Date: Mon, 16 May 2005 07:57:14 -0400 From: Mike Tancsa <mike@sentex.net> To: Andrea Venturoli <ml.diespammer@netfence.it> Cc: freebsd-questions@freebsd.org Subject: Re: NFS read-ahead? Message-ID: <9f2h819gnc2938un66cpouvnrbti39adsv@4ax.com> In-Reply-To: <428873A1.1060209@netfence.it> References: <4286313A.3080102@netfence.it> <eu9d81doft5cn61br59aimchmgeg471e8g@4ax.com> <428873A1.1060209@netfence.it>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 16 May 2005 12:19:13 +0200, in sentex.lists.freebsd.questions you wrote: >Now, the manual pages says for -a I should give values greater than 1 if= =20 >bandwidth*delay is high. I'm on a swithed FastEthernet, so=20 >bandwidth=3D100Mb/s and delay is around 0.2 ms of ping roundtrip time. >As for -r it says it should be used when there are many "fragments=20 >dropped due to timeout"; yet I see almost none. > >Do you have any suggested values to start with, before I make random=20 >experiments? Hi, Offhand, not really, but if you look around via google for mount_nfs freebsd "-r" a lot of people suggest r=3D32768,-w=3D32768. Perhaps use that as a starting point and experiment to see what gives you the best performance for your situation. ---Mike -------------------------------------------------------- Mike Tancsa, Sentex communications http://www.sentex.net Providing Internet Access since 1994 mike@sentex.net, (http://www.tancsa.com)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9f2h819gnc2938un66cpouvnrbti39adsv>