It is true that certain tools only run at PCS Bridges and that cluster’s queue can get quite long, requiring a longer-than-usual wait (as compared to other tools/clusters).
Plus, right now that cluster is down for scheduled maintenance. Details: https://portal.xsede.org/user-news/-/news/item/11554
Tools that run at PCS Bridges:
The best strategy is to:
- Queue up jobs that run at this cluster as soon as you can. Batch processing features such as Workflows (and Dataset Collections) can aid with this.
- Plan for the tools above taking a bit longer to complete than most other tools.
- Avoid deleting/rerunning – that will only put your new job back at the end of the queue, extending the wait time, each time it is done.
In short, queue up jobs and allow them to stay queued until execution is completed. These tools are computationally expensive and do not run well on other clusters. Practically, this means the turnover for cluster availability is a bit slower, but the resources (memory + runtime allocations) are higher. A good thing, in the big picture, but requires some patience.