➔ Slides
AuthorsCarrie Ganote 1
1 :
Indiana University (IU)AbstractGalaxy has an extremely flexible and versatile framework for job launching and allocation of resources. At Indiana University, we take advantage of the capabilities of Galaxy as well as adding a few tricks of our own. The following
topics will be explored:
- running in RAM
- job rerunning for programs that checkpoint sensibly
- Scaling with CPUs and memory using Galaxy Slots
- Job preemption and hidden tools as a way to sneak tests into production
We will go into a bit of detail about our implementation as well as the problems that pushed us to explore solutions.