 5:00 - 6:00 usage data interval.
There can also be multiple type 89 records for a given interval if the number of products to be reported causes the record to exceed the maximum allowable size (32,756 bytes). In this case, the information is continued on a subsequent record.