Server Load Balancing

Server Load Balancing
The Server Load Balancing (SLB) protocol can be considered an extension to HSRP, which
Cisco recommends should be already configured on the switches performing Server Load
Balancing. The purpose of SLB is to share the load normally associated with multiple traffic
streams terminating on a single server across several servers.
A virtual server represents a cluster of real servers. Clients connect to the virtual
address and—according to a load-balancing algorithm—to a selected real server.
Obviously, clients and servers need to be on separate LANs or VLANs for SLB to
work, because packets have to traverse the SLB switch.
Two different methods of load sharing may be used: weighted round-robin (WRR) and
weighted least connections (WLC). WRR specifies the next server to be connected to using a circular
selection, modified by a weight that allows more clients to connect to particular servers
prior to stepping to the next one. WLC connects to servers based on the number of existing
active connections, weighting this with the server capacity, which can be specified.
It is also possible to use SLB to load-share between firewalls, in which case the
real group of devices is called a firewalls farm.