JUNOS High Availability

MPLS Transition Case Studies

Now that we have looked at vendor-specific configuration characteristics of MPLS and the associated signaling protocols, let’s examine how a network can safely transition from a single-vendor to a multivendor MPLS implementation without risking availability. For MPLS transitions, devices that come into play include Provider (P) and Provider Edge (PE). The Provider devices typically serve as transit points for LSPs established between the Provider Edge devices.

This section offers two MPLS transition case studies. Case study 1 demonstrates one alternative for high availability equipment transitions: standing up a redundant pair, bringing up connectivity, transiting traffic, and bringing down the legacy pair. You confirm and monitor stability for a set time at each phase. Case study 2 provides another alternative: replacing devices within a high availability pair one at a time.

Both case studies in this chapter are based on the network shown in Figure 16-5. This topology features redundant P and PE routers that support a primary and secondary pair of LSPs. These LSPs are being used to engineer traffic bidirectionally between PE-router-1 and PE-router-3 to support control of the path used for traffic between Customer Edge (CE) devices at Acme Site 1 and Site 2. The P and PE routers are all in AS 64512 and are fully meshed using BGP. OSPF is used as the IGP, and RSVP is used to signal LSP setup. At this point, we assume that all devices in the network are ...