Status of RFC1191 (Path MTU Discovery)

[Jeffrey Mogul <[e-mail]>]

RFC1191 was issued in November, 1990, and has been a "Proposed Standard"
since then.  Proposed Standards are supposed to advance to Draft Standard
status within a reasonable period.  Due to the lethargy of the chair of the
working group (i.e., me), this step has not yet been made.

RFC1360 says:
   It is general IAB practice that no proposed standard can be promoted
   to draft standard without at least two independent implementations
   (and the recommendation of the IESG). Promotion from draft standard
   to standard generally requires operational experience and
   demonstrated interoperability of two or more implementations (and the
   recommendation of the IESG).

In spite of my sloth, there are in fact many implementations of
RFC1191, some of which are shipping in product systems.  It is
therefore in a position to meet the requirements for promotion not only
to Draft Standard, but perhaps even to Standard (except for the
mandatory time delays).

Therefore, I need to gather formal reports regarding the interoperability
of RFC1191 implementations, both with each other and also with hosts
or routers that do not implement the protocol.

As a first step, with the help of the MTU Discovery Working Group I've
compiled a list of extant RFC1191 implementations.  This is NOT a list
of RFC1191 product support (although some of the entries are products),
and is only intended to help arrange interoperability testing.  If you
know of an implementation not on my list, please let me know about it.

This is the only message that I intend to send to the entire IETF on
this topic; future discussions will take place on the working group's
mailing list (which is "[e-mail]".  If you want to be added
to the mailing list, please send a request to "[e-mail]"
(NOT to the whole group!).

People who have had operational experience (either good or bad) with
Path MTU Discovery are invited to share that with the working group.
I've had at least one report of a link MTU that did not closely match
the values in the table given in RFC1191; this is important information.

-Jeff

Status of RFC1191 "Path MTU Discovery" Implementations

17 February 1993

These are the implementations, as far as I know, that support either or
both of the end-host or router components of RFC1191.  [Briefly:
RFC1191 specifies new behavior for both end-hosts and routers, but only
the router behavior differs from existing protocol specification
standards].

Not all of these implementations are products.  Some are prototypes,
some are potentially future products.  The purpose of this list is
solely for encouraging the interoperability testing required before
Path MTU Discovery can advance to Draft Standard.

I've specifically omitted the product version numbers from this list,
to avoid raising anyone's hopes.   If you want to do an
interoperability test with a product from Vendor X, get in touch with
the contact person from Vendor X and arrange it.

Implementation	Components	Contact person
==============	===========	=========================
3Com Netbuilder	Router		[e-mail] (Rajeev Kochhar)
 
4.4BSD		Router		[e-mail] (Andrew Cherenson)

ACC		Router		[e-mail] (Art Berggreen)

cisco		Router		[e-mail] (Bob Albrightson)

DEC WANrouters	Router		[e-mail] (Elliot Gerberg)

DEC NIS routers	Router		[e-mail] (Dave Forster)

FTP Software	Host		[e-mail] (James B. Van Bokkelen)

Mentat		Router+Host	[e-mail] (Jerry Toporek)

Proteon		Router		[e-mail] (John Moy)

SGI		Router+Host	[e-mail] (Thomas Skibo)

Sun		Router+Host	[e-mail] (Bob Gilligan)

TGV		Host		[e-mail] (Michael Reilly)