I have used NFS very little since I started using Linux. Is there any chance you could use SSHFS instead? I have found this to be a very portable option that works any time SSH does. And if SSH doesn't work, it's usually pretty easy to determine why.
That said, if you have a spare switch/hub type device, I would swap it in for each different switch/hub in the chain temporarily to eliminate those as possible problems. Based on your description, this means you would only need to eliminate the netgear 8-port as the issue. Have you looked into debugging options for NFS? Like a --verbose flag? Maybe do some individual testing with SCP to see if there is a wider network issue, or if it's something specific to NFS? -wes On Sat, May 22, 2010 at 9:25 PM, Someone <[email protected]> wrote: > I have my server running at 100 mb/s connected to a Netgear 8 port > 10/100 switch where one of the ports is linked to a 10 MB hub and > from there the client machine with a RealTek 8029 10 MB nic card is > connected. > > Using a gpxe cdrom, I get a PXE boot going and everything seems fine > till it is time to mount the root filesystem. I get errors that the > NFS server is not responding interspersed with OK messages. > > Possibilities: > > 1) The speed change from 100 mbps to 10 mbps is the problem. > > 2) The nic is bad, though it works in Windows okay. > > 3) The sync mount option is causing a problem. > > /nfsroot/lfs/lfsp3/tuna \ > tuna.robinson-west.com(rw,no_root_squash,sync) > > I've been trying to following the NFS site for troubleshooting, > but I'm not getting enough information on exactly what is > causing the server to be inaccessible. > > I have tried bypassing the hub, didn't make any difference. > > _______________________________________________ > PLUG mailing list > [email protected] > http://lists.pdxlinux.org/mailman/listinfo/plug > _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
