On Sun, 2008-05-11 at 16:01 -0400, Perry E. Metzger wrote:
> Who do you favor for console servers these days? Ditto for
> addressable/switchable PDUs?
>
I hope Joe doesn't mind me answering a question directed at him,
but for us if you spec a separate console server it would be a cyclades
Alterpa
In the amd case all socket F cpu's support ecc.
Does support for ECC on AMD 64 bit processors bypass BIOS support, for
there _is_ no memory without bios support - it's pretty interesting to
read the AMD bios writer's guide, and see how the bios has to bootstrap:
detecting kinds of memory, me
Quoting Joe Landman <[EMAIL PROTECTED]>, on Mon 12 May
2008 06:03:56 PM PDT:
Perry E. Metzger wrote:
Joe Landman <[EMAIL PROTECTED]> writes:
I've been reading spec sheets, and they often don't tell you, which is
rather annoying. Thus my question.
I just randomly selected 2 motherboards from
Joel Jaeggli wrote:
> if you're operating at the component level it's a question which core
> logic chipsets are in use.
>
> In the amd case all socket F cpu's support ecc.
Does support for ECC on AMD 64 bit processors bypass BIOS support, for
the most part? Let me rephrase: if I install an Opter
Gilad,
Gilad Shainer wrote:
My apologizes. I meant the MPI includes an option to collect several MPI
messages into one network message. For applications cases, sometimes it
helps with performance and sometimes it does not. OSU have shown both
cases, and every user can decide what works best for
Joe Landman <[EMAIL PROTECTED]> writes:
>> I've been reading spec sheets, and they often don't tell you, which is
>> rather annoying. Thus my question.
>
> I just randomly selected 2 motherboards from 2 different vendors and
> on both spec sheets, they clearly defined which memory they took.
Oh,
Joe Landman <[EMAIL PROTECTED]> writes:
> Server class boards generally do ECC. Desktop class generally do not.
>
> Spec sheets are your friends.
I've been reading spec sheets, and they often don't tell you, which is
rather annoying. Thus my question.
Perry
_
Perry E. Metzger wrote:
Given that one is doing very long computations, it seems obvious to me
that you're likely to get memory errors that alter your computations
if you only run long enough. ECC memory thus seems like a near
necessity these days.
However, I've been finding it increasingly diff
Perry E. Metzger wrote:
Joe Landman <[EMAIL PROTECTED]> writes:
I've been reading spec sheets, and they often don't tell you, which is
rather annoying. Thus my question.
I just randomly selected 2 motherboards from 2 different vendors and
on both spec sheets, they clearly defined which memory t
Greg Lindahl wrote:
On Mon, May 12, 2008 at 07:39:35PM -0400, Joe Landman wrote:
Server class boards generally do ECC. Desktop class generally do not.
My local gamer PC shop was easily able to find desktop mobos with ECC
for me; they have a summary of the spec sheets from somewhere. They'd
n
Perry E. Metzger wrote:
Joe Landman <[EMAIL PROTECTED]> writes:
Server class boards generally do ECC. Desktop class generally do not.
Spec sheets are your friends.
I've been reading spec sheets, and they often don't tell you, which is
rather annoying. Thus my question.
I just randomly sele
Perry E. Metzger wrote:
However, I've been finding it increasingly difficult to figure out,
based on manufacturers blurbs, whether or not given motherboards
support ECC memory properly. Anyone have any tips on how to determine
it easily?
The motherboard manual usually refers to a document that
On Mon, May 12, 2008 at 07:39:35PM -0400, Joe Landman wrote:
> Server class boards generally do ECC. Desktop class generally do not.
My local gamer PC shop was easily able to find desktop mobos with ECC
for me; they have a summary of the spec sheets from somewhere. They'd
never actually sold one
Perry E. Metzger wrote:
Given that one is doing very long computations, it seems obvious to me
that you're likely to get memory errors that alter your computations
if you only run long enough. ECC memory thus seems like a near
necessity these days.
However, I've been finding it increasingly diff
Perry E. Metzger wrote:
Mark Hahn <[EMAIL PROTECTED]> writes:
I find that with IPMI and console redirection, it's very rarely necessary to
care about where your nodes are, at least from a sysadmin perspective.
Speaking of IPMI, are there reasonable motherboards that incorporate
it right into t
Given that one is doing very long computations, it seems obvious to me
that you're likely to get memory errors that alter your computations
if you only run long enough. ECC memory thus seems like a near
necessity these days.
However, I've been finding it increasingly difficult to figure out,
base
At Edward Waters College in Jacksonville Fl., we went to Purdue to see their
beowulf cluster and we have a cluster setup now.
Bernard Chapple
CIO
Edward Waters College
Sent via BlackBerry from T-Mobile
-Original Message-
From: Karen Shaeffer <[EMAIL PROTECTED]>
Date: Sun, 11 May 2008 13
Alan Louis Scheinine <[EMAIL PROTECTED]> writes:
[Describing machines that had IPMI installed...]
> We installed 1U cases in a rack and plugged-in the machines
> but they were turned-off.
Were the power supplies physically turned off or were these machines
"soft off"? Modern machines have both sw
On Monday 12 May 2008 07:28:15 Karen Shaeffer wrote:
> On Sun, May 11, 2008 at 04:45:46PM -0400, Joe Landman wrote:
> > Karen Shaeffer wrote:
> > >Hi Joel,
> > >Yes, but a separate and distinct power distribution for that +5 volt
> > >supply will then need to be implemented on the system board to s
Karen Shaeffer wrote:
On Sun, May 11, 2008 at 12:42:18PM -0700, Joel Jaeggli wrote:
Karen Shaeffer wrote:
Hi,
To implement full IPMI capability on the system board, then you need
two separate and independent power distribution systems on the system
board I believe, which is why they implement I
Joe Landman <[EMAIL PROTECTED]> writes:
> We have found as a safety precaution, that including a console
> server/kvm unit and having power control via addressable/switchable
> PDU is a great backup, especially when we are hundreds of km (or
> simply different timezones) from the units.
Who do yo
Karen Shaeffer wrote:
On Sun, May 11, 2008 at 02:45:05AM -0400, Mark Hahn wrote:
I find that with IPMI and console redirection, it's very rarely necessary
to
care about where your nodes are, at least from a sysadmin perspective.
Speaking of IPMI, are there reasonable motherboards that incorpor
Mark Hahn <[EMAIL PROTECTED]> writes:
> I find that with IPMI and console redirection, it's very rarely necessary to
> care about where your nodes are, at least from a sysadmin perspective.
Speaking of IPMI, are there reasonable motherboards that incorporate
it right into the design at this point
On Mon, May 12, 2008 at 08:20:21AM -0500, Geoff Jacobs wrote:
>
> It's a separate feed from the ATX supply, and I can't see how anyone
> would be stupid enough to switch it on the motherboard. It should be
> simply railed around the motherboard wherever needed. Perhaps, just to
> be sure, I should
On May 12, 2008, at 1:11 PM, Reuti wrote:
Am 12.05.2008 um 18:01 schrieb Craig Tierney:
Reuti wrote:
Hiho,
Am 12.05.2008 um 15:14 schrieb Prentice Bisbal:
It's still an RFE in SGE to get any arbitrary combination of
resources, e.g. you need for one job 1 host with big I/O, 2
with huge me
Am 12.05.2008 um 18:01 schrieb Craig Tierney:
Reuti wrote:
Hiho,
Am 12.05.2008 um 15:14 schrieb Prentice Bisbal:
It's still an RFE in SGE to get any arbitrary combination of
resources, e.g. you need for one job 1 host with big I/O, 2
with huge memory and 3 "standard" type of nodes you could
Reuti wrote:
Hiho,
Am 12.05.2008 um 15:14 schrieb Prentice Bisbal:
At a previous job, I installed SGE for our cluster. At my current job
Torque is the queuing system of choice. I'm very familar with SGE, but
only have a cursory knowledge of Torque (installed it for evaluation,
and that's it).
Robert G. Brown wrote:
> On Sun, 11 May 2008, Joe Landman wrote:
>
> > Gerry Creager wrote:
> >> Student workers are cheap and expendable...
> >
> > cold ... ;)
>
> but true. In fact, it as little as three or four months they tend to
> expend, and there ya gotta go and train a new one. Sometim
Quoting Prentice Bisbal <[EMAIL PROTECTED]>, on Mon 12 May 2008
06:18:49 AM PDT:
Jim Lux wrote:
Actually, you can order cables already pre numbered and labelled. Why
burn expensive cluster assembler time when you can pay someone
(potentially offshore) to do it cheaper.
Because that would
Hiho,
Am 12.05.2008 um 15:14 schrieb Prentice Bisbal:
At a previous job, I installed SGE for our cluster. At my current
job
Torque is the queuing system of choice. I'm very familar with
SGE, but
only have a cursory knowledge of Torque (installed it for
evaluation,
and that's it). We're abo
Karen Shaeffer wrote:
> Hi Geoff,
> Following my line of thought, the issue is only whether the PCI rail(s)
> powering all the IPMI circuitry are completely isolated from the
> rest of the system board circuitry or not. If true, then it would
> be equivalent to a daughterboard implementation. If fa
Jim Lux wrote:
> Actually, you can order cables already pre numbered and labelled. Why
> burn expensive cluster assembler time when you can pay someone
> (potentially offshore) to do it cheaper.
Because that would hurt the US economy, and the labels would probably be
made out of lead. ;-)
--
Pr
Reuti wrote:
> Hi,
>
> Am 09.05.2008 um 20:26 schrieb Prentice Bisbal:
>
>> At a previous job, I installed SGE for our cluster. At my current job
>> Torque is the queuing system of choice. I'm very familar with SGE, but
>> only have a cursory knowledge of Torque (installed it for evaluation,
>> a
John Hearns wrote:
> On Fri, 2008-05-09 at 14:26 -0400, Prentice Bisbal wrote:
>
>> 1. Interactive shells managed by queuing system
>> 2. Counting licenses in use (done using a contributed shell script in SGE)
>> 3. Separation of roles between submit hosts, execution hosts, and
>> administration h
On May 9, 2008, at 2:26 PM, Prentice Bisbal wrote:
At a previous job, I installed SGE for our cluster. At my current job
Torque is the queuing system of choice. I'm very familar with SGE, but
only have a cursory knowledge of Torque (installed it for evaluation,
and that's it). We're about to pu
On Sun, 11 May 2008, Joe Landman wrote:
Gerry Creager wrote:
Student workers are cheap and expendable...
cold ... ;)
but true. In fact, it as little as three or four months they tend to
expend, and there ya gotta go and train a new one. Sometimes they last
as long as a couple of years, th
This thread has also been used to discuss IPMI.
We installed 1U cases in a rack and plugged-in the machines
but they were turned-off.
The air-conditioning installment was not finished, the rack doors
were open but the lower part had many cables that basically
covered the rear part of the nodes. T
Greg Lindahl wrote:
Last I saw someone doing this, IPMI sharing an ethernet port with the
host led to all kinds of weird ARP problems. Whereas a dedicated port
is much easier to configure. My favorite vendors all offer a
dedicated port...
Yes, same here - I've done some experimentation on a n
I was initially sceptical when I had to do the same thing as I had bad
experiences with shared addressing over the same interface... but with the
gear over the last five years I've never had a problem with this except with
the specific combination of FreeBSD and Dell boxes. The problem lied on t
On Sun, 2008-05-11 at 19:13 -0700, Greg Lindahl wrote:
> O
> Last I saw someone doing this, IPMI sharing an ethernet port with the
> host led to all kinds of weird ARP problems. Whereas a dedicated port
> is much easier to configure. My favorite vendors all offer a
> dedicated port...
Also port nu
40 matches
Mail list logo
