Summary

  • The IETF Datatracker lists draft-ietf-spring-stamp-srpm-mpls-05 as a SPRING Working Group Internet-Draft updated on 16 August 2026.
  • The current text describes STAMP procedures for SR-MPLS policy, IGP best-path, Flexible Algorithm, and L2/L3 service paths, including one-way and loopback forms of measurement.

The important question in the new version is not whether a network can send a test packet. It is whether the measurement packet experiences the same engineered forwarding treatment as the service being judged. The Datatracker record identifies version -05 as an active SPRING Working Group draft dated 16 August; the current text describes its procedures in detail.

STAMP already supplies a two-way measurement framework. The draft applies it to SR-MPLS paths: policy segment lists, IGP best paths, Flexible Algorithm paths, and Layer-2 or Layer-3 services. It sets out two-way, one-way, loopback and loopback-with-Timestamp-and-Forward modes. In each case, the intended discipline is to make the test traffic relevant to the forwarding path rather than merely convenient for a controller to originate.

That distinction matters especially around ECMP and explicit policies. A delay or loss figure is useful only if an operator can say what path and encapsulation it sampled. The draft describes carrying test traffic with SR-MPLS headers matching the traffic under examination; it also treats matching forward and return paths as a practical issue. It does not establish that any particular network already does this, nor does it report a resulting SLA improvement.

There is also a capacity boundary. The document notes that a Session-Reflector can require slow-path processing, constraining the frequency and number of sessions. Its additional modes are presented as mechanisms for SR-MPLS measurement, not as proof that reflector pressure has been solved in deployed equipment. The draft is intended as Informational work, remains in I-D Exists state, and expires on 17 February 2027 unless it advances.