
![]() | You may have heard about the fantastic computer animation in "The Wild", a Disney film that appeared in theaters last Spring and was recently released on DVD. What you may not have heard is that Condor was used to assist with the tremendous effort of managing over 75 million renders. Read a nice letter we received from Leo Chan and Jason Stowe, film Technology Supervisor and Condor Lead, respectively. |
The UNIX builds are now availble on the website. The primary focus of 6.3.0 has been in preliminary support for MPI jobs on dedicated clusters of machines. 6.3.0 is not as well-tested as 6.2.0, and undoubtedly contains a number of bugs. It should not be used by those requiring a tried and tested system; for that we recommend the 6.2 stable series.
The UNIX and NT builds are now availble on the website, and 6.0 and 6.1 releases have been removed. The 6.2.0 manual is also available, and includes a section on how to upgrade your pool from 6.0.3 to 6.2.0. The 6.3 DAGMan will also be released very soon.
6.1.17 adds full support for Solaris 8 and some bug fixes.
6.1.17 will likely be the last version of Condor until the 6.2.0 release. New users should choose it over 6.0.3.
For complete details, you can read about what's new in this version.
DAEMON_LIST = MASTER, COLLECTOR, NEGOTIATOR, STARTD, SCHEDD COLLECTOR = $(SBIN)/condor_collector NEGOTIATOR = $(SBIN)/condor_negotiatorNaturally, if you run fewer things on your central manager, you should take them out of the DAEMON_LIST.
Along with a whole series of bug fixes and new features, 6.1.9 is the first version of Condor to fully support all main Intel Linux distributions, including the 2.2.X kernel, and all 3 versions of the C library now in use (libc5, glibc 2.0 and glibc 2.1). For more info on Condor support for Linux, please see The Condor Linux README.
For complete details, you can read about what's new in this version.
Of the major improvements in this release, the condor_schedd has been
fixed to handle clusters of hundreds/thousands of jobs more efficiently. For standard
jobs you can now also specify file buffering, as well as file remapping.
For complete details, you can read about what's new in this version.
+MemoryRequirements = 90replacing 90 with the actual memory requirements of your job in megabytes. We encourage you to continue to specify your job's virtual memory requirements with the image_size command in your job description file.