Using checkpointing to recover from poor multi-site parallel job scheduling decisions.

William M. Jones

Browse the full Middleware paper archive.