Notebook - Forbidden 403

Hi All,

All of a sudden I’m getting a “403 Forbidden” error when running Jupyter Notebook. It has been running fine until now. It is possible that a recent power cut that shutdown my Mac could have caused it.

I am on Mac, and running “jupyter notebook” in terminal opens up the normal initial view showing the home directory. But any attempt to go to another directory shows no files, and the message “Server error: Forbidden”. I get a similar error when trying to create a new workbook.

The same happens on any browser, and restarting has not helped.

The message I see in terminal is:

[E 10:37:54.546 NotebookApp] Error unpacking user from cookie: Extra data: line 1 column 2 (char 1)
[W 10:37:54.546 NotebookApp] Clearing invalid/expired login cookie username-localhost-8865
[I 10:37:54.546 NotebookApp] 302 GET /server-proxy/servers-info (::1) 1.370000ms
[W 10:37:57.644 NotebookApp] Forbidden
[W 10:37:57.644 NotebookApp] 403 GET /api/contents/Documents?type=directory&_=1675507074408 (::1) 0.570000ms referer=http://localhost:8865/tree/Documents
[W 10:38:06.725 NotebookApp] Forbidden
[W 10:38:06.726 NotebookApp] 403 GET /api/sessions?_=1675507074409 (::1) 0.660000ms referer=http://localhost:8865/tree/Documents
[W 10:38:06.726 NotebookApp] Forbidden
[W 10:38:06.727 NotebookApp] 403 GET /api/terminals?_=1675507074410 (::1) 0.680000ms referer=http://localhost:8865/tree/Documents

I can run “jupyter lab” however and this runs fine (but of course is not such a nice environment to work in!)

Can anyone point me in the right direction for troubleshooting?



do you solve this problem? I meet it toooo, so confused!

Hi - no, afraid not!

I switched to JupyterLab however, and that is running fine (although I find I need to use Chrome instead of Safari on Mac!).

The warning disappeared after I disabled password of jupyter.
I want to use jupyter in pycharm beacuse it has code completion. But I tried many times and it seems didn’t work well. I can edit the file and run the code but it still has errors.

1 Like

Here’s the upstream issue: there are a few workarounds available for now.