ARCHIVED: What is co-scheduling, and does XSEDE support it?

This content has been archived, and is no longer maintained by Indiana University. Resources linked from this page may no longer be available or reliable.

Definition

Co-scheduling gives you simultaneous access to resources that are part of different logical systems and, potentially, in different physical locations.

Co-scheduling lets you run jobs that have resource requirements beyond those of any one resource. For example, co-scheduling resources lets you:

  • Run jobs that require simultaneous access to a computational system and a visualization system
  • Access a scientific instrument to perform visualization in parallel with a simulation
  • Perform data analysis in parallel with data acquisition

Co-scheduling is distinct from on-demand computing, in which running or queued jobs may be preempted for time-critical applications (e.g., earthquake or tornado warning operations).

Implementation on XSEDE

Co-scheduling services are not yet widely in production on the Extreme Science and Engineering Discovery Environment (XSEDE), though they may be available experimentally on some resources.

Note: Co-scheduling is distinct from advance reservation. Advance reservation requests resources from a single logical system. Co-scheduling allocates resources from multiple logical systems. However, co-scheduling is often implemented atop advance reservations.

Getting help

If you need help or have questions about co-scheduling on XSEDE, contact the XSEDE Help Desk.

This document was developed with support from National Science Foundation (NSF) grants 1053575 and 1548562. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF.

This is document avwz in the Knowledge Base.
Last modified on 2018-03-01 16:50:00.

Contact us

For help or to comment, email the UITS Support Center.