Intel 170 Servers 14.1.2 iV5R2 Direct Attach Dasd, Intensive Workload Performance Comparison

14.1.2 iV5R2 Direct Attach DASD

This section discusses the direct attach DASD subsystem performance improvements that were new with the iV5R2 release. These consist of the following new hardware and software offerings :

y2757 SCSI PCI RAID Disk Unit Controller (IOA)

y2780 SCSI PCI RAID Disk Unit Controller (IOA)

y2844 PCI Node I/O Processor (IOP)

y4326 35 GB 15K RPM DASD

y4327 70 GB 15K RPM DASD

Note: for more information on the older IOAs and IOPs listed here see a previous copy of the Performance Capabilities Reference.

14.1.2.1

I/O Intensive Workload Performance Comparison

Compare 2778/2757 & 5074 vs 5094

2778/6718 5074 tower 18GB 10K rpm 2757/4326 5094 tower 35GB 15K rpm

2757/6718 5074 tower 18GB 10K rpm 2757/4327 5094 tower 70GB 15K rpm

2778/6719 5074 tower 35GB 10K rpm 2757/4328 5094 tower 140GB 15K rpm

2757/6719 5074 tower 35GB 10K rpm 40% DASD Subsystem Utilization 2757/6719 5094 tower 35GB 10K rpm

(Sec)	0.35
	0.3
Time	0.25
Response	0.2
	0.15
	0.1
System
	0.05
	0
	0	1000	2000	3000	4000	5000	6000	7000

Workload Throughput

For our workload we attempt to fill the DASD units to between 40 and 50% full so you are comparing units with more actual data, but trying to keep the relative seek distances similar. The reason is that larger capacity drives can appear to be faster than lower capacity drives in the same environment running the same workload in the same size database. That perceived improvement can disappear, or even reverse depending upon the workload (primarily because of where on the disks the data is physically located).

IBM i 6.1 Performance Capabilities Reference - January/April/October 2008
© Copyright IBM Corp. 2008	Chapter 14 DASD Performance	192