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
Bruno P Kinoshita from CWL4HPC has an HPC setup in Barcelona with x86_64 head nodes and ARM worker nodes. Toil supports ARM but only thinks about having single-architecture clusters, and the HPC virtualenv setup we use doesn’t really think about having to run the workers under a different Python interpreter.
We should try and support this cluster setup, and let you run the Toil leader on the head node and the Toil workers on ARM.
The workaround it so make sure the cluster lets worker nodes submit new jobs, and then run the leader as a job on the ARM workers.
┆Issue is synchronized with this Jira Story
┆Issue Number: TOIL-1687
The text was updated successfully, but these errors were encountered:
Bruno P Kinoshita from CWL4HPC has an HPC setup in Barcelona with x86_64 head nodes and ARM worker nodes. Toil supports ARM but only thinks about having single-architecture clusters, and the HPC virtualenv setup we use doesn’t really think about having to run the workers under a different Python interpreter.
We should try and support this cluster setup, and let you run the Toil leader on the head node and the Toil workers on ARM.
The workaround it so make sure the cluster lets worker nodes submit new jobs, and then run the leader as a job on the ARM workers.
┆Issue is synchronized with this Jira Story
┆Issue Number: TOIL-1687
The text was updated successfully, but these errors were encountered: