![]() The worker ID right now is typically based off the address of an object in memory. This could be guessed. While the worker is tied to a specific IP, there is a chance an off-path attacker could be hosted behind the same IP as the caller. They could possibly guess the worker id of an unclaimed session by observing the sequence of IDs presented to themselves, leading to them gaining access to an already authenticated SSH session. Use the python secrets module to generate a cryptographically secure token to use as the worker ID. This shoud be much harder to guess. |
||
---|---|---|
.github | ||
preview | ||
tests | ||
webssh | ||
.coveragerc | ||
.dockerignore | ||
.gitignore | ||
.travis.yml | ||
Dockerfile | ||
LICENSE | ||
MANIFEST.in | ||
README.md | ||
README.rst | ||
docker-compose.yml | ||
requirements.txt | ||
run.py | ||
setup.cfg | ||
setup.py |
README.md
WebSSH
Introduction
A simple web application to be used as an ssh client to connect to your ssh servers. It is written in Python, base on tornado, paramiko and xterm.js.
Features
- SSH password authentication supported, including empty password.
- SSH public-key authentication supported, including DSA RSA ECDSA Ed25519 keys.
- Encrypted keys supported.
- Two-Factor Authentication (time-based one-time password) supported.
- Fullscreen terminal supported.
- Terminal window resizable.
- Auto detect the ssh server's default encoding.
- Modern browsers including Chrome, Firefox, Safari, Edge, Opera supported.
Preview
How it works
+---------+ http +--------+ ssh +-----------+
| browser | <==========> | webssh | <=======> | ssh server|
+---------+ websocket +--------+ ssh +-----------+
Requirements
- Python 2.7/3.4+
Quickstart
- Install this app, run command
pip install webssh
- Start a webserver, run command
wssh
- Open your browser, navigate to
127.0.0.1:8888
- Input your data, submit the form.
Server options
# start a http server with specified listen address and listen port
wssh --address='2.2.2.2' --port=8000
# start a https server, certfile and keyfile must be passed
wssh --certfile='/path/to/cert.crt' --keyfile='/path/to/cert.key'
# missing host key policy
wssh --policy=reject
# logging level
wssh --logging=debug
# log to file
wssh --log-file-prefix=main.log
# more options
wssh --help
Browser console
// connect to your ssh server
wssh.connect(hostname, port, username, password, privatekey, passphrase, totp);
// pass an object to wssh.connect
var opts = {
hostname: 'hostname',
port: 'port',
username: 'username',
password: 'password',
privatekey: 'the private key text',
passphrase: 'passphrase',
totp: 'totp'
};
wssh.connect(opts);
// without an argument, wssh will use the form data to connect
wssh.connect();
// set a new encoding for client to use
wssh.set_encoding(encoding);
// reset encoding to use the default one
wssh.reset_encoding();
// send a command to the server
wssh.send('ls -l');
Custom Font
To use custom font, put your font file in the directory webssh/static/css/fonts/
and restart the server.
URL Arguments
Support passing arguments by url (query or fragment) like following examples:
Passing form data (password must be encoded in base64, privatekey not supported)
http://localhost:8888/?hostname=xx&username=yy&password=str_base64_encoded
Passing a terminal background color
http://localhost:8888/#bgcolor=green
Passing a terminal font color
http://localhost:8888/#fontcolor=red
Passing a user defined title
http://localhost:8888/?title=my-ssh-server
Passing an encoding
http://localhost:8888/#encoding=gbk
Passing a font size
http://localhost:8888/#fontsize=24
Passing a command executed right after login
http://localhost:8888/?command=pwd
Passing a terminal type
http://localhost:8888/?term=xterm-256color
Use Docker
Start up the app
docker-compose up
Tear down the app
docker-compose down
Tests
Requirements
pip install pytest pytest-cov codecov flake8 mock
Use unittest to run all tests
python -m unittest discover tests
Use pytest to run all tests
python -m pytest tests
Deployment
Running behind an Nginx server
wssh --address='127.0.0.1' --port=8888 --policy=reject
# Nginx config example
location / {
proxy_pass http://127.0.0.1:8888;
proxy_http_version 1.1;
proxy_read_timeout 300;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Host $http_host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Real-PORT $remote_port;
}
Running as a standalone server
wssh --port=8080 --sslport=4433 --certfile='cert.crt' --keyfile='cert.key' --xheaders=False --policy=reject
Tips
- For whatever deployment choice you choose, don't forget to enable SSL.
- By default plain http requests from a public network will be either redirected or blocked and being redirected takes precedence over being blocked.
- Try to use reject policy as the missing host key policy along with your verified known_hosts, this will prevent man-in-the-middle attacks. The idea is that it checks the system host keys file("~/.ssh/known_hosts") and the application host keys file("./known_hosts") in order, if the ssh server's hostname is not found or the key is not matched, the connection will be aborted.