Intel 170 Servers, AS/400 RISC Server, 7xx Servers manual Comparing Performance Data

15.4 Comparing Performance Data

When comparing the performance data in this document with the actual performance on your system, remember that the performance of save and restore operations is data dependent. If the same backup device was used on data from three different systems, three different rates may result.

The performance of save and restore operations are also dependent on the system configuration, most directly affected by the number and type of DASD units on which the data is stored and by the type of storage IOAs being used.

Generally speaking, the Large Database File data that was used in testing for this document was designed to compact at an approximate 2.8 to1 ratio. If we were to write a formula to illustrate how performance ratings are obtained, it would be as follows:

((DeviceSpeed * LossFromWorkLoadType) * Compaction) = MB/Sec * 3600 = MB/HR /1000 = GB/HR.

But the reality of this formula is that the “LossFromWorkLoadType” is far more complex than described here. The different workloads have different overheads, different compaction rates, and the backup devices use different buffer sizes and different compaction algorithms. The attempt here is to group these workloads as examples of what might happen with a certain type of backup device and a certain workload.

Note: Remember that these formulas and charts are to give you an idea of what you might achieve from a particular backup device. Your data is as unique as your company and the correct backup device solution must take into account many different factors.

The save and restore rates listed in this document were obtained on a dedicated system. A dedicated system is one where the system is up and fully functioning but no other users or jobs are running except the save and restore operations. All processors and Memory were dedicated to the system and no partial processors were used. Other subsystems such as QBATCH are required in order to run concurrent and parallel operations. All workloads were deleted before restoring them again.

IBM i 6.1 Performance Capabilities Reference - January/April 2008
© Copyright IBM Corp. 2008	Chapter 15. Save/Restore Performance	246