Performance based "Priority groups"
Hi,
I am quite new to LTM so probably there is some obvious solution available and I just missed it...
There is Priority Group Activation feature that resolves issues when there are some problem with pool members or nodes. For me it's more HA feature. I wonder if there is some way to configure pool in such way: 1. Let's say we have 10 pool members 2. Under normal load connections are distributed to 5 members 3. When load on all 5 members exceeds some threshold additional pool members are being made active 4. When load lowers additional no new connections are directed to additional members
Is that possible via GUI or maybe with iRule?
The idea is to have for example 5 members that are servers located in my data center and 5 additional configured in the cloud. Cloud members should be used only when servers in DC are overloaded. Kind of Priority Group Activation but not based on servers availability but servers load. Probably some of LB algorithms could came close but as far as I understand still not exactly the same results can be attained.
Piotr