@james-tn , I'm working with a large research center that has AzureML and wants to run horizontally scaling Dask workflows. I'm hoping you can help me answer these questions:
-
What would you suggest to be the best way today to run Dask on an AzureML Cluster? Looks like dask-on-ray is not really supported anymore?
-
If there was funding to develop a package to run Dask clusters on AzureML Clusters more easily, what approach would you suggest?