Statically Configured Pseudo-Wires

Static MPLS PWs are configurable point-to-point emulated circuits that have statically configured MPLS PW labels. Static PWs do not use targeted Label Distribution Protocol (LDP) to negotiate setup and exchange peer status. They can use any type of MPLS tunnel Label Switch Path (LSP). When used in conjunction with static routes and static LSPs, no routing protocol (such as OSPF or ISIS), and no label distribution protocol (such as LDP or RSVP-TE) are needed to provision and manage static PWs. Managing this kind of network can provide a disruptive architectural solution for building large backhaul networks that are easy to provision, operate, and incrementally expand. Because protocols are no longer required to set up emulated circuits over MPLS, you now have the capability to proactively, or on-demand, verify end-to-end PW connectivity, to provide remote endpoint status, and offer options to configure redundant PWs to maintain network high availability.

Statically configured PWs provide greater administrative and management control over the network. It also allows MPLS PWs to be configured across a network when no label distribution protocol is running. This can simplify operational management and reduce equipment interoperability issues that can arise when deploying routing packet networks.

You can use the configure mpls label max-static command to configure the max number of static labels (labels reserved for static configuration). The maximum number of static labels depends on the underlying hardware platform, and at least 100 labels are reserved for dynamic or signaled labels, such as those used by LDP and RSVP-TE. Use the show mpls label usage command to display the current label ranges and usage.

Ethernet PWs for L2VPN VPLS/VPWS can be statically configured. The PWs in a VPLS can be a mix of both signaled and statically configured types, but the corresponding peer PW must be configured with the same type.

Static PWs are created by adding a peer with configured labels. If the configured labels are not in the allowable range, or are already in use by some other statically configured entity such as static LSPs, then the command is rejected. Once a static PW is created, the labels for that PW can be changed without deleting and re-adding the peer. The L2VPN can remain operational during the change; however, the PW will go down and come back up. If the configured PW labels are accepted, but have not yet replaced the “in-use” labels in hardware, the show l2vpn detail commands outputs an additional line showing the “pending” rx and tx labels. This line will only be shown if necessary, and generally, would not be shown since this is only a transient condition with a small window for its occurrence.

Since static PWs are not signaled, the remote parameters, such as remote Virtual Circuit (VC) status and remote I/F MTU, are not "none" for Ethernet PWs. The local VC status is still calculated and displayed, but is not sent to the peer. Additionally, since the “standby” VC status bit cannot be signaled, PW redundancy cannot be configured for L2VPNs that have static PWs, and Hierarchical Virtual Private LAN Services (H-VPLS) is not supported. There is no OAM support for static PWs, and VCCV is not supported.

Use the show l2vpn command to display PWs that are statically configured. Since static PWs are not signaled, a static PW in a state similar to an LDP PW in signal state, will display in a down state.