Installation

Install your own server as systemd service using a simple to run install script

Script is hosted on Techahold and supported on our Discord.

Currently the script will download and setup the Relay and Signal Servers (hbbr and hbbs), generate configs and host them on a password protected web page for simple deployment to clients.

Requirements

You need to have Linux installed, script is tested working with CentOS Linux 7/8, Ubuntu 18/20 and Debian. A server with 1 CPU, 1 GB RAM and 10 GB disk is plenty to run RustDesk.

How to Install the server

Please setup your firewall on your server prior to running the script.

Make sure you have got access via SSH or otherwise setup prior setting up the firewall. The example commands for UFW (Debian based) are:

ufw allow proto tcp from YOURIP to any port 22

If you have UFW installed use the following commands to configure the firewall (port 8000 only needed if you want to use the auto generated install files):

ufw allow 21115:21119/tcp
ufw allow 8000/tcp
ufw allow 21116/udp
sudo ufw enable

Run the following commands:

wget https://raw.githubusercontent.com/techahold/rustdeskinstall/master/install.sh
chmod +x install.sh
./install.sh

There is also an update script on Techahold’s repository.

Install your own server as systemd service using deb file for debian distros

Please Download deb files yourself and install with apt-get -f install <filename>.deb or dpkg -i <filename>.deb.

Set up your own server instance manually.

STEP 1: Download server-side software programs

Download.

Platform versions provided:

  • Linux
  • Windows

The tutorial below is based on Linux build.

There are two executables and a folder:

  • hbbs - RustDesk ID/Rendezvous server
  • hbbr - RustDesk Relay server

They are built on CentOS Linux 7, tested on CentOS Linux 7/8 and Ubuntu 18/20.

Server Requirements

The hardware requirements are very low; the minimum configuration of a basic cloud server is enough, and the CPU and memory requirements are minimal. You can also use a Raspberry Pi or something similar. Regarding the network size, if the TCP hole punching direct connection fails, the relay traffic will be consumed. The traffic of a relay connection is between 30 K/s and 3 M/s (1920x1080 screen) depending on the resolution settings and screen update. If it is only for office work demand, the traffic is around 100 K/s.

STEP 2: Run hbbs and hbbr on your server

We suggest you use PM2 for managing your service.

Option 1

Run hbbs/hbbr without PM2.

./hbbs -r <relay-server-ip[:port]>
./hbbr
Option 2

Run hbbs/hbbr with PM2.

pm2 start hbbs -- -r <relay-server-ip[:port]>
pm2 start hbbr

PM2 requires Node.js v16+, if you fail to run PM2 (e.g. you can not see hbbs/hbbr in pm2 list), please download and install the Node.js LTS version from https://nodejs.org. If you want to make hbbs/hbbr auto-run after reboot, please check out pm2 save and pm2 startup. More about PM2. Another good tool for your logs is pm2-logrotate.

The -r parameter of hbbs is not mandatory, it is just convenient for you not to specify a relay server on the controlled client side. You do not need to specify port if you are using default 21117 port. The relay server specified by the client has a higher priority than this.

By default, hbbs listens on 21115 (TCP), 21116 (TCP/UDP) and 21118 (TCP), hbbr listens on 21117 (TCP) and 21119 (TCP). Be sure to open these ports in the firewall. Please note that 21116 should be enabled both for TCP and UDP. 21115 is used for the NAT type test, 21116/UDP is used for the ID registration and heartbeat service, 21116/TCP is used for TCP hole punching and connection service, 21117 is used for the Relay services, and 21118 and 21119 are used to support web clients. If you do not need web client (21118, 21119) support, the corresponding ports can be disabled.

  • TCP (21115, 21116, 21117, 21118, 21119)
  • UDP (21116)

Please run with the -h option to see help if you want to choose your own port.

STEP 3: Set hbbs/hbbr address on client-side

Key

Different from the old version, the key in this version is mandatory, but you don’t need to set it yourself. When hbbs runs for the first time, it will automatically generate a pair of encrypted private and public keys (respectively located in the id_ed25519 and id_ed25519.pub files in the running directory), whose main purpose is for communication encryption.

If you did not fill in the Key: (the content in the public key file id_ed25519.pub) in the previous step, it does not affect the connection, but the connection cannot be encrypted.

cat ./id_ed25519.pub

If you want to prohibit users without the key from establishing non-encrypted connections, please add the -k _ parameter when running hbbs and hbbr, for example:

./hbbs -r <relay-server-ip[:port]> -k _
./hbbr -k _

If you want to change the key, remove the id_ed25519 and id_ed25519.pub files and restart hbbs/hbbr, hbbs will generate a new key pair.

If you are using docker-compose and keys don’t exist, the start of containers will create different keys in hbbs and hbbr folders.

You could create keys manually in hbbs and copy them to hbbr before starting the containers.

Or you could stop the hbbr container and copy the keys from hbbs to the hbbr folder, and then restart the container.