JUNOS High Availability

Multivendor Reality Check

Before launching into this chapter, we need to pause for a brief reality check. Realistically speaking, network administrators concerned with high availability are not eager to replace a device that is functioning acceptably unless there’s a good reason to do so. While achieving a multivendor state may in some cases be viewed as a good reason, it seldom stands as a sole valid justification for a massive undertaking like replacing devices in a network. That being said, here are some valid reasons—reasons that may justify the labor involved in replacing a device in a high availability network:

Improved performance and stability

Demands for throughput increase over time, and at a certain point the deployed platforms are unable to keep up. Platforms tend to become less stable as they age. Both situations are unacceptable for a high availability network.

Legacy platform replacement

Platforms that are no longer supported by the manufacturer or are nearing end of life (EoL) are not appropriate for a production network with high availability requirements.

Change in functional requirements

As time passes, new services emerge and can result in functional requirements that equipment from the incumbent vendor is unable to provide. Though this might not impact high availability for existing customers, the startup of a new service has revenue implications that can’t be ignored.

Better network manageability and reduction in support costs

These two go hand in hand. As a rule, ...