SNItch -- the SNI-based Switch
This tool switches incoming TLS connections based on the SNI contained in them. It is assumed that the full SNI extension fits in the first record transmitted.
Read more about the architecture of SNItch.
The SNItch is made to listen to an address and port, which default to any address and port number 443, respectively. Use -l to override the address and -p to override the port. Addresses are interpreted as IPv6 addresses, but you may place :: in front of an IPv4 address if you like.
The configuration file is assumed to live at /etc/snitch.conf and if not, the -c option can be used to introduce another filename.
The configuration file the explains how forwarding takes place. Any line that does not start with whitespace or a # character must be of the following format:
label inthost intport [flags...]
Each of the phrases is separated by whitespace. Trailing whitespace is optional, and will be ignored. So, it is okay to end a line immediately after the port number. It is not acceptable to start a configuration line with whitespace. None of the terms mentioned above may contain a space, and with the exception of [flags...] none of them is empty.
The label is the name used in SNI. It may be a DNS-published name, or something internal if both ends see fit to using that.
The inthost is an IPv6 address of an internal host. Once again, prefix IPv4 addresses with :: if you have a nostalgic mood.
The intport is a port number to connect to.
The optional [flags...] are whitespace-separate words that detail what needs to be done with the traffic while in transit. For now, there are no flags defined.