Juniper MX Series, 2nd Edition

Chapter 6. MX Virtual Chassis

As the number of devices in your network grows, the operational burden increases. A common term in the networking industry is “stacking,” which is a concept where multiple devices can be joined together and managed as a single device. A common problem with stacking is the lack of intelligence in the implementation, which leads to high availability problems. All too often, vendors implement a stacking solution that simply designates master and slave devices to provide bare minimum functionality. In simple stacking implementations, the failure of a networking device requires a lengthy mastership election process and no synchronization of kernel state such as routing protocol adjacencies, routing tables, and MAC address tables.

Virtual chassis really shows off Juniper’s engineering prowess. When looking at a standalone chassis, there are many things that are a given: dual Routing Engines, nonstop routing, nonstop bridging, and graceful Routing Engine switchover. Virtual chassis was designed from the ground up to include these critical features and provide a true, single virtual chassis.

Why should you accept anything less? To simply refer to virtual chassis as “stacking” is an insult.

What Is Virtual Chassis?

Virtual chassis is very similar to a distributing computing concept called a single system image (SSI), which is a cluster of devices that appears to be a single device. However, simply appearing to be a single device isn’t good enough; all of ...