On 15/07/15 08:25, Michael Gutteridge wrote: > I'll put in a plug for Slurm. I'd say it's got the features you need > and it scales up really well. > > Accounting: it can write to file, but you're best off configuring the > database backend (MySQL, Mariadb). Then there's both the built-in sacct > tool and (if you're so inclined) queries directly against the database. > > It can also be set up to track job resource usage.
+lots for Slurm. If you build it with HDF5 support (we've not tried this yet) then it can apparently collect time series profile data on a variety of areas for you that would not be practical to store in the slurmdbd backend. That can be exported to CSV if you want with sh5util. http://slurm.schedmd.com/hdf5_profile_user_guide.html There's a presentation on it from the Slurm User Group 2013 here: http://slurm.schedmd.com/SUG13/profile_hdf5.pdf cheers! Chris -- Christopher Samuel Senior Systems Administrator VLSCI - Victorian Life Sciences Computation Initiative Email: [email protected] Phone: +61 (0)3 903 55545 http://www.vlsci.org.au/ http://twitter.com/vlsci _______________________________________________ Beowulf mailing list, [email protected] sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
