1. Introduction
Pseudowire signaling requires that the Pseudowire Type (PW Type) be identical in both directions. For certain applications the configuration of the PW Type is most easily accomplished by configuring this information at just one PW endpoint. In any form of LDP-based signaling, each PW endpoint must initiate the creation of a unidirectional LSP. By the procedures of [CONTROL], both Label Mapping messages must carry the PW type, and the two unidirectional mapping messages must be in agreement. Thus within the current procedures, the PW endpoint that lacks configuration must wait to receive a Label Mapping message in order to learn the PW Type, prior to signaling its unidirectional LSP. For certain applications this can become particularly onerous. For example, suppose that an ingress Provider Edge (PE) is serving as part of a gateway function between a layer 2 network and layer 2 attachment circuits on remote PEs. Suppose further that the initial setup needs to be initiated from the layer 2 network, but the layer 2 signaling does not contain sufficient information to determine the PW Type. However, this information is known at the PE supporting the targeted attachment circuit. In this situation, it is often desirable to allow the initiation of the two LSPs that compose a pseudowire to remain independent. A means is needed for allowing a PW endpoint (lacking a priori knowledge of the PW Type) to initiate the creation of an LSP. This document defines a wildcard PW Type to satisfy this need.
1.1. Conventions and Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [KEYWORDS]. This document introduces no new terminology. However, it assumes that the reader is familiar with the terminology contained in [CONTROL] and RFC 3985, "Pseudo Wire Emulation Edge-to-Edge (PWE3) Architecture" [ARCH].
