Five

  • Today was SciCat. Which was great, but I, once again went off to talk to somebody else :(

    • SciCat fiddling to come later.
  • DST - Data Systems and Technologies

    • Jesper Rude Selknaes (mailto:Jesper.RudeSelknaes@esss.se)

    • 2018 became group leader

    • Bought a machine with 2 A100 machines. Waiting for scientists to justify their use.

    • Provide the infrastructure for scientific computing (compute and network).

      • HPC
        • Slurm cluster (4000 cores)
        • ZFS (0.5PB slow storage)
        • Two 850TB systems, scalable gpfs (spectrum scaler)
      • OpenStack
      • 1Gb/10Gb/25Gb/100Gb/200Gb HDR Infiniband.
      • 40Gb is bad because the lane speed doesn’t match.
      • Puppet / Foreman is used for deploy/CM
      • When you write an HDF5 file you write some blocks, then meta data, then blocks, then meta data, etc.
        • Might have to work on a driver for hdf5 stuff as the way IO is handled is sub optimal?
      • History
        • When they started the ESS stood up their own AD
        • When DMSC started they stood up an LDAP
        • They are working on merging. Hopefully it’ll go smoothly.
      • ESS has 650-700 workers
      • The plan now, how risk is mitigated, and the plan in general is much more credible.
      • “we would like to put as much of the infrastructure into the same place as possible”
    • DRAM - Data Reduction and Modeling

  • ECDC - Experimental Control and Data Curation

  • ICS - Instrument Control Systems(? didn’t get to talk about this, so could be wrong.)

  • Cool Tools

  • Thoughts

  • Max 4

    • SciDog
    • SciFish
    • (There are many Sci) This should be enhanced.