HTCondor
HTCondor is an open-source workload management system specialized for high-throughput computing (HTC). It enables researchers to efficiently schedule and run large numbers of computational jobs across clusters, grids, and cloud resources. HTCondor is widely used in scientific domains that require scalable, distributed computing for data- and compute-intensive tasks.
About
HTCondor is a workload management system that creates a high-throughput computing environment by efficiently using collections of machines to run large numbers of independent or loosely coupled jobs. It can manage resources ranging from a single cluster to combinations of campus clusters, clouds, and distributed grids, taking advantage of idle computing power and handling failures by automatically rescheduling jobs as needed.
Users describe jobs in submit files, specifying executables, inputs, resource requirements, and how many times a job should run. HTCondor’s ClassAd-based matchmaking finds suitable machines, schedules jobs, and manages priorities and fair use. This makes HTCondor well-suited for parameter sweeps, Monte Carlo simulations, large batch processing, and other scenarios involving thousands of similar jobs. For research computing centers, HTCondor can serve as the primary batch system for campus clusters or as the interface to distributed resources, including the OSG Open Science Pool.