These are chat archives for dereneaton/ipyrad

Apr 2018
Jean-Rémi Trotta
Apr 24 2018 07:49

Thanks @isaacovercast , in the "across" directory there is one file which keep growing:

-rw-r--r-- 1 jrtrotta production    199661916 Apr 24 09:35 ASSEMBLY_NAME.clust.hdf5
-rw-r--r-- 1 jrtrotta production    200007725 Apr 24 09:46 ASSEMBLY_NAME.clust.hdf5

Why the building database counter stay at 0%? Is it because you are not able to parse some outputs in order to know when this process will be done? Does this mean that at the end it will pass directly from 0% to 100%, without intermediate states? Cheers

Isaac Overcast
Apr 24 2018 13:44
@jrtrottablanc As long as the file is growing it's fine. Especially with such a big dataset as yours. The progress bar should start moving some time soon, but it'll move slow.
Chase Jalbert
Apr 24 2018 16:20
@isaacovercast Great, thanks for the tip. I figured it was an issue with the cluster. Unfortunately interactive jobs fail as well. We've made a bit of progress though. My admins think it's an issue with a security file used in the underlying parallelization library that isn't being created on our compute nodes. Thanks again
Isaac Overcast
Apr 24 2018 17:44
No problem. Each cluster requires its own special kind of voodoo, I've found. Good luck!