JUNOS High Availability

Chapter 5. Virtualization for High Availability

Picture this: you have just left a meeting with the CTO. During the meeting, you were informed that, by order of the CFO, there is no more money, but your customer support manager says the company must improve customer satisfaction statistics by increasing, and most importantly delivering, higher uptime levels in customer service-level agreements (SLAs). Everyone is looking at you to deliver service improvements with empty pockets.

Regardless of your particular business, you have a single goal: to provide reliable access to information. Whether you are in charge of a purpose-built enterprise data center or collocation facilities selling data access, specific uptime levels are defined in your customers’ SLAs. While 99.9% uptime was an acceptable measure of reliability a decade ago, the increased importance of mission-critical information demands even higher availability. Many of today’s network deployments require an uptime of 99.999%, commonly referred to as “five 9s” of availability. Some networks have even more stringent requirements, approaching an uptime of 99.99999%, or “seven 9s” of availability.

This chapter discusses how the consolidation of resources can improve your network’s high availability. Why consolidate? The bottom line is that having fewer devices in your network results in fewer opportunities for potential errors and hence in fewer failures. Here are a few specific reasons for consolidation:

Ease of management

Reducing ...