NetBSD-Users archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: nfs error diskless
On Tue, Nov 13, 2018 at 05:59:45PM +0000, Christos Zoulas wrote:
> >I am using diskless NetBSD often (real or VMs) and sometimes I
> >get nfs error 65 for <root>.
> >
> >It seems that NetBSD won't work when NFS server is reachable
> >via router and work when it is reachable directly (in the
> >same network).
> >
> >My current setup is: two PV VMs (actually copied), located in
> >different networks, but connecting to the same NFS server.
> >NFS server is directly reachable to VM1 and via router for VM2.
> >VMs are copies and both have static dhcp leases in ISC DHCPD.
> >Startup argument is "root=xennet0" for both.
> >
> >VM1 works perfectly, but VM2 stops at "Starting network" point.
> >
> >Tcpdump says kernel mount root and happily reading
> >/sbin/init, /dev and lot of other files, but when console
> >says "Starting network" all communications are stopped. I
> >traced it on Dom0 via mac-address so I know there were no
> >NFS, no DHCP or other packet sent after this point.
> >
> >Starting network.
> >nfs send error 65 for 172.20.27.4:/export/nfsroot/ospf1
> >nfs send error 65 for 172.20.27.4:/export/nfsroot/ospf1
> >
> >Only two or three error messages and then it hangs forever.
> >
> >fstab contain no root FS and ifconfig.xennet0 contain only
> >"dhcp" option (it won't work with no ifconfig file either).
> >
> >Adding "set -x" to /etc/rc.d/network did not helped.
> >
> >How may I debug this issue or maybe someone have won this
> >battle already?
>
> I would add a set -x in /etc/rc.d/network perhaps in network_start()...
I added set -x in network_start and network_start_hostname() and got
very strange output:
ddb.onpanic: 1 -> 0
+ echo 'Starting network.'
+ command echo 'Starting network.'
Starting network.
+ 'network_start_hostname'
+ set -x
nfs send error 65 for 172.20.27.4:/export/nfsroot/ospf1
+ '[' -z '' ']'
nfs send error 65 for 172.20.27.4:/export/nfsroot/ospf1
nfs send error 65 for 172.20.27.4:/export/nfsroot/ospf1
It seems that in network_start() the route is already gone. Maybe
even it hangs on a call to /etc/myname.
--
Sincerely yours,
Dima Veselov
Physics R&D Establishment of Saint-Petersburg University
Home |
Main Index |
Thread Index |
Old Index