Networker

Re: [Networker] Optimizing / performance tuning storage nodes

2009-01-08 14:13:41
Subject: Re: [Networker] Optimizing / performance tuning storage nodes
From: Fazil Saiyed <Fazil.Saiyed AT ANIXTER DOT COM>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Thu, 8 Jan 2009 13:11:22 -0600
Hi,
If i understand your configuration correctly, You should be able to use 
Brodcomm Control Suite to configure NIC Teaming and utilize both NIC for 
maximum throughput.
I am concerned about number of clients you have configured to backup from 
this SN, i would have like the ratio to be more like 50 servers to 1 SN ( 
ofcorce if you do cloning here or any of this server are superlarge that 
would prompt design re-consideration).
NIC utilization  could be improved, hopefully less then 70 peak would be 
desirable, you could try and stagger the backups to lower the utilization.
Your overall data size does not seems to  be too high.
You should be able to push the EDL close to it's capabilities with the 
tape drives, but realize that if all 5 Tape drive write at the same time, 
you may overload the FC ( assuming you have 2 GB FC connections), if you 
have  4 GB FC connection, you may be in better shape as long as back end 
FC connections  to EDL are not shared among multiple SNS.
HTH



JK <johannes.karlsson AT SKYRR DOT IS> 
Sent by: EMC NetWorker discussion <NETWORKER AT LISTSERV.TEMPLE DOT EDU>
01/08/2009 12:49 PM
Please respond to
EMC NetWorker discussion <NETWORKER AT LISTSERV.TEMPLE DOT EDU>; Please respond 
to
JK <johannes.karlsson AT SKYRR DOT IS>


To
NETWORKER AT LISTSERV.TEMPLE DOT EDU
cc

Subject
Optimizing / performance tuning storage nodes






Hi.

I want to configure a very robust storage node to write to an EDL 4200.

To do that I have zoned a DELL PE1955 blade through 2 HBS's to the EDL. 
Each HBA is directed to a dedicated FC port on the EDL and each HBA on the 

storage node has got 5 tape drives on the same VTL on the EDL. This 
storage 
node will handle about 200 clients and is going to write about 25TB of 
data 
every week, so I really need to think about throughput.

Now, the NIC is the bottleneck. The utilization is at 70% constant and 
peaking to 85%. The PE1955 blade has got 2x 1Gb Nic's, but I fail to see 
how I can utilize both of the nic's. 

Does anyone have an idea how I can enhance the robustness even further?

What would be the best way to use the NIC's?

Regards, Johannes

To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type 
"signoff networker" in the body of the email. Please write to 
networker-request AT listserv.temple DOT edu if you have any problems with this 
list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER


To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type "signoff networker" in the body of the email. Please write to 
networker-request AT listserv.temple DOT edu if you have any problems with this 
list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER