GTM VS Capacity mode

Hi, I've got some problems in finding information about VS Capacity load balancing mode on GTM.

Configuration guide for GTM provides following information:

a) General defenition of VS Capacity mode - it says that VS Capacity mode creates list of virtual servers wighted by capacity, then picks one of them, VS with the greates capacity are picked more often, then all servers are returned, etc.

b) In section "Working with the big3d Agent" I can find that The Global Traffic Manager uses the number of current connections to determine the best virtual server when a pool uses the Least Connections or VS Capacity load balancing mode.

Let's assume that GTM works in conjunction with LTM.

So, it looks like GTM measures current VS Capacity in the following way: LTM provides to GTM (via iQuery protocol) information about current number of connections to particular VS and a maximum number of connections to particular VS. From these values GTM then calculates weight of VS.

If everything really goes in the way i described i.e. capacity measured in connections and not in Kbps, for example - it would be great (because it is exactly what I need).

But, I can't find a clear confirmation in GTM related documents that it really works in such a way.

So my first question is: where can I find detailed explanation of VS Capacity mode?

But there is one more question: where and how can I define the maximum number of connection to the VS?

It seems that the maximum number of connections should be defined on LTM. But how it can be done? Does LTM calculates this value automatically? Or I should set it manually for each virtual server? Or it calculates this value from max. connections of each particular non-virtual server?

So lots of qestions... And it would be great if someone will help me.

Thanks in advance.

--

Best regards, Alex