epic.Technology
Servers
Finally we have the servers. epic.LAN has 4 servers, all of them Dell, all of them have multiple CPUs and multiple cores. Currently 2 are used for game servers and 2 are used for network servers. At epic.One, the core network server developed a fault and refused to work. We moved all it’s functionality onto one of the gameservers and the LAN carried on. Downtime from this for the core network was approximately 30 minutes and the Intranet was down for almost 2 hours whilst the database and websites were being configured and restored.
The plan for epic.TWO will be to virtualise the network servers onto one physical PC. The servers were monitored all event by Cacti and viewing the CPU/Memory usage of the servers, it was obvious the network servers were very under-utilised. For epic.TWO we will be virtualising 2 network servers onto one physical machine. The game servers will continue to be normal servers. This then gives us 3 high spec game servers and one server for the core network. Thanks to the server being virtualised, we can take a backup of virtual machines at regular intervals and set them up again on one of the other servers with minimal downtime.
Summary
This post is longer than I expected! I hope this gives a good insight into the kind of technical infrastructure and engineering that we use at epic.LAN, as well as an idea of how we’re going to improve it for next time. If you have any ideas or suggestions, please use the epic.LAN forums or post here!