JupyterLab is suddenly slow. Has anyone got the same situation or know the solution?

Hello Everyone!

I’m using jupyterlab on our lab’s CentOS system server, and the jupyterlab(v3.3.2), python and R kernel were all installed in a codna environment built by conda 3.0. Normally I ssh to the server and port forwarding the jupyterlab port to the localhost port, and it works just well for the last 6 months

Recently I just create a new conda env with conda 4.0+ and install the jupyterlab 3.4.4. It can be ported but it works very slow all of a sudden, like it will takes 15 to 20 seconds to change to the Render View( shift + R) or move one cell to another position, or split cells.

Also these kind of things happened when I swich back to the old conda env and jupyterlab 3.3.2, and also on Edge and Chrome browser. So I cant figure out the problem.

Has any one got the same issues? Or know how to solve this problem?

Thanks!

1 Like

I was seeing some performance degradation recently, but could not narrow it down. Does it also happen in Firefox? For me Firefox was not affected.

Hi! Thannks for your reply!

In my situation, those thing also happened when I use Firefox.

And also last time I didn’t mentioned, because I arrange the jupyter-lab in one of the Computation Node under the Management Node(which means I can ssh to the computation one from the management), I have to use ProxyCommand when I port forwarding.

So one of my colleague suggest me to use the Tunneling funtion in Xshell software, and I do so. It turns out jupyter-lab will be a little bit faster.

1 Like

Hi! Thannks for your reply!

In my situation, those thing also happened when I use Firefox.

And also last time I didn’t mentioned, because I arrange the jupyter-lab in one of the Computation Node under the Management Node(which means I can ssh to the computation one from the management), I have to use ProxyCommand when I port forwarding.

So one of my colleague suggest me to use the Tunneling funtion in Xshell software, and I do so. It turns out jupyter-lab will be a little bit faster.