Content

Long-running jobs present multiple problems: Most HPC clusters enforce a strict maximum job time limits, and additionally, the longer the runtime, the higher the risk (and cost!) of failure.

This session explores strategies that could be employed to work within these limits and potentially mitigate failure risks. We will look at signaling imminent job timeouts to applications, ways to minimize the impact of job failures, and possibly external tooling for checkpointing and failure recovery.
 

Access

To access the event, click on the following link: 

Big Blue Button Link <https://webconf.hrz.uni-marburg.de/n/rooms/hsy-cmh-knb-ndw/join> 

Access Code: kw7i98