Open MPI logo

MTT Users Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2006-12-21 15:52:32


As Ethan already noted, we'll be rolling out a new version of MTT
after the new year. For your Christmas pleasure, here's a summary of
the new features:

Sever-side / General:
---------------------

- Revamped web reporter / nightly e-mails -- huge improvements over
   the prior release, including:
     - Leaner, faster, and more intuitive web user interface
     - Clickable summary reports for drilling down
     - More flexible querying (e.g., multiple selection, regular
       expressions, sortable columns, and cherry-picked reports)
     - Query screen and reports on one screen (no frames needed)
     - Super short URLs
     - Display 100 rows at a time
- Distinguish between tests that pass, fail, timeout, and are
   intentionally skipped.
- Added support for latency and bandwidth performance testing and
   reporting. Verified with NetPIPE and IMB MPI test suites.

MTT client:
-----------

- Added support for environment modules (http://modules.sf.net/)
- Dramatically better memory utilization (especially for large
   numbers of test runs)
- Add some "progress bar"-like output to --verbose output so that
   users can see how far along the tests are
- Added "--trial" option to MTT client to allow testing of the MTT
   client/INI file setup itself without submitting potentially
erroneous/
   bad results data into the MTT results database
- Print [somewhat] better messages about errors with INI files
- MTT client now distinguishes between specifying tests to run and
   analyzing their results (e.g., correctness and performance)

-- 
Jeff Squyres
Server Virtualization Business Unit
Cisco Systems