You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I recently had a weird bug and I am not sure how I can reproduce it.
I opened many notebook tabs in JupyterLab (18) and had a vertical split into to panes, each having 9 tabs. Then in one of the tabs in the right pane the Kernel was all the time "Starting" and "Restarting" (kernel status indication, that circle which is either filled or not), but I was able to execute code nevertheless. The weird thing then was, that no cell knew of any variables created in another (previously run) cell.
It was as if JupyterLab was confused which Kernel to use for which notebook / tab and my hidden state was maybe in another Kernel and not available in side the same notebook for other cells.
When I opened the notebooks, I opened up to 8 or 9 notebooks at the same time, marking them in the file browser and right click "open" them. Then I dragged the resulting tabs around into the two panes.
To Reproduce
Not sure, only ever happened once to me so far.
Expected behavior
JupyterLab should, even having many notebook tabs opened, still be able to run code in the correct kernel.
Screenshots
Not available.
Desktop (please complete the following information):
OS: Ubuntu 17.10
Browser: Firefox 61.0.1
JupyterLab: 0.35.3
Additional context
The notebooks I opened are not very large. Maybe a couple of pages in height, but no more than that.
The text was updated successfully, but these errors were encountered:
ZelphirKaltstahl
changed the title
Cells have independent Python context sometimes
Cells have independent Python contexts sometimes
Dec 17, 2018
Thanks for posting. When you say "one of the tabs in the right pane the Kernel was all the time "Starting" and "Restarting" (kernel status indication, that circle which is either filled or not)", it sounds like that single notebook kept restarting its kernel for some reason? If that is the case, it's entirely consistent that each cell might have been executed in a new restarted kernel. In this case, it wouldn't have been JupyterLab getting confused, it would have been a kernel that was sporadically restarting.
It would be good to get a reproducible case to see which situation it indeed was.
Describe the bug
I recently had a weird bug and I am not sure how I can reproduce it.
I opened many notebook tabs in JupyterLab (18) and had a vertical split into to panes, each having 9 tabs. Then in one of the tabs in the right pane the Kernel was all the time "Starting" and "Restarting" (kernel status indication, that circle which is either filled or not), but I was able to execute code nevertheless. The weird thing then was, that no cell knew of any variables created in another (previously run) cell.
It was as if JupyterLab was confused which Kernel to use for which notebook / tab and my hidden state was maybe in another Kernel and not available in side the same notebook for other cells.
When I opened the notebooks, I opened up to 8 or 9 notebooks at the same time, marking them in the file browser and right click "open" them. Then I dragged the resulting tabs around into the two panes.
To Reproduce
Not sure, only ever happened once to me so far.
Expected behavior
JupyterLab should, even having many notebook tabs opened, still be able to run code in the correct kernel.
Screenshots
Not available.
Desktop (please complete the following information):
Additional context
The notebooks I opened are not very large. Maybe a couple of pages in height, but no more than that.
The text was updated successfully, but these errors were encountered: