On Tue, Aug 13, 2013 at 06:07:53PM -0700, Christopher Samuel wrote:
>
> On 14/08/13 02:59, Andy Wettstein wrote:
>
> > If proctrack/cgroup is not being used, I don't think it is possible
> > to properly track a process that does this. Since proctrack/cgroup
> > can reliably track all PIDs for a
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 14/08/13 02:59, Andy Wettstein wrote:
> If proctrack/cgroup is not being used, I don't think it is possible
> to properly track a process that does this. Since proctrack/cgroup
> can reliably track all PIDs for a step, then I think it should be
>
Here is my config and an example perl script that will go over the
memory limit without being identified. It just creates an array and
fills it up, so the array size may need to be adjusted depending on the
memory configuration. If the call to daemonize() is commented, the job
should get killed due
Experimental meaning it doesn't work as correctly as the linux plugin
does. I know when we last worked on it the cgroup plugin did not do
memory accounting correctly. I also know there is quite a bit of
functionality missing as well, (profiling and such). Basically it is
half baked at this
Moe,
In what way is it experimental? Is it possibly unstable or just not
feature-complete?
We're writing a script to independently gather statistics for our own
database and would like to use the cpuacct cgroup, thus the interest in
the jobacct_gather/cgroup plugin.
Ryan
On 08/09/2013 1
Ok. When you say it should work do you mean there is a bug in slurm that
is causing this problem?
I can send a fairly trivial example that can bypass any memory limits if
you need it.
On Fri, Aug 09, 2013 at 09:07:53AM -0700, Moe Jette wrote:
>
> I misspoke. The JobAcctGatherType=jobacct_gather
I misspoke. The JobAcctGatherType=jobacct_gather/cgroup plugin is
experimental and not ready for use. Your configuration should work.
Quoting Moe Jette :
Your explanation seems likely. You probably want to change your
configuration to:
JobAcctGatherType=jobacct_gather/cgroup
Quoting Andy
Your explanation seems likely. You probably want to change your
configuration to:
JobAcctGatherType=jobacct_gather/cgroup
Quoting Andy Wettstein :
I understand this problem more fully now.
Certains jobs that our users run fork processes in a way that the parent
PID gets set to 1. The _get
I understand this problem more fully now.
Certains jobs that our users run fork processes in a way that the parent
PID gets set to 1. The _get_offspring_data function in
jobacct_gather/linux ignores these when adding up memory usage.
It seems like if proctrack/cgroup is enabled, the jobacct_gat