[Sunhelp] Re:Solaris 7 Help

[Balaji srinivasan]

Hi All,
         Thanks for everyone who replied. I came to know that it's a bug and
it can be fixed in the following manner. Thanks to the Tech Support guy from
Akibia.  I'm attaching the document for you guys reference.

Thanks,
Balaji
-----------------------
Unix Administrator
Segue Software Inc..,
201,Spring Street
Lexington,MA-02421
Ph: 781-402-5898
----------------------------------------------------------------------
SRDB ID: 20768

SYNOPSIS: Jumbo Kernel Patch 106541-## gives utmp and utmpx  errors

DETAIL DESCRIPTION:

After applying the Jumbo Kernel Patch for Solaris 7, the system will
fail to
boot.  Customer reports the following:  "I have just installed
kernel update
patch 106541 and now my machine will not boot.  In fact, when I
choose an
offered run level, it just keeps looping and giving me an option to
choose
another run level.  The customer will see the following on his
screen:

Boot device: /iommu/sbus/espdma at f,400000/esp at f,800000/sd at 3,0:a  File
and args: 

SunOS Release 5.7 Version Generic_106541-## [UNIX(R) System V
Release 4.0]
Copyright (c) 1983-1999, Sun Microsystems, Inc.
|
INIT: Cannot create /var/adm/utmp or /var/adm/utmpx
INIT: failed write of utmpx entry:"  "
INIT: failed write of utmpx entry:"  "
INIT: SINGLE USER MODE
ENTER RUN LEVEL (0-6, s or S): 

SOLUTION SUMMARY:

                                                
This problem has primarily been seen as an issue with FCS of Solaris
2.7, 
released in October of 1998 and Jumbo Kernel Patches 106541-05 and
-07.

The following provides a quick fix to get the system up and working
again.  

        1) At the ok prompt, type boot cdrom -s
        
        2) mount the boot disk's root partition 
                ex.: mount /dev/dsk/c0t0d0s0 /a
                
        3) ls -l /a/devices/pseudo/sys*

           If there is NO MATCH, then this is the key that the
kernel patch is the culprit. Proceed to step 4.  
        
           If there is a match, then it is likely that something
else is responsible for the NO BOOT issue.  For example,
the cu just edited vfstab and the mount of /var is incorrect.

        4) Get the "postpatch" script from the 106541-05 (or -07)
           directory. It might still be on the hard drive, or it can be
moved onto the system via ftp or floppy.

        5) From the patch directory, do the following:
        
                # ROOTDIR=/a
                # export ROOTDIR
                # chmod 755 postpatch
                # ./postpatch
                
           You should get an error aboutr the driver already being
loaded.
     
        6) Do the following:
        
                ls -l /a/devices/pseudo/sys*
                
           You should now see the devices sysmsg at 0:msglog and
sysmsg at 0:sysmsg

        7) Now unmount the disk and reboot
        
                # cd /
                # umount /a
                # reboot



If the above steps don't resolve the issue, then the customer can
try the 
following:

Using CLI to workaround Bug ID 4228999:
======================================

BEFORE rebooting the server:


Recreate the devlink by executing the following commands.

        # grep sysmsg /etc/name_to_major
        sysmsg 66
        # mknod /devices/pseudo/sysmsg at 0:sysmsg c <major_num> 0
        # mknod /devices/pseudo/sysmsg at 0:msglog c <major_num> 0
        
  * Where <major_num> is the number following sysmsg in the grep
output.

Change the permissions and ownerships of the created devlinks.

        # chmod 0600 /devices/pseudo/sysmsg at 0:sysmsg 
        # chmod 0600 /devices/pseudo/sysmsg at 0:msglog
        # chown root:sys /devices/pseudo/sysmsg at 0:sysmsg
        # chown root:sys /devices/pseudo/sysmsg at 0:msglog
        
Create the links to the /dev entries.

        # cd /dev
        # rm -f sysmsg ; ln -s ../devices/pseudo/sysmsg at 0:sysmsg
sysmsg
        # rm -f msglog ; ln -s ../devices/pseudo/sysmsg at 0:msglog
msglog
        

If the preceding steps are NOT taken prior to rebooting the server,
AND it 
fails to boot with utmpx errors, boot the server to single user from
net or 
cdrom, mount the root filesystem to /a, and execute the following
commands.

Recreate the devlink by executing the following commands.

        # grep sysmsg /a/etc/name_to_major
        sysmsg 66
        # mknod /a/devices/pseudo/sysmsg at 0:sysmsg c <major_num> 0
        # mknod /a/devices/pseudo/sysmsg at 0:msglog c <major_num> 0
        
  * Where <major_num> is the number following sysmsg in the grep
output.
  
Change the permissions and ownerships of the created devlinks.

        # chmod 0600 /a/devices/pseudo/sysmsg at 0:sysmsg 
        # chmod 0600 /a/devices/pseudo/sysmsg at 0:msglog
        # chown root:sys /a/devices/pseudo/sysmsg at 0:sysmsg
        # chown root:sys /a/devices/pseudo/sysmsg at 0:msglog
        
Create the links to the /dev entries.

        # cd /a/dev
        # rm -f sysmsg ; ln -s ../devices/pseudo/sysmsg at 0:sysmsg
sysmsg
        # rm -f msglog ; ln -s ../devices/pseudo/sysmsg at 0:msglog
msglog

Unmount the root filesystem from /a and reboot from boot-device.
                                                    
                                       
                                       
                       
                                

BUG REPORT ID: 422899
PATCH ID: 106541
PRODUCT AREA: Kernel
PRODUCT: hang
SUNOS RELEASE: Solaris 7
UNBUNDLED RELEASE: FCS
HARDWARE: any





_______________________________________________________
Say Bye to Slow Internet!
http://www.home.com/xinbox/signup.html