Skip to main content.

Computing with HTCondor™

Our goal is to develop, implement, deploy, and evaluate mechanisms and policies that support High Throughput Computing (HTC) on large collections of distributively owned computing resources. Guided by both the technological and sociological challenges of such a computing environment, the Center for High Throughput Computing at UW-Madison has been building the open source HTCondor distributed computing software (pronounced "aitch-tee-condor") and related technologies to enable scientists and engineers to increase their computing throughput.

Note: The HTCondor software was known as 'Condor' from 1988 until its name changed in 2012. If you are looking for Phoenix Software International's software development and library management system for z/VSE or z/OS, click here.

Latest News [RSS]

HTCondor Week tutorials free to UW-Madison faculty, staff and students

May 11, 2016

Just a reminder that the Tuesday tutorials at HTCondor Week are free to UW-Madison faculty, staff and students. If you have any interest in using HTCondor and Center for High Throughput Computing resources, we'd love to see you on Tuesday.

However, as everyone knows, there ain't no such thing as a free lunch -- people who attend the tutorials without paying are not eligible for the lunch (you can still get snacks at the breaks, though).

HTCondor 8.5.4 released!

May 2, 2016

The HTCondor team is pleased to announce the release of HTCondor 8.5.4. This development series release contains new features that are under development. This release contains all of the bug fixes from the 8.4.6 stable release. For sites using partitionable slots, one of the fixes from the 8.4.6 release corrects a serious regression in version 8.5.3. Highlights of the release are: Fixed a bug that delays schedd response when significant attributes change; Fixed a bug where the group ID was not set in Docker universe jobs; Limit update rate of various attributes to not overload the collector; To make job router configuration easier, added implicit "target" scoping; To make BOSCO work, the blahp does not generate limited proxies by default; condor_status can now display utilization per machine rather than per slot; Improve performance of condor_history and other tools. Further details can be found in the Version History. HTCondor 8.5.4 binaries and source code are available from our Downloads page.
HTCondor Week registration deadline is May 9

April 26, 2016

The registration deadline for HTCondor Week 2016 has been extended to Monday, May 9. This will be the final extension -- we need to finalize attendance numbers for the caterers.

Also note that Wednesday, April 27, is the last day to be guaranteed to get the conference rate at the DoubleTree Hotel.

Why do supercomputers have to be so big?

April 26, 2016

In this Blue Sky Science article Lauren Michael of the Center for High Throughput Computing answers a question from an eight-year-old: Why do supercomputers have to be so big? The article can also be viewed at the Wisconsin State Journal and the Baraboo News Republic.
HTCondor 8.4.6 released!

April 21, 2016

The HTCondor team is pleased to announce the release of HTCondor 8.4.6. A stable series release contains significant bug fixes. Highlights of this release are: fixed a bug that could cause a job to fail to start in a dynamic slot; fixed a negotiator memory leak when using partitionable slot preemption; fixed a bug that caused supplemental groups to be wrong during file transfer; properly identify the Windows 10 platform; fixed a typographic error in the LIMIT_JOB_RUNTIMES policy; fixed a bug where maximum length IPv6 addresses were not parsed; a few other bug fixes, consult the version history. Further details can be found in the Version History. HTCondor 8.4.6 binaries and source code are available from our Downloads page.
Large Hadron Collider experiment uses HTCondor and Amazon Web Services to probe nature

April 5, 2016

This Amazon Web services blog post explains how scientists at Fermilab (a Tier 1 data center for the CMS experiment at the LHC) use HTCondor and Amazon Web Services to elastically adapt their computational capacity to changing requirements. Fermilab added 58,000 cores with HTCondor and AWS, allowing them to simulate 500 million events in 10 days using 2.9 million jobs. Adding cores dynamically improves cost efficiency by provisioning resources only when they are needed.
Join us at HTCondor Week 2016!

March 30, 2016

We want to invite you to HTCondor Week 2016, our annual HTCondor user conference, in beautiful Madison, Wisconsin, May 17-20, 2016. We will again host HTCondor Week at the Wisconsin Institutes for Discovery, a state-of-the-art facility for academic and private research specifically designed to foster private and public collaboration. It provides HTCondor Week attendees with a compelling environment in which to attend tutorials and talks from HTCondor developers and users like you. It also provides many comfortable spaces for one-on-one or small group collaborations throughout the week.

Our current development series, 8.5, is well underway toward our upcoming production release. When you attend, you will learn how to take advantage of the latest features. You'll also get a peek into our longer term development plans--something you can only get at HTCondor Week!

We will have a variety of in-depth tutorials and talks where you can not only learn more about HTCondor, but you can also learn how other people are using and deploying HTCondor. Best of all, you can establish contacts and learn best practices from people in industry, government, and academia who are using HTCondor to solve hard problems, many of which may be similar to those you are facing.

Speaking of learning from the community, we'd love to have you give a talk at HTCondor Week. Talks are 15-20 minutes long and are a great way share your ideas and get feedback from the community. If you have a compelling use of HTCondor you'd like to share, see our speaker's page.

You can register, get hotel details, and see the agenda overview on the HTCondor Week 2016 site.

HTCondor 8.5.3 released!

March 24, 2016

The HTCondor team is pleased to announce the release of HTCondor 8.5.3. This development series release contains new features that are under development. This release contains all of the bug fixes from the 8.4.5 stable release. Enhancements in the release include: Use IPv6 (and IPv4) interfaces if they are detected; Prefer IPv4 addresses when both are available; Count Idle and Running jobs in Submitter Ads for Local and Scheduler universes; Can submit jobs to SLURM with the new "slurm" type in the Grid universe; HTCondor is built and linked with Globus 6.0. Further details can be found in the Version History. HTCondor 8.5.3 binaries and source code are available from our Downloads page.

Software

Community

Research and Development