A Web Service for HTCondor Scheduling System

As a high throughput computing (HTC) task scheduling system, the HTCondor (hereinafter called Condor) has been widely applied in the field of high-energy physics computing and analysis, due to its open-source, installation-friendly and better scheduling algorithm. Many research institutes deployed Condor on their computing clusters to complete the resource management and task scheduling. However, Condor is lack of a user-friendly web interface. Some beginners or students, who are not familiar with Condor, have to spend time to figure out the rules about how to submit and manage their computing or analysis tasks using condor-specified commands. Fortunately, Condor provides several application programming interfaces (APIs), which makes it possible to develop relevant application or service according to the actual environment. Herein, by utilizing the Condor python interface combined with the currently popular VUE front-end architecture, we developed a web service for Condor system. Instead of using condor commands, users can submit, query and control their HTC tasks on a web browser. This work brings great convenience to Condor system users, especially for beginners and students. Besides, it also provides a real-time monitor function, which makes more convenient and intuitive for users and system managers to monitor the cluster resources and tasks' status in real time.