When using High Performance Computing (HPC) platforms, you must request your needed resources when you submit a job. This allows the scheduler to ensure enough resources are available for your job during the whole time it is running, and also know what resources will be free for others to use at the same time. Making inaccurate requests can result in longer queue times, job failures, and wasted resources.
In this session, we'll explain what 'job scaling' is and how it can help you make more accurate job submission requests.
Come along to hear from Anthony Shaw, one of our Researcher Support Analysts, who will discuss job scaling and testing best practices, such as:
- How to run small-scale tests,
- Using and interpreting the results of sacct, and
- Using the results of small-scale tests to predict resource requirements of your full size jobs
We'll have time for questions and will also record this session for posting on our YouTube Training Playlist afterwards.
More information:
Send an email to events@nesi.org.nz.
Image credit: Gerd Altmann, Pixabay
