Skip to Main Content
Science Gateways often execute computations on clusters managed by batch schedulers. These schedulers queue computations until resources are available and then execute them. Clusters are typically scarce resources so the amount of time a computation waits before it begins to execute can be significant. Gateways can potentially reduce the turn around time of their computations and improve their user experience if they use estimates of how long computations will wait in a queue before beginning to execute. This paper describes a service that provides such estimates as well as current and historical job information. An instance of this service has been deployed on TeraGrid and is available to TeraGrid science gateways.