Usage question: Existing client?

If a client is started up within a notebook without specifying the tcp address, e.g.:

from dask.distributed import Client

client = Client(n_workers=8, threads_per_worker=8)

is there any way to connect the extension to it? I don’t see the client showing up in the left side extension window, and I don’t know how to search for it.

Relatedly, what is the search bar for? I’m not clear what I’m supposed to be searching for there, or if it’s related to my question.

[ian-r-rose]

jupyter-server-proxy should be installed when installing dask_labextension, but people have run into issues there, so I was just trying to be explicit.

The clusters listing only knows about clusters that it creates itself, so if you do it out-of-band, there is no way to make them show up (at least not right now, we may tackle cluster discovery down the road).

[JamesSample]

A small addition/UI suggestion…

I had exactly the same problem running on JupyterHub and, as suggested, entering the dashboard link as

/user/{JUPYTERHUB_USER}/proxy/{port}/status

as in dask-examples solved the problem for me. However, one thing I found confusing at first was that I was originally entering the link and then clicking the “search” (magnifying glass) button. This didn’t work - it just reset the URL to http://127.0.0.1:8787/ and then nothing happened.

Entering the dashboard link as given above and pressing ENTER worked perfectly.

I think this is just my misunderstanding of what the “search” button actually does i.e. searching for an active cluster in the default location versus searching for the cluster name entered in the text box?

Maybe it would be more intuitive if entering a link in the text box and then clicking search attempted to connect to that cluster, rather than to a default cluster at http://127.0.0.1:8787/ ?