[ale] Cluster hardware fault tracking; Latest kernel development

[Dow Hurst]

We are no longer officially in the SGI camp anymore as the software base 
is moving toward Linux and the performance in our price range is too.  I 
am getting a Linux workstation from Dell soon.  Plus, my boss is 
interested in a Linux cluster in the future for our next move on 
hardware.  I want to determine if a self built cluster is really worth 
it compared to a vendor product that is prebuilt and preconfigured.  The 
amount of self checking the cluster does, and the ease of 
upgrading/expanding is really important.  SGI has the best of all worlds 
in the Origin 2000 and now 3000 series for that stuff.  I just didn't 
know if anyone has already researched this topic and had some thoughts 
or links.  I know the cost factor of a cluster is way better than the 
usual big iron but the management of hardware and software can be a big 
load.
Dow


runman at speedfactory.net wrote:

>I have experience in installing OpenBSD on a sparc SunBlade 100.  No problems, except for X (probably due to the crappy Gateway monitor).  I know that FreeBSD and NetBSD have 64 bit offerings.  The only stuff I have seen for Linux on a 64 bit Sun box was/is unstable and highly developmental, though I have not been following it really closely.
>
>  I cannot find the Suse 64 bit stuff (I guess that they are trying to only sell it to big money companies) but it is out there and is advertised on the AMD site.  Even MS is developing a 64 bit MS Server offering (yeah right - a 2 bit system configured for 32 bit processors).  I would stay away from Itanium.  It bombed when it came out and I think it will bomb again.  I would try for Opteron and AMD offerings, though I know Dow you are firmly in the SGI camp.  I looked at Linux on the SGI and it was just in development (however that was 6-7 months ago - before you came through with some OS stuff for FreeBytes).  I guess Google will be your friend here.
>
>Greg 
>
>
>  
>
>>I would like to know some links to info on building a cluster but from
>>the management viewpoint of integrating software to pinpoint hardware
>>problems.  Is there any opensource software out there that looks at the
>>hardware?  I can see if you get the right mix of hardware you could use
>>sensors and info in /proc  and also logcheck to watch what is happening
>>on a machine.  I was wondering if people have integrated this into a
>>cluster monitor package.  I know IBM stresses this as a selling point on
>>their clusters since that is part of the package you get.  I am sure
>>that SGI's Altix cluster is similar.  I am thinking of commodity self
>>built clusters and open source cluster software.
>>
>>I've been looking at the development kernels and noticed that IBM and
>>SGI are contributing quite a bit to the 64bit code for the Itanium2 and
>>Opteron processors.  At least that what it seems to me, however, I
>>haven't much experience with reading kernel changelogs.  I noticed that
>>the MIPs code seemed to be active from the SGI contributions and that it
>>might be possible to compile a kernel for an SGI Indigo, Origin, or O2
>>now.  Does anyone have any experience with the procedure for getting a
>>basic Linux install running on different architecture than x86?  I think
>>it would be quite interesting to take one of the SGI O2s I have here and
>>try to get Linux running on it.
>>Thanks for your advice!
>>Dow
>>
>>
>>
>>--
>>__________________________________________________________
>>Dow Hurst                  Office: 770-499-3428
>>Systems Support Specialist    Fax: 770-423-6744
>>1000 Chastain Rd. Bldg. 12
>>Chemistry Department SC428  Email:   dhurst at kennesaw.edu
>>Kennesaw State University         Dow.Hurst at mindspring.com
>>Kennesaw, GA 30144
>>*****************************************************************
>>This message (including any attachments) contains confidential  *
>>information intended for a specific individual and purpose,     *
>>and is protected by law.  If you are not the intended recipient,*
>>you should delete this message and are hereby notified that     *
>>any disclosure, copying, or distribution of this message, or    *
>>the taking of any action based on it, is strictly prohibited.   *
>>*****************************************************************
>>
>>
>>_______________________________________________
>>Ale mailing list
>>Ale at ale.org
>>http://www.ale.org/mailman/listinfo/ale
>>
>>    
>>
>
>_______________________________________________
>Ale mailing list
>Ale at ale.org
>http://www.ale.org/mailman/listinfo/ale
>
>  
>

-- 
__________________________________________________________
Dow Hurst                  Office: 770-499-3428
Systems Support Specialist    Fax: 770-423-6744
1000 Chastain Rd. Bldg. 12
Chemistry Department SC428  Email:   dhurst at kennesaw.edu
Kennesaw State University         Dow.Hurst at mindspring.com
Kennesaw, GA 30144
*****************************************************************
This message (including any attachments) contains confidential  *
information intended for a specific individual and purpose,     *
and is protected by law.  If you are not the intended recipient,*
you should delete this message and are hereby notified that     *
any disclosure, copying, or distribution of this message, or    *
the taking of any action based on it, is strictly prohibited.   *
*****************************************************************


_______________________________________________
Ale mailing list
Ale at ale.org
http://www.ale.org/mailman/listinfo/ale