I have a user that is saying that his Jupyter connections are timing out often like 3-4 times each day. We are running Ondemand 2.0.20. Any ideas on how to diagnose what is causing the timeouts? We also have an instance of Ondemand 1.18 where he does not get the timeouts.
Hi, sorry for the issue. To be sure on my end, this is just a single user having the issue? Are you seeing anything in the logs from apache or ondemand-nginx in /var/log/ondemand-nginx for that user when this happens?
What browser are they using, and has it been the same behavior across more than just the one browser?
Right now it is just this one user since the 2.0.20 version is not open to everybody. I did not see any errors in the /var/ondemand-nginx/user/error.log and there are no errors in the /var/log/httpd24/error.log file. I will find out which browser he is using and ask if he can replicate the issue with another browser.
He has been using Google Chrome and will see if he can reproduce it with Firefox.
The user tried firefox and is not seeing an issue with the timeouts. The Chrome version he is using is 98.0.4758.80.
I would check the logs from Jupyter. It’s in their session directory like ~/ondemand/data/sys/dashboard/batch_connect/sys/bc_osc_jupyter.
Find a session that was giving them trouble and look to see if Jupyter is throwing any errors. My guess is - is that it’s Jupyter itself that’s timing out trying to respond and not apache. Otherwise you’d have something more systemic going on.
Are you on a current version of Jupyter?
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.