[Beowulf] Re: Help: Building a disked Beowulf / Chaitanya Krishna, Brown et al

[Ed Karns]

On Aug 25, 2005, at 3:39 AM, beowulf-request at beowulf.org wrote:

>> ... the resources that I have are these:
>>
>> 1 Intel Pentium 4 3 Ghz Procs   10
>> 2 Intel Mother boards                10
>> 3 200 GB SATA Hard disks       10
>> 4 120 GB IDE Hard disks           10
>> 5 Network cards                          10 + 1 (1 extra for master)
>> 6 Some already present switches
>>
>> All the IDE drives will be primary (the OS will reside on this) and
>> the SATA drives will be use as secondary drives for storage)
>>
>> My plan (and requirement) is the following:
>>
>> 1 To get the cluster up and running parallel jobs.
>> 2 The way I intend to do 1 is this. Install the OS (SuSE 9.3 Pro) on
>> the master and install bare-bones ( I am not sure, but may be  
>> something
>> like kernel, NFS and/or NIS, SSH, etc) on the rest of the nodes so
>> that I am able to run parallel jobs as well as serial jobs on the
>> nodes. Will require help on this.
>>
>
> Your hardware looks perfectly reasonable for a small cluster.   
> Let's hope that your NICs and switches "match" in some way --  
> enough ports, gigabit ports and gigabit cards, whatever.  One has  
> to wonder a bit about why the nodes have both a 120 GB IDE and 200  
> GB SATA drive instead of e.g. 2x[120,200] GB SATA only.  I've never  
> mixed drives like this and would expect that it works but would  
> worry that it might do something to performance (Mark Hahn usually  
> is the answer man as far as the overall IDE drive subsystem is  
> concerned:-).

* Hardware suggestions:

Take the "master" off of any alternate network until complete  
debugging of the cluster is accomplished (unplug it, at the least,  
and remove that "alternate" NIC if possible) ... set the whole thing  
up as a completely stand alone cluster until it works as required.

I would also wonder about the switches (10/100baseT = :>]  or Gigabit  
switches = :>) ?)  If two switches, then "balance" the loads = same  
number of CAT5 connections on each switch = 5 & 5 plus jumper, if  
three switches = 3 & 3 & 4 plus jumpers.

All BIOS configurations on all systems should closely match,  
especially I/O port configurations. All NIC (network cards) should  
match = brand name and model type where possible.

Although I do not speak from authority on this type of x86 cluster,  
my best guess to increase performance would be to use the SATA  
drivers completely for the OS and cluster work and use the IDE drives  
for mirroring and backup (the exact reverse of your  
configurations) ... keeping the IDE drives off of the cluster if  
possible ... or even remove the IDE drives from the systems and make  
a RAID array for the cluster network.
web-pages
* OS and software configurations = Trust Mr. Hearns', DGS' and Mr.  
Brown's suggestions.

 From John Hearns:
" ... You could do worse than to consult Robert Brown's web-pages,  
Google for Brahma and Duke University. Also get a copy of the OReilly  
book on Linux clustering, the latest one. ..."

 From DGS:
" ... You should look into some of the cluster "toolkits".  Free ones  
include OSCAR, ROCKS, Warewulf, and oneSIS.  My favorite is Warewulf,  
though ROCKS is probably the nearest you can get to a "cluster in a  
box" for free. ..."

 From Robert Brown:
" ... I'm assuming that the NICs are PXE-capable and that you've got  
a KVM setup that you can move from machine to machine somehow to set  
the BIOS and manage at least the initial install. ..."


Ed Karns
FireWireStuff.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.beowulf.org/pipermail/beowulf/attachments/20050825/670d50e5/attachment.html>