I am trying to run a genome-wide analysis using UK Biobank data, and during a recent run found that chromosomes 1 and 2 had not finished in five days (compared to <40hr for all other chromosomes). On looking at the logs, it appears that the workflow analysis step for these two chromosomes never got past the localization stage ("Localizing input: gs://.... -> /cromwell_root/....").
Might this have to do with hard disk provisioning? I wasn't sure based on the documentation and behavior whether the disk space is dynamically allocated as needed versus listening to the resource requested in the WDL (in this case, "local-disk 250GB HDD").
Please sign in to leave a comment.