[slurm-users] practical tips to budget cluster expansion for a research center with heterogeneous workloads?

[Graziano D'Innocenzo]

Dear all, thanks everybody for your comments.
XDMOD looks definitely like the way to go. We will see to get that deployed.
Thanks again for your help,

--
Graziano D'Innocenzo (PGP key: 9213BE46)
Systems Administrator - ADAPT Centre

On Thu, Mar 21, 2019 at 5:16 PM Alex Chekholko <alex at calicolabs.com> wrote:
>
> Hi Noam,
>
> Right, xdmod is a standard LAMP stack webapp.  You can see some pictures of the graphs in the web interface in a google image search here
> https://www.google.com/search?q=xdmod&source=lnms&tbm=isch&sa=X
>
> It may also require a fairly beefy database backend, depending on how many millions of jobs you want to process.
>
> I'm not aware of command-line tools that produce pretty graphs suitable for consumption by upper management :)
>
> Regards,
> Alex
>
> On Thu, Mar 21, 2019 at 10:03 AM Noam Bernstein <noam.bernstein at nrl.navy.mil> wrote:
>>
>> On Mar 21, 2019, at 12:38 PM, Alex Chekholko <alex at calicolabs.com> wrote:
>>
>> Hey Graziano,
>>
>> To make your decision more "data-driven", you can pipe your SLURM accounting logs into a tool like XDMOD which will make you pie charts of usage by  user, group, job, gres, etc.
>>
>> https://open.xdmod.org/8.0/index.html
>>
>>
>> Am I correct that xdmod requires a running web server?  It appears to require apache as a software prerequisite.  If so, does anyone know of any similar tools that do not require running a web server?
>>
>> thanks,
>> Noam