Filter
Top Products
6-2 Px Series Application Switch Installation and Configuration Guide
The Px series application switch supports the following scheduling policies:
Persistence Modes
Persistence, also called stickiness, is the maintenance of a session between a client and a
particular web server. Most medium- and large-scale websites have multiple, redundant
web servers that are run in parallel and contain duplicate information. When a client
Table 6-1: Scheduling Policies
Specifier Policy Description
rr round robin Distributes clients evenly across the web server group by passing each
new connection request to the next server in line. The first client
connection is sent to the first server, the second to the second server,
and so on, until each server has a connection. When each server has
its first connection, the next client connection is sent to the first server,
the next to the second, and so on.
Round robin is the simplest way of balancing web traffic, and is best
suited for environments where the performance level of all servers is
about equal, and all servers provide the same content.
wrr weighted
round robin
Similar to round robin, except that you can apply a weight to each
server. For example, if server A has a weight of two, and server B has
a weight of one, server A receives two connections for each connection
given to server B.
Weighted round robin is useful if all servers provide the same content,
but some are faster than others.
lc least
connections
Passes a new connection to the server having the least number of
active sessions. It distributes clients based on the server with the least
connections currently open.
The least connections policy provides the best performance under most
conditions.
wlc weighted
least
connections
As in weighted round robin, assigns a weight to each server. The weight
can be based on a number of things, including:
■ Processor speed
■ Network connection speed
Connections are distributed using the servers with the least connections
by weight. As in weighted round robin, weighted least connections can
be used if servers have very different connection handling capacities.
Weighted least connections has the advantage of not overloading older,
slower servers with too many connections.