RFC 25 UNKNOWN ARPANET · CONSTRAINTS · LINKS

No High Link Numbers

A proposal to keep link numbers at or below 63.

RFC EDITOR

Crocker, S.D., "No High Link Numbers", RFC 25, DOI 10.17487/RFC0025, October 1969, <https://www.rfc-editor.org/info/rfc25>.

BIBTEX

@misc{rfc25,
  series = {Request for Comments},
  number = 25,
  howpublished = {RFC 25},
  publisher = {RFC Editor},
  doi = {10.17487/RFC0025},
  url = {https://www.rfc-editor.org/info/rfc25},
  author = {S.D. Crocker},
  title = {{No High Link Numbers}},
  pagetotal = 1,
  year = 1969,
  month = oct,
}

PUBLISHED: October 1969
AUTHORS: S.D. Crocker
STREAM: Legacy
SUGGESTED MAXIMUM LINK NUMBER: 63
MORE REALISTIC NUMBER: 32
NUMBER THE AUTHOR SAYS IS SUFFICIENT: 64
DOI: 10.17487/RFC0025

RFC text

Original text · RFC Editor View on rfc-editor.org ↗

Network Working Group                                      Steve Crocker
Request for Comments: 25                                            UCLA
                                                         30 October 1969




                           NO HIGH LINK NUMBERS




Because it may be desirable to reserve one or more link numbers for
instrumentation purposes, and because 256 link numbers are many more
than are needed, we suggest that no link number over 63 be used.  At
UCLA, we will implement our tables to take advantage of this limitation.
We also note that 32 may be even more realistic, but 64 is certainly
sufficient.

ISO Plain Language

On this page

PROBLEMThe protocol had 256 link numbers, more than the author thought were needed, and higher numbers might be useful for instrumentation.

MECHANISMThe note suggests that no link number over 63 be used, while noting that 32 may be more realistic and 64 is sufficient.

DESIGN SUGGESTIONUCLA says it will build its tables to take advantage of the proposed limit; the note does not report wider adoption.

§1Why Use Fewer Link Numbers

The note says 256 link numbers are many more than needed. It also says it may be desirable to reserve one or more numbers for instrumentation.

On these grounds, the author suggests that no link number over 63 be used.

The author thinks the protocol allows more link numbers than are needed. Keeping the highest numbers unused could leave some available for instrumentation.

§1Proposed Limit and UCLA's Plan

The proposed rule is that no link number over 63 be used. The note says UCLA will implement its tables to take advantage of this limitation.

It adds that 32 may be a more realistic number, while 64 is certainly sufficient.

The note recommends staying at or below link number 63. UCLA says it will design its tables for that limit, though the author thinks 32 may often be enough.

§1Scope of the Note

RFC 25 gives a short recommendation and says how UCLA plans to implement its tables. It does not specify a formal allocation procedure or report that every site adopted the proposed limit.

This is a brief proposal with an implementation intention from UCLA. The note does not describe a network-wide adoption decision.