Using Appfabric cache with just one node and HA on

I am looking to use MS-Velocity a.k.a Appfabric Cache in a project that is starting by next month (so I am also assuming Q3 will have a go-live licens开发者_如何学Pythone). The project is a web application which will have at least 40K users in 6 months down. And we know for sure that cache server can be of only one node to start with and we will scale horizontally once the user base increases.

After reading this article, I was wondering that if I have to use Velocity with HA on, can I run one node just to start with and increase the nodes in cluster as soon as the load increases?

---

I think the next release will be a Release Candidate and will come with a Go Live license, so this is probably a sensible assumption ;-)

Creating a cache with HA enabled from the start strikes me as quite a neat idea - while you've only got one server it'll behave as a normal cache, but when you scale out the HA bits will kick in instead of you having to re-create the cache.

Bear in mind that you need Windows Server Enterprise Edition for HA.