On Mon, 19 Nov 2007, Joe Landman wrote:
> Hmmm... I am not sure if Chris wants to "buy" support for the
> distro.
Correct - we see now value in it for us.
> Last I checked the only two distros I knew of with any sort
> of HPC focus are Scyld and Caos. Everything else is focused upon
> LAMP se
On Mon, 19 Nov 2007, Ashley Pittman wrote:
> On Sat, 2007-11-10 at 18:33 +1100, Chris Samuel wrote:
>
> This is an interesting proposition, you seem to imply it's easier
> to change distribution that it is to customise the one you are
> already using, I'm wondering what your reasoning for this is.
Ashley Pittman wrote:
On Sat, 2007-11-10 at 18:33 +1100, Chris Samuel wrote:
On Fri, 24 Aug 2007, Michael Will wrote:
The first advise is to stay away from redhat for file servers since
they have some bursty I/O bugs and don't support XFS.
We've run through Fedora (OK, but release cycle too qui
On Sat, 2007-11-10 at 18:33 +1100, Chris Samuel wrote:
> On Fri, 24 Aug 2007, Michael Will wrote:
> > The first advise is to stay away from redhat for file servers since
> > they have some bursty I/O bugs and don't support XFS.
>
> We've run through Fedora (OK, but release cycle too quick), Ubunt
I can assure you that here in the US I have made then
VERY aware of that fact. Since we have such a large
cluster they really want to move us into RHEL in the
worse way.
We use XFS under Fedora and contrained to using RHEL,
EXT3. When the client complains about performance, I
just tell them to
On Sun, 11 Nov 2007, Buccaneer for Hire. wrote:
> A stock kernel will from Redhat will not give you the performance
> you need.
Indeed, and the people I know at Red Hat in Australia are well aware
of my thoughts on their restrictive choice of filesystems.. :-)
cheers!
Chris
--
Christopher Sam
--- Chris Samuel <[EMAIL PROTECTED]> wrote:
> On Fri, 24 Aug 2007, Michael Will wrote:
>
> > I tested several NFS server configurations with a
> 19 node cluster.
>
> Same, but with 150+ nodes and about 500+ CPUs.
>
> > The first advise is to stay away from redhat for
> file servers since
> > t
On Fri, 24 Aug 2007, Michael Will wrote:
> I tested several NFS server configurations with a 19 node cluster.
Same, but with 150+ nodes and about 500+ CPUs.
> The first advise is to stay away from redhat for file servers since
> they have some bursty I/O bugs and don't support XFS.
Amen.
We've
bytes (2.1 GB) copied, 18.5161 seconds, 116 MB/s
What does it accomplish?
Thanks.
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Glen Dosey
Sent: Thursday, August 23, 2007 6:10 PM
To: [EMAIL PROTECTED]
Cc: Beowulf
Subject: Re: [Beowulf] Network
On Fri, 2007-08-24 at 19:23 +0200, Bogdan Costescu wrote:
> On Thu, 23 Aug 2007, Glen Dosey wrote:
>
> > What really gets me is that while my NFS reads are around ~50MB/s ,
> > the writes are basically at wire speed, slowing down to and holding
> > at about ~90MB/s when we exceed the 4GB file si
On Thu, 23 Aug 2007, Glen Dosey wrote:
What really gets me is that while my NFS reads are around ~50MB/s ,
the writes are basically at wire speed, slowing down to and holding
at about ~90MB/s when we exceed the 4GB file size. That would seem
to indicate to me the server has no problem dealing
On Fri, 2007-08-24 at 08:45 -0400, Lawrence Sorrillo wrote:
> I am uncertain what this step does
>
> now we unmount the NFS share, recreate the file on the server, and remount
> it to clear the client cache but leave it cached on the server
>
> [EMAIL PROTECTED] ~]# dd if=/mnt/array3/file.dd
Glen,
Are your NFS clients and servers networked with gigabit ethernet? If not,
just hit . If so, keep reading...
If your network can handle jumbo frames (MTU of 6000 or 9000, rather than
the standard MTU of 1500) then shifting to jumbo frames might improve your
NFS performance.
I have a
ity and typos)
>
> -Original Message-
> From: "Glen Dosey" <[EMAIL PROTECTED]>
> To: [EMAIL PROTECTED]
> Cc: "Jeff Blasius" <[EMAIL PROTECTED]>; "Beowulf"
> Sent: 8/23/2007 6:09 PM
> Subject: Re: [Beowulf] Network Filesystems perfor
Message-
From: "Glen Dosey" <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Cc: "Jeff Blasius" <[EMAIL PROTECTED]>; "Beowulf"
Sent: 8/23/2007 6:09 PM
Subject: Re: [Beowulf] Network Filesystems performance
On Thu, 2007-08-23 at 15:53 -0400, Joe Landman wrote:
On Thu, 2007-08-23 at 15:53 -0400, Joe Landman wrote:
> Since you indicated RHEL4, its possible that something in kernel is
> causing problems. RHEL4 is not known to be a speed demon.
All the current testing is on RHEL5 actually. 64bit . It offered better
performance than RHEL4. Everything in he
gured?
Michael
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
On Behalf Of Glen Dosey
Sent: Thursday, August 23, 2007 12:16 PM
To: Jeff Blasius
Cc: Beowulf
Subject: Re: [Beowulf] Network Filesystems performance
I am using the async flag. I'm also using the Infi
--
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
On Behalf Of Glen Dosey
Sent: Thursday, August 23, 2007 10:12 AM
To: Beowulf
Subject: Re: [Beowulf] Network Filesystems performance
Perhaps I should just ask the simple question. Does anyone on-list
obtain greater than 40 MB/s performance from
Glen Dosey wrote:
> I am using the async flag. I'm also using the Infiniband now instead of
> the GigE and I've increased my default TCP window up to 6MB. None of it
> makes a difference. It's still around 40MB/s.
BTW:
dstat, iftop, and atop are your friends.
http://dag.wieers.com/home-made/dst
Glen Dosey wrote:
> Perhaps I should just ask the simple question. Does anyone on-list
> obtain greater than 40 MB/s performance from their networked filesystems
> ( when the file is not already cached in the servers memory ) ?
Yes.
>
> (Yes it's a loaded question because if you answer affirmati
I am using the async flag. I'm also using the Infiniband now instead of
the GigE and I've increased my default TCP window up to 6MB. None of it
makes a difference. It's still around 40MB/s.
Thanks for the thought.
On Thu, 2007-08-23 at 15:11 -0400, Jeff Blasius wrote:
> Yes.
>
> Just a shot in
Perhaps I should just ask the simple question. Does anyone on-list
obtain greater than 40 MB/s performance from their networked filesystems
( when the file is not already cached in the servers memory ) ?
(Yes it's a loaded question because if you answer affirmatively, then I
know who to interrogat
Reply inline
On Tue, 2007-08-21 at 17:24 +1000, Chris Samuel wrote:
> On Tue, 21 Aug 2007, Glen Dosey wrote:
>
> > I've left out a lot of detail to keep this succinct. I can provide
> > it when necessary.
>
> I think the details you've omitted mean there's not enough information
> there to actu
On Tue, 21 Aug 2007, Glen Dosey wrote:
> I've left out a lot of detail to keep this succinct. I can provide
> it when necessary.
I think the details you've omitted mean there's not enough information
there to actually come to any conclusions! :-)
What distro on the clients and servers ?
What k
I have tried adjusting a number of kernel parameters to increase the TCP
window size but it has not had any affect.
you didn't mention any kernel versions, nfs block sizes, etc...
___
Beowulf mailing list, Beowulf@beowulf.org
To change your subscriptio
I am having issues with read performance on our cluster using NFS ( and
also SMB) . Any initial file transfer to the clients seems to be limited
to ~40MB/s. This is not a limitation of the underlying array, as I can
read the same file on the server at ~160MB/s. Similarly, after the file
has been re
26 matches
Mail list logo
