slt is a dead-simple TLS reverse-proxy with SNI multiplexing (TLS virtual hosts).

That means you can send TLS/SSL connections for multiple different applications to the same port and forward them all to the appropriate backend hosts depending on the intended destination.

slt multiplexes connections to a single TLS port by inspecting the name in the SNI extension field of each connection.

You configure slt with a simple YAML configuration file:

bind_addr: ":443"

user: pi

frontends:
  v1.example.com:
    backends:
      -
        addr: ":4443"

  v2.example.com:
    backends:
      -
        addr: "192.168.0.2:443"
      -
        addr: "192.168.0.1:443"

Sometimes, you don't actually want to terminate the TLS traffic, you just want to forward it elsewhere. slt only terminates the TLS traffic if you specify a private key and certificate file like so:

frontends:
  v1.example.com:
    tls_key: /path/to/v1.example.com.key
    tls_crt: /path/to/v1.example.com.crt

slt performs simple round-robin load balancing when more than one backend is available (other strategies will be available in the future):

frontends:
  v1.example.com:
    backends:
      -
        addr: ":8080"
      -
        addr: ":8081"

Running slt is also simple. It takes a single argument, the path to the configuration file:

./slt /path/to/config.yml

Just cd into the directory and "go build". It requires Go 1.1+.

Just cd into the directory and "go test".

I run slt in production handling hundreds of thousands of connections daily.

Apache