Can someone explain some server jargon?
 

EV1 says their private racks are "connected to our
network through a dedicated 100 Mbps link."

But this is clearly not enough bandwidth if you go and
fill the rack with 20 x 10 Mbps unmetered boxes.

On their private rack info page:

http://www.ev1servers.net/english/privateracks.asp

they say, "If a second 3COM switch is required (when a rack is
used for unmetered servers), a setup fee of $1,000 is required."

So the "switch" is the "100 Mbps link", is that right?

Are there any drawbacks or limitations associated with having 2 x
100 Mbps switches on a rack as opposed to having a single
200 Mbps switch, eg would it affect a load balancing setup?

And what's the difference between a managed and unmanaged switch?

I think so but you better ask the support about it.

Hello, calling all server gurus :glugglug

A managed switch can be managed remotely through an interface, an unmanaged switch can not. But why do you think 100mbit won't be enough for 20x 10mbit? You could definately fit 15 on it, with some overselling and it won't slow down. Unless it's for your own project and you expect them to be maxed out 24/7.

Thanks for the explanation :)

Quote:

But why do you think 100mbit won't be enough for 20x 10mbit? You could definately fit 15 on it, with some overselling and it won't slow down. Unless it's for your own project and you expect them to be maxed out 24/7.

It's for my own project. I am looking at the possibility of using multiple load-balanced 10Mbit boxes in a private rack with the bandwith allowances aggragated into a single allowance of 200Mbit/mo (assuming 20 boxes). There's no way that I'lll go even close to using 200Mbit initially. I just want to make sure that I am not going to end-up with a scenario where as things expand, I am not able to make maximum use of the available bandwidth and rack space.