D-Link DFL-2500 - Manual (Page 282)

SLB also means that network administrators can perform maintenance tasks on servers or
applications without disrupting services. Individual servers can be restarted, upgraded, removed, or
replaced, and new servers and applications can be added or moved without affecting the rest of a
server farm, or taking down applications.
The combination of network monitoring and distributed load sharing also provides an extra level of
protection against Denial Of Service (DoS) attacks.
NetDefendOS SLB is implemented through the use of SLB_SAT rules in the IP rule set and these
rules offer administrators a choice of several different algorithms to distribute the load. This allows
the tailoring of SLB to best suit the needs of the network.
There are four issues to be considered when using SLB:
1. The target servers across which the load is to be balanced
2. The load distribution mode
3. The SLB algorithm used
4. The monitoring method
Each of these topics is discussed further in the sections that follow.
10.3.2. Identifying the Servers
The first step is to identify the servers across which the load is to be balanced. This might be a
server farm which is a cluster of servers set up to work as a single "virtual server". The servers that
are to be treated as a single vitual server by SLB must be specified.
10.3.3. The Load Distribution Mode
No single method of distributing the server load is ideal for all services. Different types of services
have different needs. In the IP rule set the administrator can configure rules for specific services.
SLB will then filter the packet flow according to these rules.
NetDefendOS SLB supports the following distribution modes:
Per-state Distribution In this mode, SLB records the state of every connection. The
entire session will then be distributed to the same server. This
guarantees reliable data transmission for that session.
IP Address Stickiness In this mode, all connections from a specific client will be sent
to the same server. This is particularly important for SSL
services such as HTTPS, which require a consistent connection
to the same host.
Network Stickiness This mode is similar to IP stickiness except that by using a
subnet mask, a range of hosts in a subnet can be specified.
10.3.4. The Distribution Algorithm
There are several ways to determine how a load is shared across a server farm. NetDefendOS SLB
supports the following algorithms:
Round Robin The algorithm distributes new incoming connections to a list of servers on
a rotating basis. For the first connection, the algorithm picks a server
randomly, and assigns the connection to it. For subsequent connections, the
10.3.2. Identifying the Servers Chapter 10. Traffic Management
282