You can obtain continuous performance data for long term reporting, analysis and trending. Acquire can read through an entire dumped SMF dataset, or a concatenation of them, and produce data for a correspondingly large number of measurement intervals in a single output file, or multiple output files if more than one system’s data is present in the input file(s).
If you need to find out what timestamps and SMF IDs are represented in the input SMF dataset(s), you can carry out a SCAN run, as described above. We recommend that you do not capture high volume data like CICS, DB2, job detail, MQ or WAS unless you are sure you want to, and can cope with the potentially very large work file requirements, and large volumes of data in your Capacity Database—many hundreds of megabytes per day is not uncommon for large systems.
To capture data for a succession of measurement intervals, you perform a COLLECT run, under control of the JCL member COLLECT. The output from this run is a dataset containing data for all the measurement intervals, ready for import to the Capacity Database. To update the Capacity Database with this kind of data, use the Syncsort Capacity Management Control Center, as described in the System Manager online help.
Refer to the sections Overview of Acquire JCL and Acquire parameter specifications for more detailed information.