20 KiB
Introduction
This is an A to Z guide that will help you get WeTTY up and running on a Debian based system. It covers the key configuration areas by using copy and paste commands. This will help you install this application and get it securely up and running with minimal system interference and reversible changes. It should also provide enough information to allow you to understand and extend that configuration for your personal requirements.
Note: Some of these configurations are optional, such as self signed SSL and public key authentication. The purpose of the guide is to show you how to correctly understand, configure, install and use these options should you wish to use them but they are not required to use WeTTY in general.
Required dependencies
Node
- WeTTY requires node v14 or greater. We will install this locally for a non root user later in the guide.
python
- This should be installed by default but we will include it in our apt-get
command to be safe.
build-essential
- We need this specifically for node-gyp
to build packages when using npm
or yarn
to install packages.
As the root
user run these commands:
apt update
apt install -y build-essential python
If you have no root access and just want to check the dependencies are installed you can use these commands:
dpkg -s python | grep Status:
dpkg -s build-essential | grep Status:
If the package is installed you will see this result:
Status: install ok installed
Create a local user account
For this guide, unless specifically stated, you should not use a root
account to install and run WeTTY. Please use an existing local account or create one now.
Note: Whichever user runs WeTTY should be the same user you wish to authenticate with via ssh
to keep this guide simple.
If you need to create a local user account you can run this command:
Important note: replace username
with a user name of your choosing and create a password when prompted
adduser --gecos "" username
Switch to your local user now and open an ssh
session to continue with this guide.
Install node locally
To install and manage node
as a local user we are going to use Node Version Manager. This is an established solution for installing and managing multiple versions of node without needing root
access. This will allow you to install and use multiple versions of node
at the same time.
This command will download and install nvm
and reload our shell.
bash <(curl -s https://raw.githubusercontent.com/nvm-sh/nvm/master/install.sh) && source ~/.profile
This command will install the latest version of the v14 branch, which is the minimum required version for WeTTY.
nvm install 14
You can now call node
to check it works using this command.
node -v
Your result should look something like this.
v14.14.0
Note: There is an important consideration with the nvm
method. node
is only in the local user's path through sourcing of the ~/.nvm/nvm.sh
which is done when the user logs in and the shell sources the user's .bashrc
file. So for some applications who are not aware of this local shell environment node
will not be usable unless we provide a full path and nvm
commands will also be unavailable. The way we over come this issue for the needs of this guide is by using this command substitution to provide the full path, where applicable:
$(source ~/.nvm/nvm.sh && nvm which 14)
Why? This command will always provide us with the path to the most current version of node 14
installed via nvm
regardless of other versions of node
installed with nvm
.
Generate OpenSSL certificates
Why? So that later we can configure WeTTY to work with https
and make sure we interact with WeTTY over a secure connection at all times.
Make the required directory using this command:
mkdir -p ~/.ssl
Generate the self signed openssl
certificates we will use to encrypt our web traffic when using WeTTY using this command:
Note: we are using ecdsa
using the secp384r1
curve. Tested to be compatible with Chrome and Firefox browsers.
openssl req -x509 -nodes -days 1095 -newkey ec:<(openssl ecparam -name secp384r1) -subj "/C=GB/ST=None/L=None/O=None/OU=None/CN=None" -out ~/.ssl/wetty.crt -keyout ~/.ssl/wetty.key
Now give these file and folders the correct permissions using these commands:
chmod 700 ~/.ssl
chmod 644 ~/.ssl/wetty.crt
chmod 600 ~/.ssl/wetty.key
This is all we need to do for now in regards to https.
Generate the ssh key file
Why? So that later we can set up automatic login via ssh
. Our instance will authorise using this key file stored locally.
Make the required directory, if it does not exist, using this command:
mkdir -p ~/.ssh
Create the ssh
private key using ed25519
that we need to authorise our local connection, using this command:
ssh-keygen -q -C "wetty-keyfile" -t ed25519 -N '' -f ~/.ssh/wetty 2>/dev/null <<< y >/dev/null
Important Note: You must add the public key to your authorized_keys
file in order to be able to log in using your ssh
key file when accessing WeTTY via a web browser.
Copy the key to our ~/.ssh/authorized_keys
file, using this command:
cat ~/.ssh/wetty.pub >> ~/.ssh/authorized_keys
Now give these file and folders the correct permissions, using these commands:
chmod 700 ~/.ssh
chmod 644 ~/.ssh/authorized_keys
chmod 600 ~/.ssh/wetty
Optional: A housekeeping command. If you need to remove all entries of the WeTTY public key with the comment wetty-keyfile
from the ~/.ssh/authorized_keys
file use this command. Otherwise ignore this.
sed -r '/^ssh-ed25519(.*)wetty-keyfile$/d' -i ~/.ssh/authorized_keys
Install WeTTY
Note: we are using -g
for npm
or global
for yarn
along with --prefix ~/
so that the application's symbolic link is installed to our ~/bin
directory and available in our local user's PATH
.
As your local user run these commands:
To make sure the local user's ~/bin
directory exists and is in the PATH
please run the following command.
mkdir -p ~/bin && source ~/.profile
Now use npm
to install the yarn
packet manager.
npm install -g yarn --prefix ~/
Then use yarn
to install wetty
.
yarn global add wetty --prefix ~/
Once successfully installed the application should be available in your local user's PATH
. To test the installation was successful please use this command:
wetty -h
Accessing the web interface via our external IP
If you are using your external IP and not a domain to access WeTTY this step needs to be done here because it is not easy to do in the next steps if WeTTY is running in the terminal.
This command will generate the correct URL you need to visit after using the start up commands in the following section.
echo https://$(curl -s4 icanhazip.com):3000
Please make make a note of this URL now.
Running WeTTY
Now we have all the ground work done we can focus on our WeTTY server configuration settings.
For example, the below command would provide a https
instance with automatic ssh
authorisation using our wetty
private key on port 3000
accessible at https://IP:3000
.
Important note: This command will run in your current terminal session and not in the background. The key combination of CTRL
+ c
will exit the application.
wetty --host 0.0.0.0 -port 3000 --title wetty -base / --ssh-key ~/.ssh/wetty --ssh-host localhost --ssh-user $(whoami) --ssh-port 22 --ssh-auth publickey --ssl-key ~/.ssl/wetty.key --ssl-cert ~/.ssl/wetty.crt
Since you may not need all these settings we will look through what each one does below so that you can decide how to best configure your instance.
Environment settings explained
Let's break it down so that we can understand what's being done and why.
--host 0.0.0.0 -p 3000 --title wetty --base /
--host 0.0.0.0
- defines the interface we want to bind to. Using 0.0.0.0
means that we bind to all available interfaces so using this setting just works. When we use nginx we can change this to --host 127.0.0.1
in order to prevent generic port access to the application and force traffic through our nginx reverse proxy URL.
-p 3000
- defines the local listening port. You will use this port to connect via the remotely accessible web server or when configuring a reverse proxy through nginx.
--title wetty
- an optional setting to set the window title for this wetty
session.
--base /
- changes the default base URL setting from /wetty/
to define the remote URL. We use --base /
to make wetty
accessible on the URL format https://IP:3000
instead of https://IP:3000/wetty
but we would change this back if we use nginx to reverse proxy the application.
SSH settings explained
These settings are all specific to ssh
and will enable you to automatically log into you ssh
session for the selected user.
--ssh-key ~/.ssh/wetty --ssh-host localhost --ssh-user $(whoami) --ssh-port 22 --ssh-auth publickey
--ssh-key ~/.ssh/wetty
- we are telling WeTTY to load our ssh
key file that we generated earlier.
--ssh-host localhost
- optional setting telling WeTTY to connect the host localhost
--ssh-user $(whomai)
- defines our ssh
username. In this case via the command substitution of whoami
which will not require your input of a username.
--ssh-port 22
- optional setting to set the ssh
port we need to connect to.
--ssh-auth publickey
defines the accepted authentication types. You do not have to use the key file and you can instead require a password but setting this to --sshauth password
. You can specify both --sshauth publickey,password
--ssh-config configfile
- (not used for this guide) alternative ssh configuration file. From ssh(1):
If a configuration file is given on the command line, the system-wide configuration file (/etc/ssh/ssh_config) will be ignored. The default for the per-user configuration file is ~/.ssh/config.
SSL settings explained
These settings are specific to openssl
to make WeTTY load https webserver so that all data is transmitted over a secure connection.
--ssl-key ~/.ssl/wetty.key --ssl-cert ~/.ssl/wetty.crt
--ssl-key ~/.ssl/wetty.key
- tells WeTTY to load our openssl
generated key file.
--ssl-cert ~/.ssl/wetty.crt
- tells WeTTY to load our openssl
generates certificate file.
Optional - load settings via a configuration file
As of WeTTY v2 there is official support for a configuration file used with the flag --conf
to specify the location of this file.
Create the directory where we will store this configuration file.
mkdir -p ~/.config/wetty
Use nano
to open a file for editing.
nano ~/.config/wetty/config.json
Here is the template config.json
you need to use.
Note: To be validated json the below json example should have the // ...
comments removed. With all comments removed the example is valid json. They are in the example to help explain the options and won't stop wetty from loading if you leave them in place. Lines you do not need can be commented out but should be removed if you want the json to pass validation.
"ssh": {
"user": "username", // default user to use when ssh-ing
"host": "localhost", // Server to ssh to
"auth": "publickey,password", // shh authentication, method. Defaults to "password", you can use "publickey,password" instead'
"pass": "password", // Password to use when ssh-ing
"key": "/home/username/.ssh/wetty", // path to an optional client private key, connection will be password-less and insecure!
"port": 22, // Port to ssh to
"knownHosts": "/dev/null" // ssh knownHosts file to use
},
"server": {
"base": "/wetty/", // URL base to serve resources from
"port": 3000, // Port to listen on
"host": "0.0.0.0", // listen on all interfaces or can be 127.0.0.1 with nginx
"title": "WeTTy - The Web Terminal Emulator", // Page title
"bypassHelmet": false // Disable Helmet security checks
},
"forceSSH": false, // Force sshing to local machine over login if running as root
"command": "login", // Command to run on server. Login will use ssh if connecting to different server
"ssl": {
"key": "/home/username/.ssl/wetty.key",
"cert": "/home/username/.ssl/wetty.crt"
}
}
Press ctrl
+ x
and then press y
to save then press enter
to confirm and exit nano
.
System Environment Variables
Note: We will not be using this section to configure WeTTY. We are simply documenting it.
There are some environment variables you can export that can be used by WeTTY to configure an instance.
BASE
PORT
TITLE
SSHUSER
SSHHOST
SSHAUTH
SSHPASS
SSHKEY
SSHPORT
KNOWNHOSTS
FORCESSH
COMMAND
These can be used in the following way
export PORT=3000
There are currently no environment settings for variables not listed above.
Systemd service settings
We will use a local user systemd
service file to manage the wetty
service.
First, create the required directory, if it does not exist.
mkdir -p ~/.config/systemd/user
Systemd service
Here is an example template of how to use service file with hardcoded values you can set in the wetty.service
file with all options enabled.
Use nano
to open a file for editing.
nano ~/.config/systemd/user/wetty.service
The copy and paste this code.
Note: This is an example service file based on all the options documented and configured so far. You may not want all these option enabled so please remove or modify the ExecStart
command based on your needs.
[Unit]
Description=wetty
After=network-online.target
[Service]
Type=simple
ExecStart=/bin/bash -c "$$(source /home/$$(whoami)/.nvm/nvm.sh && nvm which 12) /home/$$(whoami)/bin/wetty --host 0.0.0.0 -p 3000 --title wetty --base / --ssh-key /home/$$(whoami)/.ssh/wetty --ssh-host localhost --ssh-user $$(whoami) --ssh-port 22 --ssh-auth publickey --ssl-key /home/$$(whoami)/.ssl/wetty.key --ssl-cert /home/$$(whoami)/.ssl/wetty.crt"
Restart=always
RestartSec=2
TimeoutStopSec=5
SyslogIdentifier=wetty
[Install]
WantedBy=default.target
Press ctrl
+ x
and then press y
to save then press enter
to confirm and exit nano
.
Optional - Systemd service with config file
Here is the example using our pseudo configuration file. All modifications to the start up of wetty
will be done by editing the ~/.config/Wetty/config
file and then reloading the wetty.service
.
Use nano
to open the file for editing.
nano ~/.config/systemd/user/wetty.service
The copy and paste this code.
Note: This ExecStart
assumes the location of your config.json
to be ~/.config/wetty/config.json
. Please make sure you use the correct location for this file.
[Unit]
Description=wetty
After=network-online.target
[Service]
Type=simple
ExecStart=/bin/bash -c "$$(source /home/$$(whoami)/.nvm/nvm.sh && nvm which 14) /home/$$(whoami)/bin/wetty --conf /home/$$(whoami)/.config/wetty/config.json"
Restart=always
RestartSec=2
TimeoutStopSec=5
SyslogIdentifier=wetty
[Install]
WantedBy=default.target
Press ctrl
+ x
and then press y
to save then press enter
to confirm and exit nano
.
Activating your service
The you can enable and start your service.
systemctl --user enable --now wetty
Managing your services
These commands will help you manage your service.
systemctl --user daemon-reload
systemctl --user status wetty
systemctl --user start wetty
systemctl --user stop wetty
systemctl --user restart wetty
systemctl --user disable --now wetty
systemctl --user enable --now wetty
Nginx reverse proxy
If you want to use nginx as a reverse proxy here is the configuration file you can use.
Please modify these specific environment settings:
Why? This will disable generic port access to the application and force traffic via the nginx reverse proxy.
--host 127.0.0.1
Why? This change is so that our application does not attempt to load as the web root of /
for nginx.
--base /wetty/
Now you can use this nginx configuration file.
Note: we are using https
with https://127.0.0.1:3000/wetty;
because we configured wetty
to run via https
using our self signed ssl certificates. If you chose not to run WeTTY with a self signed certificate you should changes this to http://127.0.0.1:3000/wetty;
The copy and paste this into the https
server block of your enable server configuration file.
location /wetty {
proxy_pass https://127.0.0.1:3000/wetty;
#
proxy_pass_request_headers on;
#
proxy_set_header Host $host;
#
proxy_http_version 1.1;
#
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_set_header X-Forwarded-Protocol $scheme;
proxy_set_header X-Forwarded-Host $http_host;
proxy_set_header X-NginX-Proxy true;
#
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection $http_connection;
proxy_read_timeout 43200000;
proxy_set_header X-Forwarded-Ssl on;
#
proxy_redirect off;
proxy_buffering off;
}
Press ctrl
+ x
and then press y
to save then press enter
to confirm and exit nano
Now you would need to reload nginx service using this command:
systemctl restart nginx
Accessing the web interface via nginx
Visit the URL format https://YourIPorDomain/wetty
and you can access WeTTY.
This command will generate the correct URL you need to visit it you are not using a domain.
echo https://$(curl -s4 icanhazip.com)/wetty
Protecting your instance of WeTTY
Disclaimer: It is not recommended by this guide that you run an instance of WeTTY on your server with no access control in place.
If you chose to not use a password to login in you should protect your instance behind either:
2: Authelia
Configuration reference
wetty -h
configuration options for reference.
--help, -h Print help message [boolean]
--version Show version number [boolean]
--conf config file to load config from [string]
--ssl-key path to SSL key [string]
--ssl-cert path to SSL certificate [string]
--ssh-host ssh server host [string] [default: "localhost"]
--ssh-port ssh server port [number] [default: 22]
--ssh-user ssh user [string] [default: ""]
--title window title [string] [default: "WeTTy - The Web Terminal Emulator"]
--ssh-auth defaults to "password", you can use "publickey,password"
instead [string] [default: "password"]
--ssh-pass ssh password [string]
--ssh-key path to an optional client private key (connection will be
password-less and insecure!) [string]
--ssh-config Specifies an alternative ssh configuration file. For further
details see "-F" option in ssh(1) [string] [default: ""]
--force-ssh Connecting through ssh even if running as root [boolean] [default: false]
--known-hosts path to known hosts file [string]
--base, -b base path to wetty [string] [default: "/wetty/"]
--port, -p wetty listen port [number] [default: 3000]
--host wetty listen host [string] [default: "0.0.0.0"]
--command, -c command to run in shell [string] [default: "login"]
--allow-iframe Allow wetty to be embedded in an iframe, defaults to allowing
same origin [boolean] [default: false]