With your previous suggestions 8 months ago we bought a Tyan S4881 server
with 8 dual-core Opteron CPUs with 64GB RAM. Now we will buy new ones (2
more for the time being) and we will eventually planning to form a cluster
from these servers, which will have at most 8 boxes. Now, as you guess, the
Le dimanche 30 juillet 2006 00:53, Robert G. Brown a écrit :
[...snip...]
> Y'know, except for actually going to a classroom and teaching, I could
> do most of the rest of what I do sitting right where I'm sitting now, in
> between trips down to the lake in search of dinner...
>
> rgb
We all k
As Joe mention, the way we handle this is by using cluster schedulers
sitting on robust hardware platforms that are capable of handling
large numbers of job submissions without problems. Grid Engine and
Platform LSF are two capable products that come to mind and scale well.
The fact that
Rather than banning them or publishing behaviour, we simply disable
their access to the queue and won't re-instate it until they have
contacted us to indicate they have rectified their ways.
eg. rather than enforce hard disk quotas, which may result in a job
loosing data, we simply detect w
Diego M. Vadell wrote:
Maybe you can collect some logs and make a list of misbehaving users. Then
you can warn the users that you are collecting that information, that
bypassing the system will degrade the performance for everybody, and that
you may post that information. Sometimes the perspec
> Hi Jerry:
>> the other example is that use system call and ssh
>> to each node and run stuff and bypass the scheduler...
Torque 2.1.2 has just been released. It comes with a pam module that, if I
understood it right, makes harder (though not impossible) for users to
bypass the batch system. The
Hi, Thanks, Joe.
I am not meaning to "ban" anything immediately, I am just curious how often
this happen to the HPC community.
Perl/shell is really strong tool, one example is to use loop to submit huge
mount of jobs and puts burden on scheduler server, the other example is to have
one job sit idl
Hi, all:
We are going to build new cluster and we are thinking about various solutions
for file I/O. We have lots embarrassed parallel applications that read huge data
(giga byte) only in the beginning and write only at the end concurrently, and
each node run job independently. What is the bes
Hi Jerry:
Xu, Jerry wrote:
Hi, Thanks, Joe.
I am not meaning to "ban" anything immediately, I am just curious how often
this happen to the HPC community.
Perl/shell is really strong tool, one example is to use loop to submit huge
mount of jobs and puts burden on scheduler server,
Thats what t
Hi Jerry:
Its generally a good idea to talk to your users, understand what it
is they are doing, and see if you can help them, rather than simply
"banning" things. The result of bans of deeply embedded practices
usually results in some ... exciting ... meetings, emails, and telephone
calls
Hi, I am maintaining a cluster while lots user uses perl to submit tons of jobs
which seems to me like abusing the system.
Does everybody meet the same situation? Many user us system call in the perl to
do "qsub", shall I ban this? I don't know exactly why it is bad, but it looks to
me really bad
Dear Sir, I want to use Oscar software that include the MPICH2, or upgrade my Oscar MPICH to MPICH2 , but I dont khnow how . Please give me some helps about upgrading oscar libraries or Oscar verjen that include MPICH2. Best Regards.**Reza MiraniElectronic department Iran A
Hi All,
With your previous suggestions 8 months ago we bought a Tyan S4881 server
with 8 dual-core Opteron CPUs with 64GB RAM. Now we will buy new ones (2
more for the time being) and we will eventually planning to form a cluster
from these servers, which will have at most 8 boxes. Now, as you gues
13 matches
Mail list logo
