Jump to: navigation, search

PTP/meetings/scalability/20120209

< PTP‎ | meetings

February 9, 2012 Scalability Meeting

Date: February 9, 2012 Time: 10:00 AM EST

Call in information:

Toll Free (US): 888-426-6840
Toll Free (Germany): 0800-000-1018
Toll Free (UK): 0808-234-5071
Passcode: 2221402

Attendees

  • Greg Watson, IBM
  • Dave Wootton, IBM
  • Claudia Knobloch, Juelich
  • Wolfgang Frings, Juelich
  • Jay Alameda, NCSA
  • Chris Navarro, NCSA
  • Steve Brandt, LSU

Minutes

  1. Actions from last meeting
    • BGQ license issue resolved. Still need to obtain new API. Will ask for help if it is not forthcoming.
    • The GridEngine package was renamed to all capitals. Caused some problems with Mac users, but resolved now.
  2. Status Reports
    • Juelich
      • Starting to look again at nattable as a possible replacement for jobs table views
      • Checked in new version 1.15 of LML DA driver, mainly internal changes
      • Added TORQUE_ALPS driver for Cray systems
      • Now generating usage display records that will be available when the node display enhancements are approved
      • Current version of BG adapter is not self deployable, will work on a version that can be installed by user for Juno
      • Question regarding Jaguar and restrictions to the number of sftp sessions allowed. There is a bug open on this, but not sure if there is a solution.
    • IBM
      • Opened CQ on Juelich contribution
      • Bug fixes in PE RM
      • Changes to get debugging to work
      • Still have issues with working directory.
    • NCSA
      • Still working on environment management support. Working with C/C++ sync, RMs..
      • Will also contribute lonestar and generic gridengine XML configuration
      • KISTI tutorial went very well
      • Provided test code to demonstrate SWT issue. Opened bug against SWT
      • Still an issue with localhost. Need to fix for SR2
      • Managed to get debugger up and running
      • Question regarding what RMs will be available in 5.0.5? We have monitoring support for GridEngine, LL, OpenMPI, PBS, PE, Torque and Torque/ALPS. No control configuration for GridEngine but can add this if NCSA contributes.
    • LSU
      • Planning to demonstrate shared job viewer
      • Want to be able to use own monitoring script. Need this functionality in Juno.
  3. Actions Arising
    • Greg to look through the EMS code to decide where it lives
    • Greg to create wiki page to document different options for working directories. Jay to provide their requirements.
  4. Meeting adjourned; Next meeting Feb 23, 2012 to be announced on ptp-scaling list.