From owner-freebsd-questions Tue Jan 24 20:55:31 1995 Return-Path: questions-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id UAA25609 for questions-outgoing; Tue, 24 Jan 1995 20:55:31 -0800 Received: from simon.chi.il.us (simon.chi.il.us [199.245.227.17]) by freefall.cdrom.com (8.6.9/8.6.6) with SMTP id UAA25603 for ; Tue, 24 Jan 1995 20:55:29 -0800 Received: by simon.chi.il.us (Smail3.1.28.1 #2) id m0rWzkN-000NB2C; Tue, 24 Jan 95 22:54 CST Message-Id: Date: Tue, 24 Jan 95 22:54 CST From: steve@simon.chi.il.us (Steven E. Piette) To: alan@picard.isocor.ie, terry@cs.weber.edu Subject: Re: Slow ftp transfer times on Ethernet Cc: questions@FreeBSD.org, alan@buster.internet-eireann.ie Sender: questions-owner@FreeBSD.org Precedence: bulk Yes, NFS writes are suppost to be synchronous for the reasons Terry mentions. NO! Current Sun systems DO NOT use Asynchronous writes by default. You can use a Prestoserve NVRAM to cache synch disk writes on Sun's but otherwise there still synch in both versions of Solaris. I'd start looking at nfsstat on both systems and taking snoop traces if you have a Solaris 2.X machine to get more info. > From: terry@cs.weber.edu (Terry Lambert) > Subject: Re: Slow ftp transfer times on Ethernet > To: alan@picard.isocor.ie (Alan Byrne) > Date: Tue, 24 Jan 95 10:06:51 MST > Cc: questions@FreeBSD.org, alan@buster.internet-eireann.ie > > > This is because NFS writes are synchronus; this converts into an apparent > request/response before the client is permitted to send the next packet. > > The current Sun and SVR4 NFS servers are configured by default to use > async writes, and are apparently faster because of this, at a trade-off > in reliability (crash your server, and your client believes he has > written data that actually did not get to disk). > > Terry Lambert > terry@cs.weber.edu > --- > Any opinions in this posting are my own and not those of my present > or previous employers. >