
So as you can see the sizing is based upon events per second. So how excactly can you measure you events per second? Let me clarify this by showing you the devices we monitor and all together the events per second we generate. Here' our monitored devices:
4 x Windows Server
121 x Cisco 2811 IOS Routers
384 x Cisco 3750 IOS Switches
117 x PIX 506E Firewalls
2 x PIX 520 Firewalls
2 x PIX 515E Firewalls
7 x Cisco 3825/3845 Routers
2 x Cisco 4215 IPS 5.1 Sensors
6 x Unix Servers with Snort
So all together we have a relatively large infrastructure monitored by MARS. So events per second comes to what excactly? At peak usage during the day we generate about only 57 events per second. Which CS-MARS version do we run? We're currently running the 100e, which is capable of 3000 events per second! Wow... that's pretty damn powerful! We bought this size knowing that our infrastructure will soon include more servers along with NAC reporting to MARS. Hopefully this sizing overview helps when making the decision to purchase a CS-MARS appliance. If you ever need any help or recommendations, just ask!