instead of the monkeypatch we did to keep the backport patch small
requiring tornado 5 simplifies things a ton because tornado.concurrent.Future is asyncio.Future
Eliminated the Kernel and Kernelspec handlers. The Websocket (ZMQ)
channels handler still remains. This required turning a few methods
into coroutines in the Notebook server.
Renamed the Gateway config object to GatewayClient in case we want
to extend NB server (probably jupyter_server at that point) with
Gateway server functionality - so an NB server could be a Gateway
client or a server depending on launch settings.
Add code to _replace_ the channels handler rather than rely on position
within the handlers lists.
Updated mock-gateway to return the appropriate form of results.
Updated the session manager tests to use a sync ioloop to call the
now async manager methods.
When kernels are culled, the kernel is terminated in the background,
unbeknownst to the session management. As a result, invalid sessions
can be produced that appear to exist, yet cannot produce a model from
the persisted row due to the associated kernel no longer being active.
Prior to this change, these sessions, when encountered via a subsequent
call to `get_session()`, would be deleted and a KeyError would be raised.
This change updates the existence check to tolerate those kinds of sessions.
It removes such sessions (as would happen previously), but rather than
raise a KeyError when attempting to convert the row to a dictionary,
it logs a warning and returns None, which then allows `session_exists()`
to return False since the session was removed (as was ultimately the
case previously).
Calls to `get_session()` remain just as before and have the potential
to raise `KeyError` in such cases. The difference now being that the
`KeyError` is accompanied by a message indicating the cause.
Fixes#4209
rather than initializing in main and starting in background, which seemed to cause occasional issues.
Also fix checking for dead notebook, which is no longer a subprocess.