I had this idea few days ago, im not sure if im wrong on everything or if this can be a good idea:
Custom Antiddos:
Player enter on website with his game account, go to control panel and click on button: LET ME PLAY.
This button (LET ME PLAY) will get automatic the IP from the web browser and add on the table of database: SRO_VT_ACCOUNT._AllowedIP.
This query wont work as INSERT INTO, will work as UPDATE and WHERE (when a guy register new account, the ID + IP are added to this table)
Then, with a virtual host with linux (and iptables):
Script for make a text file with all the IPs from SRO_VT_ACCOUNT._AllowedIP.
Iptables DENY ALL except IPs from that text file.
Make iptables redirect ALL to the host where files are. So this should work as 'proxy'.
And from the host where server files are need configure firewall for deny all and just accept ports (normally 15779,15881,15884) from the proxy linux host.
OR/AND:
Then, a little program as launcher, with a simple 'iframe' to the website where we added that button (LET ME PLAY). So when a player enter with user and password, his IP update at AllowedIP table and START button of launchers appear.
(This wont be a problem for Mbot or others since after login at launcher, you can open bot without problem (ports are open then).
That mean, no more DDOS, maybe DOS (but thats not hard to block), lets say we add a max petitions per IP (at iptables ofcourse).
Also, a procedure at the sql that is executed every 24h (it checks the last logout, if it happen 24hours ago, then delete it from AllowedIP table).
And also, only 1 IP per Account are allowed at _AllowedIP table (That mean every account will have his own allowed IP [Remember we are talking about the iptables, so actually we are talking about ports not about login ingame]).
(For prevent a program to create random accounts for example on infected pcs (with troyan), we add Captcha at create account.)
This custom antiddos will stop also the exploits from example on Agentserver?: ofcourse no, thats just for prevent 'botnets' and things like that. So you will still need to fix those bugs anyway.
And my noby question: Firewall (iptables) will be able to do this for example at Transport Layer? (I think yes but im not sure).
Custom Antiddos:
Player enter on website with his game account, go to control panel and click on button: LET ME PLAY.
This button (LET ME PLAY) will get automatic the IP from the web browser and add on the table of database: SRO_VT_ACCOUNT._AllowedIP.
This query wont work as INSERT INTO, will work as UPDATE and WHERE (when a guy register new account, the ID + IP are added to this table)
Then, with a virtual host with linux (and iptables):
Script for make a text file with all the IPs from SRO_VT_ACCOUNT._AllowedIP.
Iptables DENY ALL except IPs from that text file.
Make iptables redirect ALL to the host where files are. So this should work as 'proxy'.
And from the host where server files are need configure firewall for deny all and just accept ports (normally 15779,15881,15884) from the proxy linux host.
OR/AND:
Then, a little program as launcher, with a simple 'iframe' to the website where we added that button (LET ME PLAY). So when a player enter with user and password, his IP update at AllowedIP table and START button of launchers appear.
(This wont be a problem for Mbot or others since after login at launcher, you can open bot without problem (ports are open then).
That mean, no more DDOS, maybe DOS (but thats not hard to block), lets say we add a max petitions per IP (at iptables ofcourse).
Also, a procedure at the sql that is executed every 24h (it checks the last logout, if it happen 24hours ago, then delete it from AllowedIP table).
And also, only 1 IP per Account are allowed at _AllowedIP table (That mean every account will have his own allowed IP [Remember we are talking about the iptables, so actually we are talking about ports not about login ingame]).
(For prevent a program to create random accounts for example on infected pcs (with troyan), we add Captcha at create account.)
This custom antiddos will stop also the exploits from example on Agentserver?: ofcourse no, thats just for prevent 'botnets' and things like that. So you will still need to fix those bugs anyway.
And my noby question: Firewall (iptables) will be able to do this for example at Transport Layer? (I think yes but im not sure).