Work planned for cluster shutdown on 2013-01-15

shutdown plan

The following services will be shut down

  1. all compute nodes - possibly with the exception of "radiobenk" server (n1663, ...n1675)
  2. all gpu nodes
  3. all thumpers except s12 and s04 (disk is still be resilvered)
  4. all data servers d01..d37
  5. HSM clients
  6. mds010 and mds020
  7. Shall we shutdown 6780 and DDN?
  8. Power down racks and not needed LCPs

action items

Work planned, please rank if necessary

  • Update baseline firmware on 6780, when?
  • do we have a newer DDN firmware to apply?
  • install new PDU and power strips in racks
  • let electricians change power lines for special racks
  • change fibre channel cards from mds010 to mds020, also change 10G NIC (change FAI/DHCP)
  • move mds020 to spot where mds010 is now
  • move einstein-wug to einstein-rack?
  • move einstein-db to einstein-rack?

Special things to prepare

  • Need long power cables, especially power cables with Schuko-C14 to keep some machines running while power is being worked on
    • need detailed plan, which machines need this!

-- CarstenAulbert - 10 Jan 2013
Topic revision: r1 - 10 Jan 2013, CarstenAulbert
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback