Path: utzoo!attcan!uunet!lll-winken!ncis.llnl.gov!ncis!helios.ee.lbl.gov!pasteur!ucbvax!BBN.COM!jchao
From: jchao@BBN.COM
Newsgroups: comp.protocols.tcp-ip
Subject: Re: Bfly redirect problem
Message-ID: <8901181822.AA19049@ucbvax.Berkeley.EDU>
Date: 18 Jan 89 16:54:36 GMT
References: <890117150322.558487@RADC-MULTICS.ARPA>
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 25

Tim and Jim 

Thank you for reporting and forwarding the redirects log. 
We are trying to understand the problem and hope to resolve it 
soon.  To help us track down the problem we would appreciate 
it very much if you could in the future include in the log the 
source and destination IP addresses of the original packets 
(if they are readily available).  It would also be extra nice 
if a short description or a simple diagram be included in the
msg showing how your host/gw is connected to Arpanet/Milnet/Internet.  

We suspect a possible configuration problem on the reston-dcec-mb.ddn.mil
gateway (10.6.0.20, 26.21.0.104) and we have temporarily turned 
loadsharing off for that gatewayw.  This should solve the looping redirects 
between reston-dcec-mb.ddn.mil and mclean-mb.ddn.mil as reported by Tim 
for now. We are studying John's report.  It may be a transition problem 
related to LSI11 GGP extrap hop and interaction between
ICMP net redirects from Butterfly Mailbridges (loadsharing) 
and ICMP net redirects from LSI11 EGP servers.

Please cc future msgs to mb-transition@BBN.COM.  Thank you for your 
assistence and cooperation.

John Chao
BBNCC Gateway Development