I'm about to begin a production deployment of shavlik protect for our server environment. Our environment consists of ±800 windows servers distributed across three data centres. The three DCs have excellent connections between them so bandwidth and latency are not an issue, for this reason my intention is not to use distribution servers.
I am however, curious about the use of multiple console servers. Currently I've got one console server deployed with 4 cores and 8GB of memory. My intention is to patch servers without agents. We have a resonabally short maintence window of 2 hours, once per week for most of our environments.
Will one server provide sufficent performance for an environment of this size, or should we look at a distributed console model?