You are here
The Benefits of GLOST for Many Jobs
This webinar is presented by SHARCNET, an HPC consortium that is part of Compute Ontario, one of four regional partners of Compute Canada. You do not need a SHARCNET or Compute Canada account to attend this session.
* * * * * * * * * * * *
This webinar will look at GLOST, the Greedy Launcher Of Small Tasks. It is a tool, provided as a module on Compute Canada clusters for running many sequential jobs of short or variable duration, or doing parameter sweeps. If your project requires running many thousands of small jobs then setting them up within a GLOST input file can reduce the effort to simply submitting a single job! It also provides a convenient solution to the problem of running many jobs on clusters which impose maximum job submission limits.
In the webinar several, hands on examples will be demonstrated to show how to setup and submit GLOST jobs and compare their performance with other approaches.
By attending the webinar, users should gain a clear understanding what GLOST does, how to use it and ultimately decide if it can help improve a project's research workflow on Compute Canada clusters.
How to Join the Webinar:
At the time of the session, click this link to join the videoconference:
For more information or to ask questions about this session, contact email@example.com.