Compaq ProLiant 6000 Compaq DLT Tape Array II: High-Performance Backup of Ente - Page 29

Test Feed Speed Relationships

Get Compaq ProLiant 6000 PDF manuals and user guides View all Compaq ProLiant 6000 manuals
Add to My Manuals
Save this manual to your list of manuals

Page 29 highlights

ECG075.0997 (cont.) DLT Tape Array II ... TEST 11 FEED SPEED RELATIONSHIPS Windows NT 4.0 Cheyenne ARCserve 6.0 for Windows NT By image, single and multiple jobs 2 GB 4:1 RAIT-0 (not significant when using one drive) ProLiant 5000, two Pentium Pro 200 processors, 256K cache, 256 MB RAM; and ProLiant 6000, four Pentium Pro 200 processors, 256 K cache, 256 MB RAM Embedded or SMART-2/P SCSI (53C825) Controller (see details below) 35/70 GB DLT drive(s) Wide-Ultra SCSI-3 Cards (All Compaq off-the-shelf products) To demonstrate the relationship of feed speed to backup speed. Single and multiple job tests were performed to demonstrate the relationship of bandwidth to bottlenecks within the operating system The tests were executed in four environments: • The server used five 4-GB Fast-Wide SCSI drives, each formatted under Windows NT as a single logical drive. All drives were connected to the embedded SCSI (53C825) controller. Tests A and B used only one drive. Test C used four logical drives (each on a separate physical drive). • The server data drives used the SMART-2/P Array Controller to operate twelve 4-GB Fast-Wide SCSI drives configured as a single RAID-0 drive. Windows NT was used to configure the array as a single drive letter (stripe set) for Tests A and B. Test C used four logical drives created by assigning three physical disk drives in RAID-0 to each logical disk using the array configuration utility and formatting them under Windows NT. • The server data drives used two SMART-2/P Array controllers to operate twenty-four 4-GB Fast-Wide SCSI drives formatted as one RAID-0 drive (and one Windows NTstriped drive) for Tests A and B. Test C used four logical disks created by assigning one channel of each SMART-2/P Array Controller to each logical drive (six drives per channel) in a RAID-0. Each Smart-2/P Array Controller card was installed on a separate PCI bus within the server, for performance balancing. • The server data drives used four SMART-2/P Array controllers to operate fifty-six 4GB Fast-Wide SCSI drives formatted as one RAID-0 drive (and one Windows NTstriped drive) for Tests A and B. Test C used four logical disks, created by assigning one channel of each SMART-2/P Array Controller to each logical drive (seven drives per channel) in a RAID-0. Each SMART-2/P Array Controller card was installed on a separate PCI bus within the server, for performance balancing.

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
7
%
(cont.)
DLT Tape Array II
ECG075.0997
T
EST
11
F
EED
S
PEED
R
ELATIONSHIPS
Windows NT 4.0
Cheyenne ARCserve 6.0 for Windows NT
By image, single and multiple jobs
2 GB
4:1
RAIT-0 (not significant when using one drive)
ProLiant 5000, two Pentium Pro 200 processors, 256K
cache, 256 MB RAM; and ProLiant 6000, four Pentium Pro
200 processors, 256 K cache, 256 MB RAM
Embedded or SMART-2/P SCSI (53C825) Controller (see
details below)
35/70 GB DLT drive(s)
Wide-Ultra SCSI-3 Cards
(All Compaq off-the-shelf products)
To demonstrate the relationship of feed speed to backup speed. Single and multiple job tests
were performed to demonstrate the relationship of bandwidth to bottlenecks within the
operating system
The tests were executed in four environments:
•
The server used five 4-GB Fast-Wide SCSI drives, each formatted under Windows NT
as a single logical drive. All drives were connected to the embedded SCSI (53C825)
controller. Tests A and B used only one drive. Test C used four logical drives (each on
a separate physical drive).
•
The server data drives used the SMART-2/P Array Controller to operate twelve 4-GB
Fast-Wide SCSI drives configured as a single RAID-0 drive. Windows NT was used
to configure the array as a single drive letter (stripe set) for Tests A and B. Test C used
four logical drives created by assigning three physical disk drives in RAID-0 to each
logical disk using the array configuration utility and formatting them under Windows
NT.
•
The server data drives used two SMART-2/P Array controllers to operate twenty-four
4-GB Fast-Wide SCSI drives formatted as one RAID-0 drive (and one Windows NT-
striped drive) for Tests A and B. Test C used four logical disks created by assigning
one channel of each SMART-2/P Array Controller to each logical drive (six drives per
channel) in a RAID-0. Each Smart-2/P Array Controller card was installed on a
separate PCI bus within the server, for performance balancing.
•
The server data drives used four SMART-2/P Array controllers to operate fifty-six 4-
GB Fast-Wide SCSI drives formatted as one RAID-0 drive (and one Windows NT-
striped drive) for Tests A and B. Test C used four logical disks, created by assigning
one channel of each SMART-2/P Array Controller to each logical drive (seven drives
per channel) in a RAID-0. Each SMART-2/P Array Controller card was installed on a
separate PCI bus within the server, for performance balancing.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.