A
anonymous
We have 2 servers, J1 and F1, and a 3rd machine accessing
them called A1. Both servers also access themselves and
each other through mapped drives. To keep drive letters
consistent each server also maps its local drive. We have
noticed the following read/write speeds using a small VB
app which writes then reads a 10MB file using the put/get
statements to write/read a large random byte array.
J1 Read Write (Mbits/sec)
local 6400 158
mapped local 1045 883
mapped F1 2133 388
F1 Read Write (Mbits/sec)
local 1766 156
mapped local 826 813
mapped J1 2438 445
A1 Read Write (Mbits/sec)
mapped F1 1219 279
mapped J1 1138 201
c:\temp 1164 100
J1 = Dell System
Precision Workstation 650
IDE Controller Intel 82801DB Ultra ATA storage Controller -
24CB
2 Disks
Dual Xeon 2.8Ghz without Hyper Threading
Serive Pack 4
2048Mb RAM
F1 = Dell System
Poweredge 1750
Dual Xeon 2.8Ghz Hyper Threading
Service Pack 4
2048Mb RAM
RAID 1
A1 =
PowerEdge 1400Sc
Dual 800ghz
1024Mb
Service PAck 2
Questions:
1) Why is the local read speed on J1 so much faster than
on F1
2) Why is it faster on F1 to read from J1 than to read
locally
them called A1. Both servers also access themselves and
each other through mapped drives. To keep drive letters
consistent each server also maps its local drive. We have
noticed the following read/write speeds using a small VB
app which writes then reads a 10MB file using the put/get
statements to write/read a large random byte array.
J1 Read Write (Mbits/sec)
local 6400 158
mapped local 1045 883
mapped F1 2133 388
F1 Read Write (Mbits/sec)
local 1766 156
mapped local 826 813
mapped J1 2438 445
A1 Read Write (Mbits/sec)
mapped F1 1219 279
mapped J1 1138 201
c:\temp 1164 100
J1 = Dell System
Precision Workstation 650
IDE Controller Intel 82801DB Ultra ATA storage Controller -
24CB
2 Disks
Dual Xeon 2.8Ghz without Hyper Threading
Serive Pack 4
2048Mb RAM
F1 = Dell System
Poweredge 1750
Dual Xeon 2.8Ghz Hyper Threading
Service Pack 4
2048Mb RAM
RAID 1
A1 =
PowerEdge 1400Sc
Dual 800ghz
1024Mb
Service PAck 2
Questions:
1) Why is the local read speed on J1 so much faster than
on F1
2) Why is it faster on F1 to read from J1 than to read
locally