Compaq ProLiant 6000 Compaq DLT Tape Array II: High-Performance Backup of Ente - Page 25

Etwork, Erformance, Indows, Erver, Ulling From An, Ntranet, Emote, Lient

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 25 highlights

ECG075.0997 (cont.) DLT Tape Array II ... TEST 9 NETWORK PERFORMANCE  WINDOWS NT SERVER PULLING FROM AN INTRANETWARE REMOTE CLIENT Windows NT 4.0 (server) pulling IntranetWare 4.11 Server (as client) Cheyenne ARCserve 6.0 for Windows NT Single job 2 GB 4:1 ARCserve 6 for NetWare high-performance push agent option RAIT-0 ProLiant 5000, two Pentium Pro 200 processors, 256K cache, 256 MB RAM 80-GB array (two SMART-2/P Array Controllers with twenty 4-GB drives) 35/70 GB DLT drive(s) Wide-Ultra SCSI-3 cards (All Compaq off-the-shelf products) 10/100TX - Ethernet over CAT-5 cable through simple hub TCP/IP protocol To test the throughput of a multiple server backup over a single network connection. The throughput of more than 32 GB/hr for four jobs is close to the limit of the 100TX wire, though the NT client backup agent helps slightly. The Windows NT client does allow multiple backup streams; the network is still the limit, in most cases. For this reason, it is not surprising that the Windows NT numbers are very close to the Novell IntranetWare numbers in the network tests. Using multiple 100TX-controller cards on a switched hub to independent clients may allow network backups to approach local backup speeds; this would be scaleable to the limits of the tape drives. IntranetWare client agent did not support multiple streams to the Windows NT backup server. The reverse test, with IntranetWare 4.11 server pulling from a base of Windows NT Clients, was not considered for two reasons. If using Cheyenne JETserve, it is strictly an image backup, used as a local backup of the NetWare server's drives. Also, in comparing Cheyenne ARCserve for Windows NT to Cheyenne ARCserve for NetWare, it was noted that the Novell solution was not as robust a candidate. The NT based solution was initially chosen to best illustrate the objective. This is not to say that Cheyenne ARCserve for NetWare would not perform adequately in a Novell environment. Later tests show that all Novell solutions perform as well as the NT solutions.

  • 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
9
N
ETWORK
P
ERFORMANCE
W
INDOWS
NT S
ERVER
P
ULLING FROM AN
I
NTRANET
W
ARE
R
EMOTE
C
LIENT
Windows NT 4.0 (server) pulling IntranetWare 4.11
Server (as client)
Cheyenne ARCserve 6.0 for Windows NT
Single job
2 GB
4:1
ARCserve 6 for NetWare high-performance push agent
option
RAIT-0
ProLiant 5000, two Pentium Pro 200 processors, 256K
cache, 256 MB RAM
80-GB array (two SMART-2/P Array Controllers
with twenty 4-GB drives)
35/70 GB DLT drive(s)
Wide-Ultra SCSI-3 cards
(All Compaq off-the-shelf products)
10/100TX – Ethernet over CAT-5 cable through simple
hub TCP/IP protocol
To test the throughput of a multiple server backup over a single network connection.
The throughput of more than 32 GB/hr for four jobs is close to the limit of the 100TX wire,
though the NT client backup agent helps slightly. The Windows NT client does allow multiple
backup streams; the network is still the limit, in most cases. For this reason, it is not surprising
that the Windows NT numbers are very close to the Novell IntranetWare numbers in the
network tests. Using multiple 100TX-controller cards on a switched hub to independent clients
may allow network backups to approach local backup speeds; this would be scaleable to the
limits of the tape drives.
IntranetWare client agent did not support multiple streams to the Windows NT backup server.
The reverse test, with IntranetWare 4.11 server pulling from a base of Windows NT Clients,
was not considered for two reasons. If using Cheyenne JETserve, it is strictly an image backup,
used as a local backup of the NetWare server’s drives. Also, in comparing Cheyenne ARCserve
for Windows NT to Cheyenne ARCserve for NetWare, it was noted that the Novell solution
was not as robust a candidate. The NT based solution was initially chosen to best illustrate the
objective. This is not to say that Cheyenne ARCserve for NetWare would not perform
adequately in a Novell environment. Later tests show that all Novell solutions perform as well
as the NT solutions.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.