help understanding pod names

In an attempt to control costs, am trying to figure out which users in my pangeo google cloud cluster have been creating dask kubernetes pods. All I have to work with are the kubernetes API calls logged by google cloud into bigquery.

The challenge is that it is hard to parse the names of the dask workers, which look something like this: dask-jovyan-3569756c-e2j5ht. I have concluded that the first random looking string is associated with the cluster, and the second is associated with the individual worker. Is there any hope of tracing this back to individual users?

[mrocklin]

I think that a broader solution would be to push on this upstream in JupyterHub.

Ideally when users arrive in their session in JupyterHub their username reflects their login.

Ideally if you login to ocean.pangeo.io, you could do something like this

$ echo $USER
rabernat

Instead of the behavior today

$ echo $USER
jovyan

If this change was made, then everything else would follow naturally

@yuvipanda

[jacobtomlinson]

Going to close this as solutions seems to have been found.