New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Jupyter won't open or trust notebooks that worked previously #174
Comments
I also created a python 2.7 conda env, and I get the same error. |
The file it's trying to open is Is it possible that your home directory is on an NFS mount? That seems to be the most common cause of issues with sqlite databases. We try to catch these errors and automatically recreate the database file, but maybe we're missing this one. |
Nope, that's one of the ones I tried moving. Tried again just now to make sure, and it didn't help. The file did get created again, but PRAGMA foreign_keys=OFF;
BEGIN TRANSACTION;
/**** ERROR: (10) disk I/O error *****/
ROLLBACK; -- due to errors
It appears to be Lustre [that's what I get when I do Anything I can do to help diagnose this? Or maybe just workaround signatures? |
Try configuring the notebook to store the trust database on a local scratch disk, or if that's not possible, the special The config option is |
Thanks, that works. Would it help to print a warning if a |
I think it could do something like we do with the IPython history database: if it fails once, move the database aside and try to recreate it (it already does this). If recreating it fails, switch to an in-memory database and give a warning. |
Ah, I see that's been in master for the past 5 months, but not in the version I currently get from conda.
I'm just creating a PR to try this. |
This error should be handled by the newest version of |
Thanks! |
I was getting the error because my disk was full. I moved some of the stuff and its fine now. |
For me, removing nbsignatures.db from directory '~/.local/share/jupyter' is work. |
This happened to me recently with old notebooks, and also wasn't able to create new notebooks. There were two problems: a) couldn't open old notebooks (unhandled error) I did two things: a) vim .jupyter/jupyter_notebook_config.py and then c.NotebookNotary.db_file = ':memory:' Now everything seems to work :) |
This still doesn't work on NFS mounts for:
|
I have encountered this error intermittently when running Jupyter in an official Docker image, specifically, with this command:
The error appears on the console when attempting to open an As of this point, it's not clear to me exactly what causes me to sometimes have the error occur and sometimes not have it occur. Sometimes opening a file causes the problem and sometimes it doesn't. Restarting the container seems to help. I'm on OS X using Chrome and Docker (as opposed to podman, etc.), in case those details are relevant. The particular image version (if I'm doing this right, with |
Just confirmed: When I get this error, shutting down the container and restarting it lets me open the file without errors. |
I'm having a problem with both jupyter notebook and just plain jupyter trust, which both seem to have the same underlying issue. Notebooks that were working previously are no longer working. I had updated everything with conda recently, but I tried just totally reinstalling conda (python 3.5) and get the same result.
I've found this ipython issue, which is slightly different (different error message), but in any case the solution to that doesn't solve this. In fact, I tried completely removing my ipython and jupyter profile directories, and nothing changed. I also found this traitlets issue that at least reports the same exception, but that was fixed a few months ago, so it shouldn't be an issue, right??? Finally, there's also this issue, but all my directories have the correct permissions and ownership.
If I try to load the notebooks, I see a pop-up saying
and the log says
If I try to trust the notebooks from the command line, I see something like this:
Any workaround or fix for this?
The text was updated successfully, but these errors were encountered: