White Paper Consolidation of a
VM1 |
|
| VMn | Svc | Cons | |||
| vSwitch 1 |
|
|
| vSwitch 0 | |||
|
| |||||||
|
|
|
|
|
|
| ||
10 GbE VMDq |
|
|
|
|
| 1 GbE | ||
|
| Force 10 | S50N | |||||
|
| ESL Network |
|
|
|
Firewall
Internet
Figure 13.. Live Internet test setup..
Counter- | Counter- |
Strike 1.6* | Strike 1.6* |
Windows | Windows |
Server 2003* | Server 2003* |
32 bit | 32 bit |
VM1 | VMn |
| VMware ESX 3.5 U1* |
Intel® Xeon® Processor 7400
Figure 14.. Test software stack..
ESL game testing
At ESL labs, we ran a series of tests cases that included private and public Internet testing. All these tests revealed that there is no impact on
As part of this testing, we used an Intel Xeon processor 7400 server
2 GB memory, Windows 2003 server (32 bit) and Counter-Strike
1.6(three game servers running per VM). We are not using any kind of CPU affinity or memory reservations. Also, on the ESX server we enabled the “NetQueue” feature and by default a maximum of 16 queues per 10 GbE port are created.
As a part of load generation we used real players who connected to the game servers on the local LAN and through the Internet, and “bots,” which are an emulation of a player playing a game. These bots run on the game server as a
10