NotebookApp not being culled

JupyterHub discuss
1

Hi all,
I just tried setting

c.NotebookApp.shutdown_no_activity_timeout = 60 * 60

in my JupyterHub configuration.

However, after spawning a notebook yesterday evening it was still running today morning.
Do you see any mistake?

2

Hi!
Could you show us your full jupyterhub configuration, and also tell us what Jupyter applications you were running in addition to notebook? Have you enabled the idle culler GitHub - jupyterhub/jupyterhub-idle-culler: JupyterHub service to cull idle servers and users ?

If you’re relatively new to JupyterHub you may find it’s easier to install JupyterHub using The Littlest JupyterHub — The Littlest JupyterHub v0.1 documentation

3

Thanks for your answer @manics, here is my (stripped down) jupyterhub config:

c.JupyterHub.allow_named_servers = True
c.PAMAuthenticator.open_sessions = False
from jupyterhub.auth import PAMAuthenticator
import pamela
from tornado import gen
class KerberosPAMAuthenticator(PAMAuthenticator):
    @gen.coroutine
    def authenticate(self, handler, data):
        """Authenticate with PAM, and return the username if login is successful.
        Return None otherwise.
        Establish credentials when authenticating instead of reinitializing them
        so that a Kerberos cred cache has the proper UID in it.
        """
        username = data['username']
        try:
            pamela.authenticate(username, data['password'], service=self.service, resetcred=pamela.PAM_ESTABLISH_CRED)
        except pamela.PAMError as e:
            if handler is not None:
                self.log.warning("PAM Authentication failed (%s@%s): %s", username, handler.request.remote_ip, e)
            else:
                self.log.warning("PAM Authentication failed: %s", e)
        else:
            return username
c.JupyterHub.authenticator_class = KerberosPAMAuthenticator
c.JupyterHub.bind_url = 'http://:8686/jupyter/'
c.JupyterHub.default_url = 'home'
c.JupyterHub.named_server_limit_per_user = 5
c.Spawner.default_url = '/lab'

import batchspawner
c.JupyterHub.tornado_settings = {
    'slow_spawn_timeout': 30,
}
c.Spawner.http_timeout = 120
c.Spawner.start_timeout = 60
c.JupyterHub.spawner_class = 'wrapspawner.ProfilesSpawner'
c.BatchSpawnerBase.req_nprocs = '2'
c.BatchSpawnerBase.req_runtime = '48:00:00'
c.BatchSpawnerBase.req_memory = '12gb'
c.BatchSpawnerBase.req_prologue = '''...'''

c.NotebookApp.shutdown_no_activity_timeout = 60 * 60

I was only running a number of python kernels in idle-mode on the worker.

4

Have you tried adding the idle-culler as suggested in my previous post?

5

I tried the idle-culler now with '--timeout=3600 and it works great, thanks :slight_smile:

Does this mean that c.NotebookApp.shutdown_no_activity_timeout has no effect?

6

It’s a bit complicated :smiley:

See this post

and the follow-ups, especially about turning on debug logging and ServerApp vs NotebookApp.

1 Like
7

I see, so as I understand the NotebookApp.shutdown_no_activity_timeout culler only culls, when there are no active kernels.
So I would have to enable kernel culling as well if I wanted that setting to work, right?

8

I’m not sure, most of what I know is from the linked thread!