PP-719: Enhance setUp in PTL specifically for Cray platforms

Community discussion is located here: http://community.pbspro.org/t/pp-719-enhance-setup-in-ptl-specifically-for-cray-platforms/463


Overview:

PBSTestSuite.setUp() calls several functions. The changes made in the settings in these functions are often not desirable on Cray because the MoM, scheduler, and server settings get reverted to plain Linux out-of-box configurations.


Out-of-box configuration of PBS on Cray:

  • PBS is installed and configured on Cray which result in specific settings. These initial settings are referred to as the out-of-box settings below.  These settings should also apply to the Cray ALPS simulator when PBS is properly configured.  These are settings for a freshly installed and configured PBS on Cray or Cray ALPS simulator.

  • As of current support for Cray CLE 5.2 and 6.0, below are the out-of-box configuration settings per PBS daemon:

1)  MoM

    • These lines should appear in PBS_HOME/mom_priv/config:
    •         $vnodedef_additive 0
    •         $alps_client <path to ALPS API>
    •         $usecp *:<Absolute path to /home directory>  <Absolute path to /home directory>

      Notes: $usecp setting above is used to enable local copy of job output of jobs that ran on compute nodes.
      $vnodedef_additive 0 will allow MoM to tell the server that there are vnodes missing, which the server will mark as stale.

2) Scheduler

    • PBS_HOME/sched_priv/sched_config will have 'vntype' in the resources line.

3) Server

    • Based on what is returned by the ALPS inventory :
      • vnodes representing the compute nodes will get created
      • the Cray-specific custom resources (e.g. PBScraynid, PBScrayhost, etc) will get created in PBS_HOME/server_priv/resourcedef.
      • the appropriate Cray-specific custom resources will be added to the vnodes.
      • since the mom config $vnode_per_numa_node is unset by default, there will be no PBScrayseg attribute on the vnodes representing the compute nodes.
    • Server settings :
      • flatuid = true
      • scheduling = true

      • the default queue is workq


Interface:  MoM.__init__() 

  • Visibility: Public
  • Change Control: Stable
  • Synopsis: Modifications to include initial MoM config settings for Cray.
  • Standing of the interface : modified interface
  • Details
    • if on real Cray or Cray ALPS simulator, then initialize the PBS_HOME/mom_priv/config to have these:
    • $vnodedef_additive 0

    • $alps_client <path to ALPS API>

    •                $usecp *:<Absolute path to /home directory>  <Absolute path to /home directory>


Interface:  Scheduler.revert_to_defaults() 

  • Visibility: Public
  • Change Control: Stable
  • Synopsis: Modifications to include 'vntype' in sched_config out-of-box settings for Cray.
  • Standing of the interface : modified interface
  • Details
    • After copying the out-of-box scheduler configuration in PBS_EXEC/etc/pbs_sched_config to PBS_HOME/sched_priv/sched_config :
      • If platform is a real Cray or Cray ALPS simulator, then add "vntype" to resources line of PBS_HOME/sched_priv/sched_config.


Interface:  PBSTestSuite.revert_mom()

  • Visibility: Public
  • Change Control: Stable
  • Synopsis: Modifications to have additional steps of deleting and creating nodes back, for retaining Cray specific resources.
  • Standing of the interface : modified interface
  • Details
    • If on a Cray or Cray ALPS simulator then perform below:
      • Delete all nodes and create them back.


Interface: Server.revert_to_defaults()

  • Visibility: Public
  • Change Control: Stable
  • Synopsis: Modifications to hooks settings for Cray.
  • Standing of the interface : modified interface
  • Details
    • If on a Cray or Cray ALPS simulator then perform below:
      • Restore the default 'PBS_translate_mpp' hook by:
        • Copying PBS_EXEC/lib/python/altair/pbs_hooks/PBS_translate_mpp.HK to PBS_HOME/server_priv/hooks/PBS_translate_mpp.HK
          • In this way, the shipped version (i.e. default version) of the hook will be in PBS_HOME for any PTL test to enable if desired.