Reboot problem !!! urgent plse !!!!!!!!!!!

Operatin System : Solaris 5.9
Server : Sun Fire 3800

Shutdown and reboot i receive the following messages .....

un Fire 3800
OpenFirmware version 5.15.2 (08/04/03 10:27)
Copyright 2001-2003 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
SmartFirmware, Copyright (C) 1996-2001. All rights reserved.
4096 MB memory installed, Serial #xxxxxxx.
Ethernet address 8:0:20:e8:e0:c2, Host ID: xxxxxxxxx.

Rebooting with command: boot
SunOS Release 5.9 Version Generic_112233-03 64-bit
Copyright 1983-2002 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
WARNING: /ssm@0,0/pci@1c,700000/pci@2/SUNW,isptwo@4/sd@6,0 (sd21):
Corrupt label; wrong magic number

Hardware watchdog enabled
configuring IPv4 interfaces: hme0.
Hostname: server_name
The / file system (/dev/rdsk/c0t0d0s0) is being checked.
/dev/rdsk/c0t0d0s0: UNREF DIR I=12231 OWNER=root MODE=40755
/dev/rdsk/c0t0d0s0: SIZE=512 MTIME=Dec 12 09:26 2005 (RECONNECTED)
/dev/rdsk/c0t0d0s0: DIR I=12231 CONNECTED. PARENT WAS I=2
/dev/rdsk/c0t0d0s0: LINK COUNT DIR I=2 OWNER=root MODE=40755
/dev/rdsk/c0t0d0s0: SIZE=1024 MTIME=Jan 5 15:54 2006 COUNT 27 SHOULD BE 26 (ADJUSTED)
/dev/rdsk/c0t0d0s0: LINK COUNT lost+found I=3 OWNER=root MODE=40700
/dev/rdsk/c0t0d0s0: SIZE=8192 MTIME=Nov 22 14:46 2005 COUNT 3 SHOULD BE 4
/dev/rdsk/c0t0d0s0: LINK COUNT INCREASING
/dev/rdsk/c0t0d0s0: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.

WARNING - Unable to repair the / filesystem. Run fsck
manually (fsck -F ufs /dev/rdsk/c0t0d0s0). Exit the shell when
done to continue the boot process.

Type control-d to proceed with normal startup,
(or give root password for system maintenance): resuming system initialization
mount: the state of /dev/dsk/c0t0d0s0 is not okay
and it was attempted to be mounted read/write
mount: Please run fsck and try again
/sbin/rcS: /etc/dfs/sharetab: cannot create
failed to open /etc/coreadm.conf: Read-only file system

INIT: Cannot create /var/adm/utmpx

INIT: failed write of utmpx entry:" "

INIT: failed write of utmpx entry:" "

INIT: SINGLE USER MODE

Type control-d to proceed with normal startup,
(or give root password for system maintenance):
single-user privilege assigned to /dev/console.
Entering System Maintenance Mode

What's happened ????
I tryed run manually command ( fsck -F ufs /dev/rdsk/c0t0d0s0 ) but ask me some things that i don't know )

For example : * Phase 4 - Check Reference Counts
LINK COUNT DIR I=3 OWNER=root MODE=40700
SIZE=8192 MTIME=Nov 22 14:46 2005 COUNT 2 SHOULD BE 3
ADJUST? no

UNREF DIRECTORY I=12231 OWNER=root MODE=40755
SIZE=512 MTIME=Dec 12 09:26 2005
CLEAR?

UNREF FILE I=15956 OWNER=root MODE=100444
SIZE=706 MTIME=Dec 7 11:53 2005
CLEAR? no

UNREF FILE I=15980 OWNER=root MODE=100644
SIZE=352 MTIME=Dec 7 11:53 2005
CLEAR?

UNREF FILE I=15996 OWNER=root MODE=100644
SIZE=12246 MTIME=Dec 7 11:54 2005
CLEAR? n

UNREF FILE I=224809 OWNER=root MODE=100644
SIZE=95080448 MTIME=Dec 7 11:54 2005
CLEAR?
LINK COUNT DIR I=247411 OWNER=root MODE=40755
SIZE=512 MTIME=Dec 7 16:06 2005 COUNT 0 SHOULD BE 2
ADJUST? no

UNREF FILE I=347550 OWNER=ldap MODE=100755
SIZE=614660 MTIME=Nov 20 18:22 2002
CLEAR?
UNREF FILE I=347551 OWNER=ldap MODE=100755
SIZE=9636 MTIME=Nov 20 18:20 2002
CLEAR? no

UNREF FILE I=392005 OWNER=ldap MODE=100755
SIZE=730288 MTIME=Nov 20 19:10 2002
CLEAR?
UNREF FILE I=425706 OWNER=root MODE=100644
SIZE=0 MTIME=Dec 7 11:54 2005
CLEAR? no

UNREF FILE I=448096 OWNER=ldap MODE=100644
SIZE=1426216 MTIME=Nov 20 18:30 2002
CLEAR?
UNREF FILE I=448105 OWNER=ldap MODE=100755
SIZE=80656 MTIME=Nov 20 18:24 2002
CLEAR? no

UNREF FILE I=448109 OWNER=ldap MODE=100755
SIZE=41744 MTIME=Nov 20 18:19 2002
CLEAR?

UNREF FILE I=448110 OWNER=ldap MODE=100755
SIZE=4419884 MTIME=Nov 20 18:44 2002
CLEAR? no

UNREF FILE I=448111 OWNER=ldap MODE=100755
SIZE=60344 MTIME=Nov 20 18:24 2002
CLEAR?

** Phase 5 - Check Cyl groups

FILE SYSTEM STATE IN SUPERBLOCK IS WRONG; FIX? no

58433 files, 1579802 used, 350113 free (2729 frags, 43423 blocks, 0.1% fragmentation)

I response no to all because i don't know what's mean ..........

The HD is done ??

Plse help me !!!!!!!!!!!!!!!!!!!!!!

Tnks in advance

Run it again and respond "yes" to everything. That will probably work. When you respond "no" no attempt to fix anything is happening. If fsck can't fix it, the data on your disk is probably lost. But the disk itself may be ok is a hardware sense.

If you really shutdown correctly, something is very wrong with your shutdown procedure. This looks like someone just pulled the plug instead of doing an orderly shutdown.

What me makes a little scary is that you are using a single boot disk on a 200.000 $ + server which probably houses important data.

...
/ : /dev/(r)dsk/c0t0d0s0
...

Normally, and I dont know one server where not, the boot disk is at least mirrored, for the best in two seperate disk trays over different controllers in different I/O boards. Adding a spare disk / contegency disk is also a good option to do.

The 3800 has space for two I/O cPCI boards , so you can mirror the boot disk with no single-point-of-failure when you have two disk trays like the D240.

But that your filesystem crashed is another story...�
Hope you can restore from tape ..

I've noticed that a number of people tend to use 'reboot' on systems, but have placed either . or /etc/ in their PATH. Solaris has a /etc/reboot which is a symbolic link to halt, thus no orderly shutdown:

$ ls -l /etc/reboot
lrwxrwxrwx 1 root root 16 Apr 17 2002 /etc/reboot -> ../usr/sbin/halt

I tend to favor shutdown for reboot:

shutdown -i6 -g0 -y

and power off:

shutdown -i5 -g0 -y

Cheers,

Keith

sun actually recommends using "/usr/sbin/init 6" for reboots and "/usr/sbin/init 5" for powerdowns --- shorter and still does basically the same thing ...

The nice thing about shutdown is you can append a system message to the end of the call. I list the reason for the shutdown as the message so system logs contain a bit more information:

shutdown -i6 -g0 -y "=== System going down to upgrade RAM - kduffin - CMR 1145 ==="

Cheers,

Keith

Sun has two shutdown commands. The UCB shutdown is a dangerous c program that does do some logging. The real shutdown command is shell script and I don't see any logging there.

I came from an HP-UX environment. On HP-UX, using init to shutdown is like pulling the plug. I got real used to the shutdown command. This is one of the reasons Sun admins have a rough time with HP-UX. They usually manage one shutdown via the init command... although they never do it again. This also may explain why Sun recommends using init!

On Suns, I use the shutdown command because it warns the users... including making an attempt at warning nfs clients. I try to give 5 - 10 minutes warnings. I can abort the shutdown if someone calls. This whole warning stuff gives me some political cover.

Ouch - you're right. No syslog, just 'wall'.

Run this from single user:

fsck -Fufs -y /dev/rdsk/c0t0d0s0