Re: [Ltsp-discuss] "ioctl SIOCSARP: Network is unreachable" error

[Jim M. <jam@McQuil.com>]

Peter,

If they are all on the same physical network, and you
get 'Network is unreachable', that probably means that the
workstation is sending an Arp request, but not getting an arp
reply from the server.

It would be interesting to run tcpdump on the server and watch
the traffic for one of the workstations.

You can do this:

   tcpdump host x.x.x.x

where the x's are replaced with the IP address of the workstation.

We already know that Etherboot is getting the kernel loaded, but for
some reason the network driver in the kernel is unable to talk on the
net.

On the server, try listing the arp table.  You can use 'arp -an' to
list all of the entries.  Take a look in the listing for MAC
address of the workstation.  Entries don't stay in the arp cache
for very long, so you should try to do it immediately after the
failure.

Also, do a 'tail -f /var/log/messages' and watch for the bootp requests
to be answered.

Also, is there a switch on the network or a hub?

If it's a switch, try cycling the power on the switch.

Hope that helps,
Jim McQuillan
ja...@lt...


Peter Bates wrote:
> 
> Hello all...
> 
> First of all, I'd like to thank everyone
> involved in the LTSP project for such great work...
> 
> We've been running a number of diskless terminals
> using the LTSP configs and stuff for a while now
> for email access, and all is fine...
> 
> Fiddling about today, however, I suddenly have two
> terminals (they're 200MHz PCs with 3Com 3C905C NICs)
> failing to boot at all, and I'm just getting this error:
> 
> Apr 24 20:10:53 icecube bootpd[721]: ioctl SIOCSARP: Network is unreachable
> 
> from the system doing the bootp...
> 
> The back-end server bit is running on a RH 6.2 box,
> but is a seperate system from the bootp server, but
> all of this has worked fine for ages, until I started
> changing some X-settings today, and that is surely
> not causing the problem here, because the other
> terminals are fine...
> 
> Does anyone know what is causing this, as hardware
> failure on two machines next to each other seems unlikely... ?
> 
> --
> ---------------------------------------------------------------->
> Peter Bates, Systems Support Officer, Network Support Team.
> London School of Hygiene & Tropical Medicine.
> Telephone:0207-927 2124 / Fax:0207-436 5389 / Pager: 07625 255362
> 
> _______________________________________________
> Ltsp-discuss mailing list
> Lts...@li...
> http://lists.sourceforge.net/lists/listinfo/ltsp-discuss