Re: Serious Problems with diskless clients

Wakko Warner (wakko@animx.eu.org)
Thu, 26 Sep 2002 07:18:47 -0400


> my diskless clients have some severe problems on one of my servers.
> Sometimes (right now most of the time) everything just hangs at the same place when
> starting up the kernel. Here are the last messages I get (right before this
> IP-Config is running and looks OK):
>
> NET4: Unix domain sockets 1.0/SMP for Linux NET4.0
> ds: no socket drivers loaded !
> Looking up port of RPC 100003/2 on 192.168.0.235
> portmap: server 192.168.0.235 mot responding, timed out !
> Root-NFS: Unable to get nfsd port number from server, using default
> Looking up port of RPC 100005/1 on 192.168.0.235
> portmap: server 192.168.0.235 mot responding, timed out !
> Root-NFS: Unable to get mountd port number from server, using default
> mount: server 192.168.0.235 not responding, timed out
> Root-NFS: Server returned error -5 while mounting /netclients/192.168.0.87
> VFS: Unable to mount root fs via NFS, trying floppy
> VFS: Insert root floppy and press ENTER
>
> I am thinking right now that we have some problems with network hardware,
> but maybe its a Software problem. Could someone tell me what the 'Looking up
> port of RPC 100003/2 on 192.168.0.235' in kernel startup is doing an why it
> could fail ?
>
> We have Kernel 2.4.10 on both server and clients (I also tried 2.4.19, but
> it changed nothing).

I have 2 diskless machines both tftping the kernel from the network and they
work just fine. Both are using kernel 2.4.19 vanalla with a small patch to
force ip=auto if nfs=/dev/nfs

[wakko@gohan:/] uname -a
Linux gohan 2.4.19 #1 SMP Tue Sep 3 13:02:36 EDT 2002 i686 unknown
[wakko@gohan:/] mount
rod:/tftpboot/gohan on / type nfs
(rw,intr,hard,rsize=8192,wsize=8192,intr,hard,rsize=8192,wsize=8192)
/proc on /proc type proc (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
[wakko@gohan:/]

It's a very small patch to make that work w/o saying ip=auto on the command
line (I did this originally because I used a kernel disk w/o a boot loader
on some boxes at work)

--- net/ipv4/ipconfig-orig.c 2001-11-19 20:48:35.000000000 -0500
+++ net/ipv4/ipconfig.c 2001-11-19 20:56:21.000000000 -0500
@@ -1105,7 +1105,11 @@
proc_net_create("pnp", 0, pnp_get_info);
#endif /* CONFIG_PROC_FS */

- if (!ic_enable)
+ if (!ic_enable
+#if defined(IPCONFIG_DYNAMIC) && defined(CONFIG_ROOT_NFS)
+ && ROOT_DEV != MKDEV(UNNAMED_MAJOR, 255)
+#endif
+ )
return 0;

I did a cut'n'paste so I don't know if it will apply correctly but you get
the idea.

One thing I had problems with was using USB to mount the rootfs (usb hdd
actually) and I suspect the same with a USB nic (loading the kernel from
floppy).

-- 
 Lab tests show that use of micro$oft causes cancer in lab animals
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/