R
Ryan
Guest
Last nights connection problems.
As many of you know we were experiencing some network issues starting at around 3:33:45 last night. These issues started with some minor denial of service attacks from IP's based out of England or the UK, France, USA, and Japan.
Our upstream provider decided to go ahead and implement a filter to block this attack, which was very affective. The attack which was two pronged, a standard "fill up your pipe DDoS" and a UDP flood, peaked at about 85 Megs of traffic and about 15k packets per second.
The game server itself runs about 15 - 20k packets per second anyway so the packets weren't a big deal, what happened was when the filter was implemented a monitoring system kicked in and kept launching TCP connections to the server on all available ports, however the monitor for port 2593 went crazy, and basically inadvertently became a denial of service attack on the game server, thus blocking all connections.
I have asked the upstream provider to look into this and fix it for us so that it does not happen again. We will see what the outcome is, and I will keep you updated.
As many of you know we were experiencing some network issues starting at around 3:33:45 last night. These issues started with some minor denial of service attacks from IP's based out of England or the UK, France, USA, and Japan.
Our upstream provider decided to go ahead and implement a filter to block this attack, which was very affective. The attack which was two pronged, a standard "fill up your pipe DDoS" and a UDP flood, peaked at about 85 Megs of traffic and about 15k packets per second.
The game server itself runs about 15 - 20k packets per second anyway so the packets weren't a big deal, what happened was when the filter was implemented a monitoring system kicked in and kept launching TCP connections to the server on all available ports, however the monitor for port 2593 went crazy, and basically inadvertently became a denial of service attack on the game server, thus blocking all connections.
I have asked the upstream provider to look into this and fix it for us so that it does not happen again. We will see what the outcome is, and I will keep you updated.
