We are using CSS11500 for load balaning between in our two servers. We are using default round robin algorithm for connection hits. We need to modify our load balancing as per the server load. Can we check the server load "as memory or CPU usage" and then accordingly forward the request to the less loaded server.
Can we check which server has more active connections and
then forward the requests to the server on which less number of active connections exist. ( we are monitering only one service a single port ).
We are trying to use weight command to mannually redirect hits to one server as compared to second server.
Minimum configuration is used to define service this service moniter on 9501 port on each server. ( with weight 2 on one and weight 9 on another )
These are added to one content in owner configuration. Can we say balance as per the weight in each service ? Or do where do we need to adjust the weights ?
But , We are looking for the server load balancing as per the load on each server.
any link on cisco.com or any working example is highly appreciated.
Thanks in advance
You can try looking into DFP setup for CSS. DFP agents running on servers can tell CSS where to send more requests depending upon resources available on servers.
Alternatively, weights can be assigned to services and content-rule should use weighted-round-robin algorithm.