Recently, I started working on some Raspberry Pi projects, but the most frustrating part of that experience was to access the raspberry pi during development.
I had the following options:
- Use the Television at my place as a monitor over HDMI (not great pixel quality) and use wired keyboard and mouse to control. This was troublesome since I could only develop when I’m at my place and the Television is free.
- Be on the same network and ssh into the PI. This worked for a while since I could also access the ports of RPi, so I could host my applications on a certain port and access them over the RPi’s IP address.
The second option is obviously the way to go, but this needs some refining.
- Be on the same network to access ssh and other ports.
This can be solved using Ngrok. For more info:
- Access persistent shell sessions (in case of power or network drop)
This is why I needed to write a script
- Access the internet as if being browsed from the RPi (this is a use case for some future projects). Simplest solution: ssh -L 80:remotehost:80 user@myserver
Limitations of Ngrok
A free account on NGrok has the following limitations:
- HTTP/TCP tunnels on random URLs/ports
- 1 online ngrok process
- 4 tunnels/ngrok process
- 40 connections / minute
The most troublesome of the above limitations is the random port. Every time I reconnect to ngrok, I receive a different port to which I need to connect. I could buy a premium account but there’s a simple solution for this.
- Create a server
- Accept beacons from devices
- Show currently available devices
- Timeout devices when no beacon received
- Create a client
- Easy and scriptable installation
- Auto-start on boot
- Bundle Ngrok binary
- Upon ngrok start, access the available tunnels
- Send beacon to server with tunnel details
This way, any time I want to access a device, I just need to pull down a repository and set it to run on boot. This in turn will keep updating my server on the availability of my device and the port I can access it on.
If I want to access a different port, maybe over http protocol instead of tcp (ssh is supported over tcp), I can just edit a global variable in the pulled down code and restart the device. The new tunnel should show up on the server once the device is up and running.
Why not use a VPN?
In cases like my office network, I might need to be already connected to a VPN or some other network proxy. Configuring my development client device to connect to my home VPN might not always be an option.
Working on it…