Darshan events at SC16

Darshan will be featured in several community events at SC 2016; please stop by and visit us if you are interested in learning more about Darshan and other related topics:

  • Tutorial: Parallel I/O in Practice, by Robert Latham, Robert Ross, and Brent Welch.  Sunday, November 13th, 8:30am – 5pm, in room 355-B (requires tutorial registration).
  • Paper presentation: Modular HPC I/O Characterization with Darshan,
    by Shane Snyder, Philip Carns, Kevin Harms, Robert Ross, Glenn Lockwood and Nicholas Wright.  Sunday, November 13th, 3:30pm – 3:55pm, ESPT workshop in room 155-E (requires workshop registration).
  • Birds of a Feather session: Analyzing Parallel I/O, by Philip Carns and Julian Kunkel with guest presentations from Glenn Lockwood, Shane Snyder, Jakob Luettgau, Salem El Sayed, and Xiaosong Ma.  Wednesday, November 16th, 5:15pm – 7:00pm, in room 155-E.

 

Darshan 3.1.2 point release now available

Darshan version 3.1.2 is a point release meant to address a bug in versions 3.1.0 and 3.1.1 that can cause crashes in applications (or other binaries) when LD_PRELOADing the Darshan library. The bug was caused by an error in the resolution of the address of the fopen64 function call, so any binary that calls this function is affected. This bug does not affect Darshan’s methods for instrumenting statically-linked executables.
The release tarball can be downloaded HERE.
Documentation for this version of Darshan can be found HERE.

Darshan 3.1.1 is now available for download

Darshan version 3.1.1 is now available for download. This release contains the following bug fix for the recent 3.1.0 release:

  • Fix bug in calculation of the MPIIO_F_WRITE_START_TIMESTAMP counter in the MPI-IO module that could lead to incorrect timestamps in certain cases

Please let us know if you have any issues, questions, or comments using either our mailing list or the Darshan GitLab page.
The release tarball can be downloaded HERE.
Documentation for this version of Darshan can be found HERE.

New Darshan 3.1.0 release now available

Darshan version 3.1.0 is now available for release! This version contains a number of general enhancements to the code base:

  • addition of a stdio I/O library instrumentation module
    • this handles more in-depth instrumentation of file stream I/O functions like fopen(), fprintf(), fscanf(), etc.
    • this module also captures stats on the standard I/O streams (stdin, stdout, & stderr)
  • addition of a Lustre instrumentation module
    • this module provides Lustre striping details (e.g., stripe width, stripe size, list of OSTs a file is striped over)
  • addition of a new mmap-based logging mechanism that allows Darshan to generate output logs even in cases where applications don’t call MPI_Finalize()
    • Note that these logs are uncompressed and are per-process rather than per-job
  • addition of the darshan-merge utility to darshan-util to allow per-process logs generated by the mmap-based logging mechanism to be converted into Darshan’s traditional compressed per-job log files
  • modification of the POSIX module timestamp counters to also include a LAST_OPEN & FIRST_CLOSE counters to give more details on application I/O intervals

This version has been tested extensively on a number of platforms. As always, please let us know if you have any issues, questions, or comments using either our mailing list or the Darshan GitLab page.
The release tarball can be downloaded HERE.
Documentation for this version of Darshan can be found HERE.

Darshan 3.0.1 release now available

Darshan version 3.0.1 is now available for download. This version contains a number of bug fixes and enhancements to Darshan 3.0.0. For reference, the complete list of changes for this release is:
* bug fix in darshan logutil mount parsing code that was causing file paths to be matched to the first mount point with a common prefix rather than the one with the longest common prefix
* bug fix in the darshan-util bzip2 configure check that was accidentally overwriting Darshan’s LDFLAGS
* minor bug fixes to IO start time counters in all modules to set IO start time to the actual first start time rather than the first IO op to complete
* update darshan-util perl scripts to get perl bin from user’s path, rather than from /usr/lib (reported by Kay Thust)
* update Darshan’s fortran and cxx compiler wrapper generators to automatically detect MPICH library names on BG/Q
* bug fix for calculating Darshan’s agg_perf_by_slowest performance metric correctly
* add performance estimate info to darshan-job-summary graph
We have tested this release on multiple platforms, but please let us know if you have any issues or questions. You can report issues directly using Darshan’s GitLab page or by using the Darshan users mailing list.
The release tarball can be downloaded HERE.
Documentation for this version of Darshan can be found HERE.

Darshan 3.0.0 performance evaluation technical report

As part of our transition to the new modularized version of Darshan (3.0.0), we have published a technical report evaluating its performance on Edison, a production Cray XC30 system at NERSC. This technical report primarily serves as validation that Darshan’s performance remains suitable for full-time deployment on production HPC systems, but it also provides some additional background that may be useful in understanding the design changes made to Darshan’s runtime library and log file format. Our results indicate that Darshan’s new design is still lightweight enough to run full-time on production systems with little to no impact on the performance of instrumented applications.
A PDF for this technical report can be found HERE.

Darshan 3.0.0 official release is ready

Darshan 3.0.0, the new modularized implementation of Darshan, is now officially ready to be released! This is the first stable release of the Darshan 3.0 fork and builds off the experimental pre-releases we have been providing over the last few months.  We encourage users to consider upgrading to this version in the near future as it contains numerous enhancements over older versions: logging of full file paths, clearer parser and job summary output, addition of a BG/Q specific instrumentation module, future proofing for any upcoming instrumentation modules, etc.
The following changes have been made between this official release and the last pre-release (darshan-3.0.0-pre3):

  • fix bug causing compile errors when building external utilities that use the darshan-util API/library
  • update docs to give debugging tips for cases where Darshan logs are not generated
  • update darshan-runtime docs to give information on upgrading Darshan
  • fix shared library regression test script to check for potential errors with Darshan symbols rather than failing silently in these cases
  • bug fix for determining minimum non-zero counters in shared file reductions in all modules
  • loosen Darshan’s PMPI symbol check to prevent inadvertent disabling of Darshan for some MPICH builds
  • bug fix for resolving MPI_Gather and MPI_Barrier when LDPRELOADing Darshan’s shared libraries
  • add more helpful error handling when opening 2.x version log files
  • port darshan-diff utility over to new 3.0.0 log file format
  • fix numerous configure bugs on Cray systems
  • add synthetic benchmarking hooks for testing Darshan’s shutdown procedure

We have tested this version extensively on a number of platforms over the last few months and are satisfied with the new features and Darshan’s overall performance. Included in the darshan-runtime documentation (a link is provided below) is some information on how to best upgrade from a Darshan 2.x version to Darshan 3.0.0. As always, please let us know if you have any issues or questions (using either the Darshan users mailing list or by opening an issue on the Darshan GitLab page).
The release tarball can be downloaded HERE.
Documentation for this version of Darshan can be found HERE.

New point release available for Darshan 3.0.0

There is a new point release available for the modularized 3.0.0 version of Darshan. This point release contains the following changes:

  • add module-specific version fields to header to allow utilities to handle different versions of a module’s I/O data for backwards compatibility — NOTE: this breaks the log file parsing for logs obtained using Darshan-3.0.0-pre2 & Darshan-3.0.0-pre1
  • bug fix in regression test scripts for setting proper environment variables to use MPI profiling configuration for Fortran apps
  • bug fix in bzip2 log writing implementation in darshan-logutils
  • possible race conditions resolved in each module’s shutdown code
  • general code, comment, and documentation cleanup
  • addition of module-specific counter descriptions printed prior to parsing a modules I/O data in darshan-parser

As a reminder, this is still an experimental release and is not appropriate for production use at the moment, though we are getting close to an official release! We hope you can take the time to try the new version in preparation for the upcoming release, and look forward to any feedback you may have.
Documentation for this version of Darshan can be found HERE, under the Darshan 3.0.0 heading.

SC15 BoF: Analyzing Parallel I/O

Julian Kunkel and Philip Carns are co-organizing a Birds of a Feather session at SC15 in Austin, Texas, entitled “Analyzing Parallel I/O”.  The BoF will feature a variety of guest speakers to share their perspectives on the field as well as a discussion session.  This is a great opportunity to find out more about I/O analysis tools and best practices.  The BoF will be held in room 13B from 1:30 to 3:00pm on Thursday, Nov 19.  The tentative agenda is available online.