Minutes for FTB conference call - 2010 April 14th

From CIFTS
Jump to: navigation, search

Attendees

  • Argonne National Lab: Rinku
  • Oak Ridge National Lab: Hoony, Thomas
  • Ohio State University: Sonia, Raghu
  • Indiana University: Abhishek, Josh
  • University of Tennessee: Absent
  • Lawrence Berkeley National Lab: Absent

Items Discussed

  • ANL
  1. Rinku working on FTB-0.6.2 and will continue to do so for the next two weeks. 0.6.2 bug fix release targeted for end of month
  2. Harish continuing his work on FTB-enabled Cobalt development
  • ORNL
  1. Thomas working on refining the test scripts.
  2. Hoony is working on fixing some bugs for 0.6.2
  3. Rinku needs to send Hoony BGP RAS logs. Job logs from the BG/P can perhaps not be shared, Rinku is looking into this. Hoony and Rinku to figure our if the ORNL monitoring tool can be made to work on BG/P with information currently shared by ANL ALCF folks.
  • OSU
  1. OSU is working on MVAPICH migration and checkpointing. As per Raghu, these changes in MVAPICH will allow MVAPICH to checkpoint and migrate based on received FTB events. These FTB events will be currently published by FTB-IPMI component (that is within OSU plan for FTB) and FTB-IB or FTB-MVAPICH itself.
  2. Rinku suggested that OSU share the events they plan to react to with other CIFTS folks to be in sync
  3. OSU to provide more information on the FTB-related features in the next conference call
  • IU
  1. Continuing work on FTB-enabled Open MPI and the notifier component of Open MPI. FTB-related changes will be integrated in the main release in a month's time.
  2. Abhishek to send out the common MPI document to the mailing list, so that all MPI folks (especially OSU) can take a look and comment on any overlapping events.
  • LBNL
  1. No significant update
Personal tools