You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Yeah, we'll have to remove the temp_dir from the config.. I thought we already did but that was for Spark.. I just created a new issue.
So, here it looks like the temp_dir is used for the pipeline workDir as well as the temp dir of Picard (since we generally use temp_dir=/scratch). We'll need to standardize how we handle tool temp directories and pipeline workDir.
For now, I think we can just resolve #181 and release the major version as discussed to minimize the chance of breaking the cluster...
In the long run, we may want to have a specific temp dir for each process/tool as default to make the temp/workDir dir structure cleaner, debugging easier, and avoid potential temp dir conflicts (probably unlikely but just in case).
As default, maybe we can do something like
/scratch/${pipeline-name}-${runName}/tmp/${process_name} (tool name should be in the process name)
/scratch/align-DNA-${runName}/tmp/run-MarkDuplicatesSpark-GATK
/scratch/align-DNA-${runName}/workDir
Yeah, we'll have to remove the
temp_dir
from the config.. I thought we already did but that was for Spark.. I just created a new issue.So, here it looks like the
temp_dir
is used for the pipelineworkDir
as well as the temp dir of Picard (since we generally usetemp_dir=/scratch
). We'll need to standardize how we handle tool temp directories and pipeline workDir.For now, I think we can just resolve #181 and release the major version as discussed to minimize the chance of breaking the cluster...
In the long run, we may want to have a specific temp dir for each process/tool as default to make the temp/workDir dir structure cleaner, debugging easier, and avoid potential temp dir conflicts (probably unlikely but just in case).
As default, maybe we can do something like
We can discuss this next week.
Originally posted by @tyamaguchi-ucla in #180 (comment)
The text was updated successfully, but these errors were encountered: