Newsgroups: comp.parallel.pvm
Path: ukc!uknet!pipex!howland.reston.ans.net!spool.mu.edu!caen!crl.dec.com!crl.dec.com!uvo.dec.com!cicely.ilo.dec.com!jim
From: jim@cicely.ilo.dec.com (Jim Lawton)
Subject: Re: Timing Anomaly
Message-ID: <1994Jan18.145537.27529@decuk.uvo.dec.com>
Lines: 15
Sender: usenet@decuk.uvo.dec.com (USENET News System)
Reply-To: jim@ilo.dec.com
Organization: Digital Equipment Corporation
X-Newsreader: mxrn 6.18-9
References: <CJF00y.M78@nas.nasa.gov> <2gsdm4$jui@pravda.sdsc.edu> <2gtsl0$o73@belfort.daimi.aau.dk>
Date: Tue, 18 Jan 1994 14:55:37 GMT


In article <2gtsl0$o73@belfort.daimi.aau.dk>, kells@daimi.aau.dk (Kell S|nnichsen) writes:
>
>Btw., the problem disappeared with patch 5 to PVM 3.2 and unless this patch
>patches our TCP implementation, too, how could it then repair this
>syndrome?
>

The operation of this aspect of the TCP congestion control algorithm is
controlled by the TCP_NODELAY socket option. 

In PVM 3.2.5, this socket option is turned on.

/jim


