Date: Sun, 27 Nov 2005 21:24:15 -0800 From: Mike Eubanks <mse_software@charter.net> To: freebsd-stable@freebsd.org Subject: Re: NFS network load on 5.4-STABLE Message-ID: <1133155455.868.135.camel@yak.mseubanks.net> In-Reply-To: <20051127204344.GB3175@xor.obsecurity.org> References: <1132964757.831.20.camel@yak.mseubanks.net> <43891EA5.2020206@mac.com> <1133083658.838.109.camel@yak.mseubanks.net> <20051127204344.GB3175@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 2005-11-27 at 15:43 -0500, Kris Kennaway wrote: > On Sun, Nov 27, 2005 at 01:27:38AM -0800, Mike Eubanks wrote: > > On Sat, 2005-11-26 at 21:49 -0500, Chuck Swiger wrote: > > > Mike Eubanks wrote: > > > > As soon as I mount my NFS file systems, the network load > > > > increases to a constant 80%-90% of network bandwidth, even > > > > when the file systems are not in use. NFS stats on the client > > > > machine (nfsstat -c) produce the following: > > > > > > > [ ... ] > > > > Fsstat and Requests are increasing very rapidly. Both the client and > > > > server are i386 5.4-STABLE machines. Is this behaviour normal? > > > > > > Sort of. Some fancy parts of X like file-manager/exporer > > > applications tend to call fstat() a lot, but it's probably > > > tunable, and if you enable NFS attribute caching that will > > > help a lot. > > > > Thank you for the reply Chuck. It seems that it is something to do > > with Gnome. I haven't done an upgrade to 2.12 yet, but the difference > > did happen when I refreshed my user configuration to remove any stale > > config files. Using the "top -mio" command I get the following: > > > > [ ... ] > > > > That doesn't look like it is showing a problem to me. In particular > it is indicating 0 I/O. > > > +---- file-manager/explorer? > > | > > client.220312819 > server.nfs: 96 fsstat [|nfs] > > server.nfs > client.220312819: reply ok 168 fsstat POST: DIR 755 ids > > 1001/0 [|nfs] > > > > If this is enough evidence for the file-manager/explore, > > It's evidence that something is peforming NFS I/O, but it doesn't show > what. Perhaps you needed to also use the top -S flag, or to sort the > output by typing 'ototal'. > > > I'll just have > > to accept it for now. I can't find anything about tuning them. As far > > as attribute caching, do you mean the `-o ac*' options to mount_nfs? I > > also noticed two sysctl values, although, I left them unmodified. > > > > vfs.nfs.access_cache_timeout: 2 > > vfs.nfs4.access_cache_timeout: 60 > > Increase the former (you're not using nfs4). Try 60 seconds, for > example. The downside is that you'll have to wait up to a minute for > access changes on the server to be visible to the client, but that's > usually not a big deal unless you're accessing a lot of dynamically > created and destroyed files. > I made the sysctl modification. Still no luck though. The only process that had any activity using the top with the -S option, or after sorting by total, was the swapper/syncer. Even then, it was hardly active. The network traffic persists. -- Mike Eubanks <mse_software@charter.net>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1133155455.868.135.camel>