Hi, I have recently set up Darcy Cloud with my nvidia jetson nano and everything works great!
However, I wanted do some minor testing of my code by setting up a Jupyter server on my jetson and accessing it remotely from another device. This requires setting up a tunnel and forwarding of ports which is sometimes restricted by the nature of one or both of the networks that I (and the jetson) am connected to.
Is there any way this can be done via the edgectl ssh service? I just don’t want to be updating/sending several application files for these minor changes. Retrieving the jetson from its location can be bothersome at times, too.
At this moment, what you are asking can’t be done yet. If you follow us on social media or are signed up to our newsletter you will get a notification when we release our latest feature that should be a solution to your issue.
In the meantime, you can access the Nano via Darcy’s SSH in the Cloud portal or CLI. From the Nano’s SSH connection you can access your Jupyter server.
Yes, I can Darcy SSH into the Nano and initiate a Jupyter server there. However, I am having trouble accessing that server from my side (from the client side), which I understand is a feature of the future. I am not sure how to use the SSH connection to access it, as you said.
With Darcy’s SSH feature you can SSH into the Nano from any network regardless of location, pending both Nano and the machine you are connecting from are on the internet. That should eliminate the issue of having to retrieve the Nano for anything.
If you do this, I am unsure if you need to the Jupyter server if you can remotely access the device from anywhere & any network. If you do need it, my suggestion would then require you to access the server from the Nano, after having SSH-ed into the Nano.
I hope I was able to answer your question. I will also ask the team if they have any input to share as well.