input / output volumes and logs

[vfdev-5]

This project looks very interesting and I wonder about two things that I haven't seen in the README.
Let's say i'd like to submit a job with a script that fetches (read-only) data from an NFS (which can be a PV in k8s) and occasionally writes output data to another data storage (another PV). How can we specify input/output data sources ? Is it coming with JupyterHub setup?
Second question about execution logs, if they are autmatically stored somewhere and could be easily inspected ?
Thanks !

[TomAugspurger]

How can we specify input/output data sources

I don't think kbatch does anything special with persistent volumes (JupyterHub or otherwise). #36 is a feature request for mounting the JupyterHub volume. It's possible that if you include the volumes in your job definition things will work.

Second question about execution logs, if they are autmatically stored somewhere and could be easily inspected ?

The logs are stored by Kubernetes, and the CLI has a pod logs command. https://kbatch.readthedocs.io/en/latest/examples/streaming-logs/streaming-logs.html has an example.