HP Data Protector Software Performance White Paper

43
Network backup of typical files
The typical files were saved from the client server via the network (Gigabit Ethernet) to the
remote backup server and its SCSI-attached Ultrium 960 tape drive.
Table 3 shows that the tape device in test 1 and the disk device in test 3 were faster than Data
Protector in Tests 4 and 5. In this scenario, both operating systems showed the same backup
performance via network, which was the bottleneck. The Gigabit Ethernet itself has a
1,000 Mb/s or 120 MB/s limitation which is very close to the results of test 5 (Windows 108.29
MB/s and HP-UX 111.84 MB/s).
Table 3. Network backup of typical files – bottleneck determination
Test Performance
(MB/s)
CPU Load
Client
CPU Load
Backup Server
Bottleneck?
1. Windows L&TT Tape Write
3
154.00 - - No
2. Windows HPReadData Single
3
105.25 - - No
3. Windows HPReadData Parallel
3
262.24 - - No
4. Windows DP NULL Parallel 108.16 12% 12%
Yes (Network)
5. Windows DP Ultrium 960 Parallel 108.29 16% 15%
Yes (Network)
1. HP-UX L&TT Tape Write
3
158.54 - - No
2. HP-UX HPReadData Single
3
130.58 - - No
3. HP-UX HPReadData Parallel
3
265.75 - - No
4. HP-UX DP NULL Parallel 112.83 26% 26%
Yes (Network)
5. HP-UX DP Ultrium 960 Parallel 111.84 24% 26%
Yes (Network)
Recommendation: If you are backing up typical files from a fast disk via Gigabit Ethernet to a
remote Ultrium 960 tape drive, you could consider backups without multiplexing and
concurrency. In this case, the network is the bottleneck.
3
Tested in the section Evaluating tape and disk drive performance on page 32