I'm having a discussion with a software developer in my nears. He is using MySql direct connections for a software. There is no PHP/SSH/whatever-Bridge involved for the communication, neither at startup, while running or before logging out. The software goes from account login to management and so on.
As you may know, the account data can be read out by some hacking software e.g. Cheat Engine. Thats the point we discussed after I created a web-based readonly interface for some data from this database. There is no security and everyone could just access to it and make with the data (mail addresses, passwords: sadly just md5) what they want.
Now he told me that he configured the firewall and mysql server to I get mails from non-software (like my web interface) accesses to this database and an automatic backup. Now he is crying about a broken firewall and an overwhelmed mail server.
Edit:
Its a Debian-76-wheezy-64-LAMP
My question is: Is this possible?
My mind of this: No, its not as the client (with around 40 users) has always another IP and is logging in to the same mysql account as in my web interface. He is using that tactic to disallow complains about the mentioned security issues.
I appretiate any help on that question.
A little graphic:

iptablesor on any unix/linux platform usehosts.allow&hosts.denyfiles where you can allow or disallow specific IPs, Ports, Services etc.