BIG-IP VE LTM 200Mbps - Throughput is limited to total or on each virtual servers?

Hello, I'm sorry for the long text and bad english, but is hard to explain this case.

 

I'm facing issue in LTM 200Mbps and I have question about "How BIG-IP enforce the throughput limit rate" I have read this link https://support.f5.com/csp/article/K15831 but it was not conclusive for me.

 

How would the BIG-IP enforces that rate? per all TMM process or per virtual server TMM process?

 

We have been experienced that, with VE LTM 200Mbps limit, just one virtual server was unavailable when limit was reached on box (ltm logs showing up to 260Mpbs) and another virtual servers kept working normaly. The virtual server that was unreachable requests, was under attacked with DoS L7 having 12k hard connections. Meanwhile, another virtual server with few connections and requests, replies to requests even on +260Mbps.

 

In my other question here: ";, I ask about an ASM DoS issue, and the behaviors is alike to this case.

 

So the customer case 1 have ASM 200Mbps and customer case 2 have LTM 200Mbps, in both cases, the attacked virtual server stops to work and all other virtual servers kept working normally.

 

This way, I think that problem is when BIG-IP VE having throughput limit rate, it cause problems only on heavy requested virtual servers.

 

Do you know what I mean? Make it sense to you? Do you know how the througputh limit works indeed?

 

Thank you in advance.