Fedora cluster project? (was Re: [Beowulf] Opteron/Athlon Clustering)

Anand Vaidya anandv at singnet.com.sg
Wed Jun 9 20:57:33 PDT 2004


[Disclaimer: I work for a company that is IBM LS business partner]

I would suggest that you take a look at how IBM is implementing xCAT (Extreme 
Cluster Admin Toolkit http://www.xcat.org/ ) 

- They let you pick the base OS - RH7.x, RH8.x, RHEL, SL8/9 SLES etc, so your 
base OS is still vanilla RH/SL etc. so you can get support from them too, 
unlike ROCKS where the mailing list is your only saviour...

- Some progress towards Fedora, Debian

- I am requesting for Whitebox Linux and caOS integration.

- Available for IA32, AMD64, IA64 (probably PPC soon?)

- All other software (MPI, GM, PBS etc) are compiled/installed via src code, 

- They have deployed several large (900+ nodes) clusters with this.

Some features include mutliple headnodes, multiple package src nodes, pshell 
etc

The cluster kit (actually a bunch of scripts) sits on top of the OS, no need 
for any OS hacks like ROCKS. So those who wish to run free distro like Fedora 
can do that,. or for the risk averse/prorietary locked-in people, they can 
choose RHEL/SLES etc

IBM moves features from xcat into their proprietary CSM software, so if you 
need commercial support from IBM, you should be still able to buy CSM later, 
there are ways to migrate from xCAT to CSM.

The only sad part is it is not OpenSource as of now (== BSD/LGPL/GPL), but the 
key developer is working on it. BTW, The current licensing allows you to 
deploy xCAT on any hardware as long as you have some IBM hardware...

I found the idea very good..

Regards,
Anand




On Wednesday 09 June 2004 20:36, Mitchell Skinner wrote:
> On Tue, 2004-06-08 at 15:00, Robert G. Brown wrote:
> > <rant> Ah, but you see, this is a religious issue for me for reasons of
> > long-term scalability and maintainability, so I don't even think of this
> > alternative.  Or if you like, I think it costs money in the short run,
> > and costs even more money in the long run, compared to participating in
> > Fedora or CaOSity and doing it once THERE, where everybody can share it.
> > But you knew that...;-)
>
> I've been kicking around the idea of starting a fedora-oriented cluster
> distribution.
>
> ***Advantages of Fedora vs:
>
> RHEL rebuilds (Rocks, cAos)
> *  Don't need to do release engineering for the base distribution (cf.
> Rocks 3.1.0 bug where i686 kernels were installed on athlon machines)
> *  More up to date
>
> Debian
> *  More stable than unstable, less archaic than stable
>
> Rocks
> *  freely available update stream
>
> ***My goals for it would be:
> Social/political goals:
> 1. Ease of installation ("yum install cluster-master")
> 2. piggyback on other work (Fedora release engineering, mobs of people
> trying it out on commodity hardware)
> 3. Encourage outside contributions (have a completely open devel
> process, use a license without an advertising clause)
> 4. Be an integration point for applications ("yum install mpiblast")
> 5. Feed back upstream (to fedora and/or directly to maintainers)
>
> Technical goals/hopes:
> 1. Organize as a set of add-on packages, rather than a whole
> distribution (like OSCAR, but without the extra complexity of multiple
> base distributions).  This means creating SRPMs that can be fed upstream
> (unlike rocks-sge, for example).
> 2. Use RPM/anaconda to select architecture-specific files, like Rocks
> (handles heterogenous clusters more cleanly than systemimager (OSCAR) or
> network booting (warewulf))
>
> ***Potential Objections:
> 1.  "Fedora changes too frequently" - This is problematic in proportion
> to the pain of change.  One reason that change is painful is that people
> put it off, and then have to make a huge change all at once.  More
> frequent, more incremental changes can work, especially if you have the
> source to your apps.  This is assuming that the still-somewhat-untested
> fedora-legacy project doesn't work out; if it does then this objection
> is moot.  OTOH, if you want your closed-source ISV apps to be certified
> for your setup, then maybe this approach is not for you.
>
> What do people think?
>
> Mitch
>
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list