HTCondor-CE
HTCondor-CE is a specialized service that provides a standardized entry point for submitting jobs to distributed computing resources. It translates external job requests into the local scheduler’s format, ensuring interoperability across grids and national cyberinfrastructure. HTCondor-CE is commonly used in research collaborations that require cross-institutional job submission.
About
HTCondor-CE is a “Compute Entrypoint” (CE) built on HTCondor that acts as a job gateway for sites that are part of larger computing grids such as the OSG. It receives remote resource allocation requests (pilot jobs or grid jobs), authorizes them, and then routes them into the site’s local batch system (e.g., HTCondor, Slurm, PBS Pro/Torque, LSF, Grid Engine) using the HTCondor Job Router.
From an operations perspective, HTCondor-CE provides a standardized, configurable interface between external grids and local clusters, handling authentication, accounting, job transformation, and monitoring. It allows a site to contribute resources to federated HTC infrastructures while retaining control over local policies and scheduling. For research computing staff, HTCondor-CE documentation and webinars explain how to deploy and configure a CE, integrate it with site security and accounting, and debug routing and queue issues, making it an essential component for institutions that want to participate in distributed HTC collaborations.