in any SNA design, and there are multiple ways of accomplishing the same goal. Some methods
are better than others. In this chapter, we put together the last two chapters and apply them to
common SNA design models. We will talk about how to design reliable SNA networks with
redundancy. We will also discuss DLSw+ designs and how to apply QoS features to DLSw+.
By the end of this chapter, you should have a solid understanding of some of the topics on
the exam.
networks with Cisco routers and transporting LLC2 and SDLC frames, it is important that the
convergence time is less than the session expiration time. Otherwise, timeouts might occur. As
long as this is done, you will have the predictability and reliability needed to maintain the SNA
session. Let's discuss some of the methods commonly used in SNA to maintain that reliability.
(see Figure 14-2). The key to this design is that the same locally administered MAC address
appears on both FEPs. In this design, explorers from hosts 1 through 4 find multiple paths to
the FEPs by configuring the routers as DSPUs or gateways. Half of the hosts have MAC address
A, and the other half have MAC address B. Half of the PCs should connect to FEP right, and
the other half should connect to FEP left. This method provides not only load balancing, but
redundancy as well.
other words, both rings cannot be Token Ring 10--they are 10 and 20.