Date: Tue, 16 Jun 1998 17:07:51 -0500 From: "J.A. Terranson" <sysadmin@mfn.org> To: "'FreeBSD Questions'" <freebsd-questions@FreeBSD.ORG> Subject: Detailed cry for [bootp] help... Message-ID: <01BD9949.4C7BB540@slowbob.mfn.org>
next in thread | raw e-mail | index | archive | help
Ok. As i am now on my second day of this nightmare,
here's everything I know about what is/isn't going on. I
have searched the archives till I have the headache of
the century, but to no avail...
The scenario is this: I need to set up a bunch of diskless
workstations, all to run 2.2.5R, as well as the NFS/tftp/bootp
servers which service this group.
I have set up a test net of two machines, on their own wire.
SERVER:
/etc/rc.conf:
.......
amd_enable="NO"
nfs_client_enable="YES"
nfs_client_flags="-n 4"
nfs_server_enable="YES"
nfs_server_flags="-u -t 12"
mountd_flags="-nr"
nfs_reserved_port_only="NO"
rpc_lockd_enable="NO"
(I *would* enable lockd, but it is marked as broken???)
rpc_statd_enable="YES"
portmap_enable="YES"
portmap_flags=""
....
/ETC/EXPORTS:
/ -alldirs -ro -maproot=0 -network blah.blah...
/usr -alldirs -ro -maproot=0 -network ...
/tmp -network ...
/var -network ...
In the root directory is a symlink to /usr/tftpboot, therefore, in /TFTPBOOT:
drwxr-xr-x 4 root wheel 512 Jun 16 13:57 204.238.179.205
-rw-r--r-- 1 root wheel 258 Jun 16 16:23 cfg.204.238.179.205
The 204.238.179.205 directory has a full tree (/, /dev, /bin, /etc, blah..) under "rootfs" and the kernel sees the devices when probing, so I assume the /dev is ok. There is a 50k block swap file (dd's from /dev/zero) in "swapfs". The
kernel appears to see it when beginning to boot.
the contents of the cfg file are:
ip 204.238.179.205
server 204.238.179.35
netmask 255.255.255.0
hostname demo.mfn.org
kernel kernel
rootfs 204.238.179.35:/tftpboot/204.238.179.205/rootfs
swapfs 204.238.179.35:/tftpboot/204.238.179.205/swapfs
swapsize 50000
(I have also tried adding the line "rootopts rsize=1024,wsize=1024,resvport")
showmount -e shows /var, ./usr, and / all available to my netblock
ON THE CLIENT SIDE:
Booting from the floppy yields:
loading cfg.204.238.179.205
...
loading kernel
....
<normal probing>
npx0 ...
panic: nfs_mountroot: SIOCAIFADDR: 6
Now I have to assume that nfs mounts are ok,
since the kernel has been loaded, right?
The only thing that caught my eye was the load
address of the kernel when it started to boot:
...
entry 0xF0100000
I'm not used to seeing that first bit set, but then,
I've never done a diskless before, so maybe it's
normal...
Correct me if I'm wrong, but here's what it *looks* like to me:
(1) tftp is doing it's job, and then handing things off to bootp
(2) bootp is doing it's job, and handing things over to the kernel
(3) the kernel begins ok, and then after probing for devices,
attempts to mount it's filesystems (again?), and fails, causing
the panic.
On the assumption that this is the case, I have tried
(successfully) to mount these filesystems and directories
at the server ("just to see"). I have checked for /val/log/message
output from mountd etc., butt there is nothing to be had.
Has *anyone* ever gotten this to work under 2.2.5R? If so,
would you be willing to part with some setup info?
Desperately Yours,
J.A. Terranson
sysadmin@mfn.org
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?01BD9949.4C7BB540>