IPAE Meeting Notes 3/4/93

Bob Hinden <Bob.Hinden@eng.sun.com> Fri, 19 March 1993 01:27 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa18952; 18 Mar 93 20:27 EST
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa18948; 18 Mar 93 20:27 EST
Received: from Sun.COM by CNRI.Reston.VA.US id aa03747; 18 Mar 93 20:27 EST
Received: from Eng.Sun.COM (zigzag-bb.Corp.Sun.COM) by Sun.COM (4.1/SMI-4.1) id AA23696; Thu, 18 Mar 93 17:24:16 PST
Received: from sunroof.Eng.Sun.COM by Eng.Sun.COM (4.1/SMI-4.1) id AA27201; Thu, 18 Mar 93 17:24:17 PST
Received: from bigriver.Eng.Sun.COM by sunroof.Eng.Sun.COM (4.1/SMI-4.1) id AA07666; Thu, 18 Mar 93 17:23:51 PST
Received: by bigriver.Eng.Sun.COM (4.1/SMI-4.1) id AA00731; Thu, 18 Mar 93 17:23:56 PST
Date: Thu, 18 Mar 93 17:23:56 PST
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Bob Hinden <Bob.Hinden@eng.sun.com>
Message-Id: <9303190123.AA00731@bigriver.Eng.Sun.COM>
To: ip-encaps@sunroof.eng.sun.com
Cc: Bob.Hinden@eng.sun.com
Subject: IPAE Meeting Notes 3/4/93
content-length: 4663

IPAE Meeting	3/4/93 
Silicon Graphics

MCI Phone Bridge:	Mike Conn, Yakov Rekhter, Christina Huitema,
                        Carl Beame

On-site:		Bob Hinden, Dave Crocker, Erik Nordmark, Steve Deering,
                        Chuck Desostoa, Ron Jacoby

1) Review Implementations Status
--------------------------------

Christian reported that Werner Vogels / IENSC is doing a Mach
implementation.

CH is able to spend about 20% of his time on SIP implementation.  Deering
suggested looking at latest Berkeley release on UU.NET sources for latest
BSD implementation that support variable length network addresses (called
NET2).  CH would like have a module that could be added to existing
kernels.  Need to incorporate a SIP route module like MRoute module.

CD reported interest in doing an implementation but does not have the
resources to do it now.  Crocker suggested HP work with INRIA and TGV
because these implementations are based on BSD code.

SGI: Implementation underway.  Expect to have pings working in next few
weeks.

Beame / Sun Test: Telnet & FTP working.  Still doing ICMP testing.  Some
problems with ICMP checksums.  When ICMP done complete transport will be
done.  Then on to applications.  Expect a few more hours work.
Clarified: B&W only doing IPAE now.

Geoff Mulligan / KA9Q: SIP packets flowing from KA9Q.  Need to test with
other implementations.

Gary Malkin is doing extensions to SIP stuff for RIPv2.  Some discussion on
merits of Distance Vector.....  [Note:  Draft of SIP RIP now out as an
Internet Draft]

Bob Gilligan meet with James Davidson of Network General.  Network
General wants to do SIP/IPAE sniffer implementation.  They collected some
SIP/IPAE traffic traces.

2) Planning for IETF Meeting
----------------------------

What should we do for the IETF?  Give implementation status reports?
Discuss ease of implementations etc.

IPAE meeting at IETF will be  Monday at 4:30-6pm.

Carl Beame: Can attend, might have SIP working by then.

Sun to get Solaris 2.x to Beame.

IETF Demo:

Deering asked about multicast and sd/vat.

ACTION: Deering / Nordmark to see what is required to do audio/video demo
        at IETF.

Protocols to Demo:

    Telnet / FTP / Ping / TFTP (?) / Email (?)

Sun: all done now.  

B&W: same

KA9Q: Telnet, Ping,  FTP

BSD/ INRIA: No commitments ; Ping, UDP, (but needs help).  Also has BIND
code.

SGI: Ping (certain), others less certain.  Also, NetVisulizer.

ACTION: Crocker will get slot for demo.  Try for Tuesday afternoon or
        evening.

Group decided to setup Wed. March 24 to do testing for IETF demo will be
held at SGI.  Mixture of on-site testing and internet testing.

ACTION: Gilligan to document demo scenario for testing.

CH: Suggested important to have working implementations on Internet to
test against.

Sun and SGI can do this.

CH: Has been getting inquiries about doing SIP pilot.  CH will facilitate
this in Europe.

SGI will put B&W implementation on Internet.

ACTION: Jacoby to put B&W implementation up on Internet.

Carl Beame will be at Interop.  Will meet Crocker and Hinden.

Need to put together test scenario before the 24th.  Deering suggested
that we put short demo doc together.

Discussion about BIND and DNS name services.  Nordmark will look into
adding this to Sun Implementation.  CH expects to have it running soon
tool. Will need to put AE records in to make it useful.

3) IPAE Protocol Issues
-----------------------

Deering: Proposed 4 byte PAD between IP and SIP.  This would help 64bit
machines.  Wants comments from implementors.  Proposed always putting 4
bytes of zero PADs between headers.  Propose is to allow more efficient
processing of SIP packets on 64bit machines.  Only for encapsulating SIP
in IPv4.  CH: Didn't like original proposal because of implementation and
performance costs.  New proposal still requires extra overhead which will
useful in the short to medium term.  Long discussion followed on this
topic.

Group decided to leave it the way it is now (e.g. no padding between IP
and SIP headers).  View was that complexity was not worth the gain.  It
needs to be proven that this is really a win in the IPAE case.  Gain is
not clear given the added complexity to support it.


4) Next Meeting:  
----------------

	24 March	Demonstration of SIP/IPAE Dry Run at SGI


SUMMARY OF OPEN ACTIONS
-----------------------

ACTION: Deering / Nordmark to see what is required to do audio/video demo
        at IETF.

ACTION: Crocker will get slot for demo.  Try for Tuesday afternoon or
        evening.

ACTION: Gilligan to document demo scenario for testing.

ACTION: Jacoby to put B&W implementation up on Internet.