[Bioclusters] LSF 4.2 w/ Redhat 7.2 (2.4.X kernel with glibc2
.2) install problem
Alan Hart
bioclusters@bioinformatics.org
Thu, 2 May 2002 09:05:14 +0100
We'd like to pitch in here as we have a few years of LSF experience and now
a year or so of GE (solaris & linux).
Some (brief) humble opinions of ours as below ...
Enjoying the dialogue on comparisons ....
Regards
Alan Hart / Lee Scott
UNIX Team
Oxford GlycoSciences
> -----Original Message-----
> From: Ron Chen [SMTP:ron_chen_123@yahoo.com]
> Sent: 02 May 2002 02:55
> To: bioclusters@bioinformatics.org
> Subject: Re: [Bioclusters] LSF 4.2 w/ Redhat 7.2 (2.4.X kernel with
> glibc2.2) install problem
>
> Hi Kristian,
>
> Can you answer my questions about LSF vs SGE?
>
> 1) which one do you think is easier to install/use?
>
We dont think there is a big deal here - both straight forward
enough to install and maintain.
> 2) does SGE really have lots of admin overhead, and in
> which area?
>
Perhaps - more complex in areas where you want to emulate
functionality that is default in LSF - eg preemption.
Again we dont feel that there is too much of a difference here.
> 3) which features do you think is missing in SGE that
> you think is important?
>
Preemption - the facility for jobs active within a queue with lower
prioirty to be suspended by work running within higher priority queues. GE
handles this with prioiritisation within queue which isnt quite the same.
The LSF model dynamically allows jobs at lower prioirty to be suspended on
the system allowing higher prioirty work to go through. Ideally we would
like to see a product that accommodates both methodologies.
We have also found GE to be a little "buggy" during admin - eg
updating config files etc. Need a bit more scripting around GE to make it
do what you want.
Job accounting and history is poor in GE - superior in LSF.
GE is free - LSF certainly isnt !
We've had problems with master to shadow daemon failover on GE which
works fine on LSF.
We also remember working around the dual interface issue with our
LSF install - but same issues with GE ?
> (I hope to improve SGE so that people who don't have
> much money can get a free, powerful batch system)
>
> If it is commerical support issues, which most
> open-source software is having, SGE has Sun providing
> support for Solaris, and Sun partners providing
> support for other OSes--
>
> http://wwws.sun.com/software/gridware/partners/index.html
>
> I know a lot of people upset with PBS switch to SGE,
> and in some cases, LSF users switching to SGE as well,
> so I really want to understand what makes a free,
> opensource software less attractive, and the way to
> improve it.
>
> -Ron
>
> --- Kristian Vlahovicek <kristian@icgeb.trieste.it>
> wrote:
> > Hey Chris,
> >
> > We are currently in a process of evaluating LSF on
> > our 16x1.3GHz Athlons
> > running RH 7.2 and glibc 2.2. We have been running
> > SGE for some time now
> > and it was quite sufficient for what we needed, but
> > we got the possibility
> > to buy LSF at a very reasonable price so we decided
> > to give it a try :)
> >
> > WRT the installation, it went smoothly, no problems
> > with glibc. However,
> > we did not manage to get the daemons to bind to a
> > correct interface on
> > master. We migt be missing something (anyone with
> > more experience with LSF
> > administration is welcome to suggest the solution),
> > or could this be the
> > glibc incompatibility? OTOH, workers seem to be
> > working quite fine...
> >
> > basically, our master has 2 interfaces, eth0
> > (external net - public IP)
> > and eth1 (cluster net - private IP), each named
> > differently. when doing
> > hostsetup after install, it queries the hostname on
> > the master and gets
> > the public one and a) either it complains about host
> > not being listed in
> > the servers list (in case we put the private iface
> > name) or b) sets
> > everything up without complaints, but then master
> > sees only itself and the
> > rest of the cluster see each other but not the
> > master. (lshosts and
> > bhosts). Any ideas? As I said, we just started the
> > eval, so there might be
> > a trivial solution to this burried somewhere in the
> > documentation, but the
> > out-of-the-box thing does not work. Tried to see if
> > there were any
> > switches to the lim daemon to direct it to the
> > proper interface, but it
> > seems not...
> >
> > Kristian
> >
> > --
> > Dr. sc. Kristian Vlahovicek
> > Protein structure & bioinformatics Phone: +39
> > 0403757354
> > ICGEB Fax: +39
> > 040226555
> > Padriciano 99 E-mail:
> > kristian@icgeb.trieste.it
> > Trieste 34012 Italy Web:
> > http://www.icgeb.trieste.it/dna
> >
> >
> > On Tue, 30 Apr 2002, Chris Dagdigian wrote:
> >
> > >
> > > Hey folks,
> > >
> > > I've got a Platform LSF 4.2 distro that is bombing
> > out during the
> > > install process. I've tracked the problem down to
> > the install script not
> > > being able to determine the correct set of system
> > binaries it is
> > > supposed to install.
> > >
> > > The cluster is up-to-date but not bleeding-edge.
> > I'm using a mostly
> > > stock 2.4.18 kernel compiled from the official
> > Redhat Rawhide 2.4.18 rpm
> > > series. All the systems have all the latest update
> > RPM's installed.
> > >
> > > Reading the Platform website shows that they only
> > claim to be supporting
> > > 2.4 kernels with glibc2.1 which is disconcerting
> > because my systems
> > > here all have glibc2.2.4-22. They also only have
> > binaries for
> > > linux-2.4-glibc2.1 on their FTP site.
> > >
> > > I don't consider glibc2.2 all that new or exotic
> > and I don't want to
> > > downgrade so I figured I'd ask this list at the
> > same time I put in a
> > > query to Platform engineering...
> > >
> > > Anyone running LSF on linux clusters with
> > glibc2.2? Anyone else have
> > > trouble getting LSF 4.2 to install out of the box
> > on a updated RH7.2
> > > platform? I think the installer scripts might be a
> > bit flaky but the
> > > binaries seem fine.
> > >
> > > -Chris
> > >
> > > _______________________________________________
> > > Bioclusters maillist -
> > Bioclusters@bioinformatics.org
> > >
> >
> http://bioinformatics.org/mailman/listinfo/bioclusters
> > >
> >
> >
> > _______________________________________________
> > Bioclusters maillist -
> > Bioclusters@bioinformatics.org
> >
> http://bioinformatics.org/mailman/listinfo/bioclusters
>
>
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Health - your guide to health and wellness
> http://health.yahoo.com
> _______________________________________________
> Bioclusters maillist - Bioclusters@bioinformatics.org
> http://bioinformatics.org/mailman/listinfo/bioclusters
**********************************************************************
The information transmitted by this email is private and
confidential and is intended for the use of the intended
recipients specified therein.
If you are neither an intended recipient nor an employee
or agent responsible for delivery to an intended recipient,
you should be aware that any dissemination, distribution
or copying of this communication is strictly prohibited.
If you received this communication in error, please
notify us immediately.
**********************************************************************