Jupyter support installation failed

Solved!
Aniket_Chauhan
Level 1
Jupyter support installation failed

I installed one plugin and after that I am trying to build code Environment  but getting error in Jupyter kernel. It is showing Jupyter kernel : not installed, and when I am trying to Re(Install) Jupyter. It is getting error which I have attached below. And In (installJupyterSupport.log)  it is showing that (No module named ipykernel). 

 

 

 

Please let me know what might be causing the issue?

Any Suggestions??


Operating system used: Windows

0 Kudos
1 Solution
tgb417

@Aniket_Chauhan 

Can you share a bit more about your operating environment?  Your post listed Windows as your operating system.  Dataiku DSS does not run directly on MS Windows.  It runs in Linux or Mac OS on Intel hardware.  So I expect that you have one of these OSes hosting your Data Science Studio (DSS) in some way. Either in the cloud our in your local environment.

 Plug-ins tend to install python into their own code environments.  Itโ€™s my opinion that those code environments should be left alone for sole use of the plugin.  Having a separate code environment for your work with Jupiter notebooks has seemed like a good idea to me.

Finally, you did not mention how you re-installed Jupyter Notebooks.  There are several ways to install python, via PIP or Anaconda this can make a difference.  Also you might have attempted these reinstalls from within DSS, or out in the operating system where DSS is running.

So there are a bunch of variability  about what might be going on in your case.

In your case Iโ€™d suggest a support ticket with the Dataiku team.  The support team has ways to help you to share diagnostic logs about the whole environment and then make targeted recommendations about fixes from within DSS and out at the operating system level.  They have been very helpful.

Hope that helps a bit.

 

--Tom

View solution in original post

0 Kudos
4 Replies
tgb417

@Aniket_Chauhan 

Can you share a bit more about your operating environment?  Your post listed Windows as your operating system.  Dataiku DSS does not run directly on MS Windows.  It runs in Linux or Mac OS on Intel hardware.  So I expect that you have one of these OSes hosting your Data Science Studio (DSS) in some way. Either in the cloud our in your local environment.

 Plug-ins tend to install python into their own code environments.  Itโ€™s my opinion that those code environments should be left alone for sole use of the plugin.  Having a separate code environment for your work with Jupiter notebooks has seemed like a good idea to me.

Finally, you did not mention how you re-installed Jupyter Notebooks.  There are several ways to install python, via PIP or Anaconda this can make a difference.  Also you might have attempted these reinstalls from within DSS, or out in the operating system where DSS is running.

So there are a bunch of variability  about what might be going on in your case.

In your case Iโ€™d suggest a support ticket with the Dataiku team.  The support team has ways to help you to share diagnostic logs about the whole environment and then make targeted recommendations about fixes from within DSS and out at the operating system level.  They have been very helpful.

Hope that helps a bit.

 

--Tom
0 Kudos
Aniket_Chauhan
Level 1
Author

Yes, Dataiku DSS doesn't support windows  directly so that  I am using a linux  as server and using it via chrome. I have attempted these reinstalls via DSS.

tgb417

@Aniket_Chauhan 

Iโ€™d recommend a support ticket.  

--Tom
0 Kudos
dgraham
Dataiker

Hi @Aniket_Chauhan ,

To re-install Jupyter support in the code environment, the ipykernel package must already be installed in the code-env. However, in this particular case, since the code environment is a plugin managed code-env, you would not be able to make modifications to its list of requested packages. To enable Jupyter support, you should instead create a new Python code-env, which installs ipykernel and enables support for Jupyter notebooks by default.