Date: Mon, 26 Jun 1995 06:56:52 +1000 From: Bruce Evans <bde@zeta.org.au> To: bde@zeta.org.au, phk@freefall.cdrom.com Cc: bde@freefall.cdrom.com, current@freebsd.org, pete@puffin.pelican.com Subject: Re: cvs commit: src/sys/i386/isa sio.c Message-ID: <199506252056.GAA06791@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
[for slip (and ppp)] >> time rcp machine1:/kernel machine2:/tmp/kernel & \ >> time rcp machine2:/kernel machine1:/tmp/kernel & >> >> This takes about twice as long as would separate rcp's because acks get >> queued behind large amounts of data and arrive too late to keep the data >> streaming. >Have you tried making your tcp-windows bigger ? I tried a smaller tcp_sendspace/tcp_recvspace but that made the problem worse. The slowdown is (now) not nearly a factor of 2. It is from 95 seconds for rcp'ing in one direction only to 125 and 125-136 seconds (the slowdown is sometimes asymmetrical). With a tcp_sendspace/tcp_recspace of 32K instead of 16K, the slowdown is less (to 110 and 128 seconds). With large windows, transmission works OK for a while, then both sides sometimes stop sending for a second or two, then each side takes turns transmitting, then transmission works OK... With small windows, this cycle repeats more often. The problem is more obvious at lower line speeds. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199506252056.GAA06791>