Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations IamaSherpa on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Proliant ML370G2 Performance Findings

Status
Not open for further replies.

Guest_imported

New member
Jan 1, 1970
0
Server specs:
PIII 1.26GHz x 2, 512KB L2 Cache, 1GB + 256MB PC133 ECC SDRAM, 5 x 18.2GB Ultra3 SCSI-3 (on Raid 5)

The above specification/configuration was tested in isolation (single client cross cable to server) with different NIC on a same transaction and it shows different results as detail below

Built-in 10/100 NIC - response time (35 sec)
PCI 3COM 10/100 NIC - response time (24 sec)

Comparing against a desktop server (cross cable as above)
P4 - 1.6Mhz/256MB RAM/IDE HDD/Built-in 10/100 NIC
The same transaction gives a response of 17sec

Running the same transaction within the servers
Compaq server gives a response of 7 sec
Desktop server gives a response of 10 sec

From the above findings, my questions are :
1. Is it common for built-in NIC to be slower than PCI add-on NIC ?
2. Why is the strong makeup of this Compaq server does not outperform a normal desktop server ?
3. Is this test conclusive to say that the Compaq server has a flaw somewhere in its design ?


 
For a start, what OS were you using and what testing software was it? In answer to your questions (my 2 cents...):

1. Were you using the latest drivers for the onboard NIC?

2. If you had 100 clients trying to perform the same test to the desktop, you'd soon know about it - the server would cope quite adequately with it

3. NO! -----------------------------------------------------
"It's true, its damn true!"
-----------------------------------------------------
 
OS was Win2K and application was Progress application on Progress RDBMS.

1. Yes, latest NIC driver was loaded by Compaq engineer and it didn't show any improvement.

2. Granted. However, the points to which I'd like to draw feedback is that

a. the testing when done internally (with thin client and server application sitting in the same box) the Server shows excellent response compared to desktop server with the same setup as in () -> 7 secs vs 10 secs.

b. and when it was tested with external thin client connected to a server, the desktop server out perform the compaq server. -> 35 secs/24 secs vs 17 secs.

Why is there such inconsistency (Additionally, the motherboard was replaced for the sake of argument and the testing results were the same throughout).
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top