i was wondering - consider i have 2 machines - client and server. the Client is opened to the internet, and sends very sensitive data to the server.
If fake data will be sent to the server and is not identified - this will lead to a catastrophic situation.
Client certificate / authentication will not be helpful, as this is an automatic process and a compromised client certificate/credentials will be easily grabbed.
I've read the following post: Authentication and removing the ability to replay information from a compromised machine and the answer suggested to use a "secret" that will be hard to replay from a compromised machine.
So, following this information, i can think of hiding some "secret" info from time to time (when generating the "real" data) and let the server check it - if it's ommited, the machine was compromised.
I was wondering if there are any other recommended methods in denying compromised machine data?