I have a couple of Perl packages available on CPAN
Slurm::Sacctmgr
Slurm::Sshare
to provide wrappers around the sacctmgr and sshare
commands, respectively.
These will parse the output of query subcommands (
e.g. sacctmgr show XXX) and present as Perl-ish data structures,
as
Sacctmgr usage can be a bit tricky. Although there are "user", "account",
etc. entities that can be manipulated with sacctmgr, much of the time
you are dealing with "association" entities. And although sacctmgr will
allow you to view associations, modifying associations is handled by
On Tue, 2 Aug 2016, Joshua Baker-LePain wrote:
On Mon, 1 Aug 2016 at 1:16pm, Thomas M. Payerle wrote
So jobs of "paying" customers can cut ahead of jobs of "non-paying" and
"scavenger" in the queue, and will bump "scavenger" jobs. They will
If you decide to go the single partition model, you can use the
"Weight" parameter in slurm.conf to cause the standard nodes
to be preferentially used to the high-mem and GPU nodes. So jobs
only end up on high-mem or GPU nodes if they requested a lot of
memory or a GPU, or if the cluster is
I also have a bunch of Perl packages for wrappers around sshare,
and sacctmgr on CPAN.
The sshare package comes with an "sbalance" script which presents
the sshare info in an user friendly format and will work with common
accounting setups.
See
Yes, I agree with you. The TRES stuff is nice, but seems to be
added somewhat quickly, and in particular I do not see anyway to
view that information. I am looking forward to applying your sshare
patch. (I attempted once to write a patch similar to yours but had
issues I could not resolve
Something like
sacctmgr show user withassoc -p user=USERNAME --noheader | awk -F'|' '{ print
$6}'
should give you what you want.
On Wed, 12 Oct 2016, Eggleston, Nicholas J. wrote:
I?ve got kind of a weird question for you all. I?m in a place where I need to
know a list of all the
I've not used it, but when looking at srun man page (for something else)
notices a -r / --relative option which sounds like it might be what you
are looking for.
On Thu, 8 Dec 2016, Nigella Sanders wrote:
Hi Michael,
Actually, that was my first try but it didn't work. ?
srun finds it
done
wait
Setting -r2 will confine al the tasks in torus6003, as expected.
Regards,
Nigella
2016-12-08 15:27 GMT+01:00 Thomas M. Payerle <paye...@umd.edu>:
I've not used it, but when looking at srun man page (for something else)
notices a -r / --relative option which sounds
I am unaware of anything analogous to 'uniqueuser' but cleanup
is not that bad.
We have prolog and epilog scripts to handle it.
The prolog script creates a "jobtrack" directory if not already
there, and creates a file named after jobid with uid of user running
the job (e.g.
echo $SLURM_JOB_UID
On 2017-04-05 16:00, maviko.wag...@fau.de wrote:
Hello Dani and Thomas,
[ ...]
However i think i did not specify clear enough what my cluster looks
like and what i'm trying to achieve.
Compared to a regular HPC-Cluster my testing cluster consists of as
little as 5 nodes (each having the
You can define nodes with "features", and then at job submission time
require specific features.
E.g., if you had some nodes with ethernet only, some with QDR infiniband,
and some with FDR ib, you could define the QDR and FDR nodes to have
the feature qdr or fdr, respectively. Then, e.g. a
Basically, you create an allocation account for each of the
organizations and projects, with the projects having as "parent"
the allocation account corresponding to the organization it
belongs to. Then set the appropriate limits on the allocation
accounts, and create associations for users
Also, I believe the
#SBATCH ...
flags in a script must come before any non-blank/non-comment
lines in the script file in order to be honored by Slurm.
So something like
#!/bin/bash
WDIR=$PWD
#SBATCH -t 1:00
the -t 1:00 will get ignored by sbatch
On Thu, 29 Jun 2017, Lachlan Musicman wrote:
Your confusion seems to be stemming from misunderstanding job arrays, not the
--node and --ntasks-per-node options,
Job arrays are basically a short cut for submitting large numbers of similar
jobs
at once. The "sbatch array.sub" command basically submits 4 jobs, each job
being
allocated 1
On Tue, 3 Oct 2017, Christopher Samuel wrote:
On 29/09/17 06:34, Jacob Chappell wrote:
Hi all. The slurm.conf documentation says that if decayed usage is
disabled, then PriorityUsageResetPeriod must be set to some value. Is
this really true? What is the technical reason for this requirement
I think Uwe was on the right track.
It looks to me like the problem node is somehow thinking
TmpFS=/tmp rather than /local/scratch.
That seems to be consistent with what is being reported
(TmpDisk = 500 ).
I would check the slurm.conf/scontrol show config output
on the problem node and
On Wed, 4 Oct 2017, Mike Cammilleri wrote:
Hi Everyone,
I'm in search of a best practice for setting up Environment Modules for our
Slurm 16.05.6 installation (we have not had the time to upgrade to 17.02 yet).
We're a small group and had no explicit need for this in the beginning, but as
18 matches
Mail list logo
