Intel 170 Servers User Manual
Chapter 15. Save/Restore Performance
This chapter’s focus is on the IBM i operating system platform. For legacy system models, older
device attachment cards, and the lower performing backup devices see the V5R3 performance capabilities
reference.
device attachment cards, and the lower performing backup devices see the V5R3 performance capabilities
reference.
Many factors influence the observable performance of save and restore operations. These factors include:
y
y
The backup device models, number of DASD units the data is spread across, processors, LPAR
configurations, IOA used to attach the devices.
configurations, IOA used to attach the devices.
y
Workload type: Large Database File, User Mix, Source File, integrated file system (Domino, Network
Storage, 1 Directory Many Objects, Many Directory Many Objects.
Storage, 1 Directory Many Objects, Many Directory Many Objects.
y
The use of data compression, data compaction, and Optimum Block Size (USEOPTBLK)
y
Directory structure can have a dramatic effect on save and restore operations.
15.1 Supported Backup Device Rates
As you look at backup devices and their performance rates, you need to understand the backup device
hardware and the capabilities of that hardware. The different backup devices and IOAs have different
capabilities for handling data for the best results in their target market. The following table contains
backup devices and rates. Later in this document the rates are used to help determine possible
performance. A study of some customer data showed that compaction on their database file data
occurred at a ratio of approximately 2.8 to 1. The database files used for the performance workloads were
created to simulate that result.
hardware and the capabilities of that hardware. The different backup devices and IOAs have different
capabilities for handling data for the best results in their target market. The following table contains
backup devices and rates. Later in this document the rates are used to help determine possible
performance. A study of some customer data showed that compaction on their database file data
occurred at a ratio of approximately 2.8 to 1. The database files used for the performance workloads were
created to simulate that result.
#1. Software compression is used here because the hardware doesn’t support device compaction
Note the compaction factor is a number that used with the formulas in the following chapter to help describe the actual rates
observed as the lab workloads were run using the above drives. This is not the compression ratio of the data being written to
tape. I list them here to help understand what our experiments were able to achieve relative to the published drive speed.
Note the compaction factor is a number that used with the formulas in the following chapter to help describe the actual rates
observed as the lab workloads were run using the above drives. This is not the compression ratio of the data being written to
tape. I list them here to help understand what our experiments were able to achieve relative to the published drive speed.
2.5
100.0
3592E Fiber Channel
2.0
120.0
3580 Ultrium 4 Fiber Channel
2.0
120.0
5746 Half High Ultrium 4
2.0
80.0
3580 Ultrium 3 Fiber Channel)
2.8
40.0
3592J Fiber Channel
2.8
35.0
3580 Ultrium 2
2.8
18.0
5755 ½ High Ultrium-2
2.0
6.0
6258 4MM tape Drive
2.0
12.0
6279 VXA-320
2.0
6.0
VXA-2
2.0
5.0
SLR100
2.0
4.0
SLR60
2.8
#1
2.5
SAS DVD-RAM
2.8
#1
0.75 Write/2.8 Read
DVD-RAM
COMPACTION FACTOR
Rate (MB/S)
Backup Device
Table 15.1.1 backup device speed and compaction information
IBM i 6.1 Performance Capabilities Reference - January/April/October 2008
©
Copyright IBM Corp. 2008
Chapter 15. Save/Restore Performance
243