15 Sep
2003
15 Sep
'03
11:03 p.m.
On Sun, 7 Sep 2003, Jochen Kunz wrote:
This is why non-broken rarpd servers do this themselves :)
--f
root on le0
nfs_boot: trying RARP (and RFC/bootparam)
nfs_boot: client_addr=192.0.0.10 (RARP from 192.0.0.1)
nfs_boot: timeout
I had (possibly) the same error when I netbooted a SPARCstation
1+ from
a FreeBSD machine. It was quite simple. The primitive IP stack of the
netloader on the client machine can't answer arp requests. But the
server broadcasts a arp (arp, not rarp) when the tftp or NFS action is
starting. Solution was simple: Set the arp information for the client
on the host paermanent. Somthing like
arp -s 192.0.0.10 00:11:22:33:44:55 permanent
on the boot server.