SCIENTIFIC-LINUX-USERS Archives

January 2011

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Konstantin Olchanski <[log in to unmask]>
Reply To:
Konstantin Olchanski <[log in to unmask]>
Date:
Thu, 13 Jan 2011 16:59:51 -0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (31 lines)
On Thu, Jan 13, 2011 at 03:22:11PM +1100, LEES, Cooper wrote:
> 1)
> Is there better options I should use to obtain better performance?
>
> 2)
> Is there a package (preferably within Scientific Linux Packages - if not
> easily compilable) to do benchmarks on a mounted NFS filesystem?


What kind of disk arrays you have at both ends? What's your network?
What performance do you get and what kind of improvement do you want or expect?

On a 100 Mbit/sec network, NFSv3 with default options can use 100% of the network (10 Mbytes/sec data rate).

On a GigE network, NFSv3 with default options can support "rsync" file transfers
at least up to 20-30-40 Mbytes/sec, depending on the disk configuration
on both ends.

For benchmarking, your final application is the best benchmark. Otherwise,
"time rsync", "time dd if=/dev/zero of=/nfs/somewhere/bigfile" and "time cp"
are okey benchmarks, but expect surprises when you run your final application.

Also before your benchmark NFS, you should benchmark the network and your disk arrays.


-- 
Konstantin Olchanski
Data Acquisition Systems: The Bytes Must Flow!
Email: olchansk-at-triumf-dot-ca
Snail mail: 4004 Wesbrook Mall, TRIUMF, Vancouver, B.C., V6T 2A3, Canada

ATOM RSS1 RSS2