I am currently new to the GCP ecosystem. Over the past week, I have successfully created a notebook instance and have been using it. However, recently I encounter an issue whereby I am unable to open jupyter lab. Upon looking at the health status (image below). I realise the Juypter Lab API Status and Jupyterlab status are unhealthy.
Things I have tried and did not work:
Restarting the notebook instance
Restarting the Jupyter API
Restarting the Docker
Wondering if anyone has faced similar issues before? Any guidance on fixing this would be highly appreciated. Thank you in advance!
Notebook Health Status
Have you tried to run the troubleshooting 'diagnostic tool' as mentioned here:
https://cloud.google.com/vertex-ai/docs/general/troubleshooting#opening_a_notebook_results_in_a_524_a_timeout_occurred_error ?
Take a look at the output log of it and it might give you some hint what is the problem.
Related
Whenever I try to create a Kubeflow notebook server to build a pipeline from a jupyter notebook, it keeps loading forever without displaying any error.
I'm currently using a Kubeflow dashboard that's already up and running on a server, so I didn't deploy it myself and I'm not working on a local instance to use the terminal.
Any idea what the origin of the problem might be and how to solve it?
Here's a screenshot that might explain better.
So our team created a new Azure Machine Learning resource, but whenever I try to add a new notebook and try to edit it using "JUPYTERLAB" i get ERR_HTTP2_PROTOCOL_ERROR error, but the same notebook, when edited using EDIT IN JUPYTER works perfectly.
This is a blank and clean notebook, I also tried 2 different laptops and multiple browsers per laptop, same error. I also tried incognito and clearing cookies, but to no avail.
update: I seem to have accidentally replicated the issue and I now know what is causing it, the situation is that Im using my work laptop and constantly switching VPN connections, and some times, connecting to the AZURE PORTAl OUTSIDE the VPN. So, when you've worked on a notebook while inside a VPN, then you disconnected, and tried loading the notebook sometime later, you will encounter this
Have you tried creating a new Azure Machine Learning Compute Instance? Sometimes the VM has a fatal issue and a new one needs to be spun up.
Also try modifying the working Jupyter Notebook url by adding /lab to the end.
This problem has stomped me for hours, but I was finally able to fix it. What I did was I opened a terminal and did a Jupyter lab rebuild "jupyter lab build"
while creating jupyter notebook I am getting this error. I have also created the api key. Kindly help me please I am sharing the error screenshot1
Im trying to complete the very first training module offered by MS. Something Im missing that isn't detailed on the documentation of the training.
These are the instructions I'm following
https://github.com/MicrosoftDocs/mslearn-aml-labs/blob/master/labdocs/Lab01.md
All good until I have to run the second command defined on the notebook called
"01-Getting_Started_with_Azure_ML.ipynb".
And yes I entered the device login code as the instructions indicate.
Look at the attached screenshot of the error returned after running the command of the notebook.
Opened a case with Microsoft. They noticed is an issue affecting their VM servers.
This is their reply:
Hi Marbin,
Hope you are doing good. I had discussed this with our team as well. This was a known issue with workspace names with capital letters. So, the workspace name ‘ML_Battlefield’ was creating the issue. This is fixed in SDK version 1.3.0.
In the compute instance notebook, we can update the SDK version to: pip install --upgrade azureml-sdke
I am having massive trouble with using vpython in jupyter notbooks. I'm creating small animations with vpython. After a couple of minutes when I try to run a cell, it will either not show any output or will yield a error message "object could not be called".
The only fix I found for this is to restart or change the kernel. Most times it works then for the next few minutes until it stops again. This is really annoying and prevents real progress.
pictrue of error message with example code
all used objects have been imported in another cell before this one.
I am running vpython 7.3.2 and anaconda navigator 1.6.10 on a mac with High Sierra. As a browser I use Chrome.
Thank u for every hint to fix this permanently in advance!
Cheers,
Gordon
Try asking your question on the vpython forum and perhaps provide a sample notebook on github where with instructions on how to reproduce the problem.
https://groups.google.com/forum/?fromgroups&hl=en#!forum/vpython-users
Here is a link to demo vpython notebooks running in the cloud using binder service.
https://mybinder.org/v2/gh/BruceSherwood/vpython-jupyter/master?filepath=Demos
If you provide a notebook on github that demonstrates the problem then it should be reproduceable when running on mybinder.
https://mybinder.org/