A issue about batch run
When I execute multiple workflow tasks or batch runs, the priority level automatically changes from "normal" to "high." As a result, my costs increase significantly. For example, as shown in the attached image, after running the first step of the SAIGE algorithm twice, my costs went from £1.7 to £14, and it’s still not complete. I encountered this issue previously while batch-running the second step of the SAIGE algorithm for 22 chromosomes, which led to an increase in costs of nearly £100. Due to time constraints, I didn't restart the process immediately after noticing the priority change from normal to high. This has left me quite confused.
Comments
2 comments
Hi Shangyou,
By selecting high priority, on-demand instances will be selected. Low priority will select spot instances. Normal priority will try select a spot instance but will switch to on-demand after 15 minutes if no spot instances are free.
If you select low priority this should ensure that only spot instances are used.
Please see this article for more information about setting job priorities: https://dnanexus.gitbook.io/uk-biobank-rap/working-on-the-research-analysis-platform/managing-jobs/managing-job-priority
Hope this helps, thank you for using the community forum.
Please sign in to leave a comment.