From:	SMTP%"CMU-OpenVMS-IP@DRYCAS.CLUB.CC.CMU.EDU"  9-SEP-1994 09:46:06.89
To:	EVERHART
CC:	
Subj:	RE: problems with 665a on vms 5.3

Date: Fri, 09 Sep 1994 14:02:38 GMT
From: "Andy Harper, KCL Systems Manager" <udaa055@bay.cc.kcl.ac.uk>
Subject: RE: problems with 665a on vms 5.3
To: cmu-openvms-ip@DRYCAS.CLUB.CC.CMU.EDU
Cc: udaa055@bay.cc.kcl.ac.uk
Errors-to: CMU-OpenVMS-IP-Request@DRYCAS.CLUB.CC.CMU.EDU
Reply-to: CMU-OpenVMS-IP@DRYCAS.CLUB.CC.CMU.EDU
Message-id: <00984373.EA1BA500.32@bay.cc.kcl.ac.uk>

>In article <34kmeo$43m@newark-news.rutgers.edu>, laidlaw@draco.rutgers.edu
>writes:
>>	Could anyone send me pointers on what to look at for diagnosing the
>>problem?  There does not seem to be any real difference that in that
>>netstat
>>stuff between the working connections and the stuck ones but I am probably
>>overlooking something.
>
>It took me ages to fix this one.  Turned out the SLIP line I was
>connected to had a BTU of 552, which meant that if my machine running
>OpenCMU/IP sent packets that were too big, they would not get through.
>
>There is a parameter that you can set in the file INTERNET$CONFIG (or was
that INET$CONFIG?) that prevents this sort of problem from occuring.
>Sorry, but I cannot remember what the parameter is called; it's at home
>and I'm at work.


Try:

       Variable:MAX_TCP_DATASIZE:xxxx	! the MTU value
       Variable:DEFAULT_MSS:xxxx	! normally MTU-40

Andy Harper
Kings College London