kernel error when starting a python notebook

longhowlam
Level 3
kernel error when starting a python notebook



Hi,



I installed DSS on a GCP VM when I start a Python  jupyter notebook, I get the following kernel error. Any Ideas



Regards, Longhow



 



Traceback (most recent call last):

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/web.py", line 1592, in _execute

    result = yield result

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/gen.py", line 1133, in run

    value = future.result()

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/concurrent.py", line 261, in result

    raise_exc_info(self._exc_info)

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/gen.py", line 1141, in run

    yielded = self.gen.throw(*exc_info)

  File "/home/longhowlam/dataiku-dss-6.0.2/dku-jupyter/packages/notebook/services/sessions/handlers.py", line 81, in post

    user=self.get_current_user(),

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/gen.py", line 1133, in run

    value = future.result()

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/concurrent.py", line 261, in result

    raise_exc_info(self._exc_info)

  File "/home/longhowlam/dataiku-dss-6.0.2/python.packages/tornado/gen.py", line 326, in wrapper

    yielded = next(result)

  File "/home/longhowlam/dataiku-dss-6.0.2/dku-jupyter/packages/notebook/dataiku/sessionmanager.py", line 55, in create_session

    kernel_context = self.new_kernel_context(head, tail, user, session_id)

  File "/home/longhowlam/dataiku-dss-6.0.2/dku-jupyter/packages/notebook/dataiku/sessionmanager.py", line 32, in new_kernel_context

    "sessionId" : session_id

  File "/home/longhowlam/dataiku-dss-6.0.2/dku-jupyter/packages/notebook/dataiku/__init__.py", line 27, in pintercom_json_post

    raise Exception("Failed privileged call (%s): %s" % (path, _get_error_message(err_data).encode("utf8")))

Exception: Failed privileged call (jupyter/new-kernel-context): Spark Config not found: default

0 Kudos
2 Replies
Clément_Stenac
Dataiker
Hi,

You removed the Spark named config named "default", which is configured as the Spark config to use for your notebooks. We'd advise that you go in Admin > Settings > Spark and either select another config as the default for notebooks, or create a new config named "default" with no settings in it.
0 Kudos
longhowlam
Level 3
Author
Hi Clement,

Thanks this worked!

Note that the removal of default config 'silently' happend when I set up Dataiku to run notebooks on GKE. Other users may see the same behaviour?

Regards, Longhow
0 Kudos

Labels

?
Labels (3)
A banner prompting to get Dataiku