Restoring to previous Boot Environment


 
Thread Tools Search this Thread
Operating Systems Solaris Restoring to previous Boot Environment
# 1  
Old 06-07-2011
Restoring to previous Boot Environment

Hi all,

I'm fairly new to Solaris and am just getting to grips with using LU (Live Upgrade) for OS patching purposes.

Code:
worcester#uname -a
SunOS worcester 5.10 Generic_144488-12 sun4v sparc SUNW,SPARC-Enterprise-T5220

I have successfully created and patched a new BE (boot environment) using the lucreate/luactivate commands.

On activating the patched BE I received the following instructions:

Code:
  worcester#luactivate patching20110606
  A Live Upgrade Sync operation will be performed on startup of boot environment <patching20110606>.
  
  **********************************************************************
The target boot environment has been activated. It will be used when you reboot. NOTE: You MUST NOT USE the reboot, halt, or uadmin commands. 
You MUST USE either the init or the shutdown command when you reboot. If you do not use either init or shutdown, the system will not boot using the target BE.
  **********************************************************************
  
  In case of a failure while booting to the target BE, the following process needs to be followed to fallback to the currently working boot environment:
  
  1. Enter the PROM monitor (ok prompt).
  
  2. Boot the machine to Single User mode using a different boot device
  (like the Solaris Install CD or Network). Examples:
  
       At the PROM monitor (ok prompt):
       For boot to Solaris CD:  boot cdrom -s
       For boot to network:     boot net -s
  
  3. Mount the Current boot environment root slice to some directory (like /mnt). You can use the following commands in sequence to mount the BE:
  
       zpool import rpool
       zfs inherit -r mountpoint rpool/ROOT/patching
       zfs set mountpoint=<mountpointName> rpool/ROOT/patching
       zfs mount rpool/ROOT/patching
  
  4. Run <luactivate> utility with out any arguments from the Parent boot environment root slice, as shown below:
  
       <mountpointName>/sbin/luactivate
  
  5. luactivate, activates the previous working boot environment and
  indicates the result.
  
  6. Exit Single User mode and reboot the machine.
  
  **********************************************************************
  
  Modifying boot archive service
  Activation of boot environment <patching20110606> successful.

On rebooting the server, everything comes back up fine and my OS is now fully patched. This is a test server I am working on and want to get recovery procedures documented before I apply this to my live servers. Therefore i tried rolling back to the previous BE (pre patched)

Following the instructions given when running the luactivate command, i booted into single user mode and first ran
Code:
zpool import rpool

This presented me with lots of failures saying
Code:
Cannot mount '/export': failed to create mountpoint
Cannot mount '/rpool': failed to create mountpoint

for all my mountpoints. Is this correct?

I ignored these errors as I've read in some other forums that this can be ignored (not sure how true this is..)

The next few steps were, it seems, completed successfully
Code:
zfs inherit -r mountpoint rpool/ROOT/patching 
zfs set mountpoint=/mnt rpool/ROOT/patching
zfs mount rpool/ROOT/patching

The final step was to run
Code:
<mountpointName>/sbin/luactivate

This resulted in an error of
Code:
luactivate: ERROR: Live Upgrade not installed properly (/etc/default/lu not found).

Could anyone offer any assistance or see that I'm missing anything out please?

Any other output that is needed, please let me know.

thank you.

Last edited by pludi; 06-12-2011 at 12:42 PM..
# 2  
Old 06-07-2011
You did boot into single user mode from the network or DVD? Correct?
# 3  
Old 06-07-2011
Yes - I booted, from DVD, into single user mode using
Code:
boot cdrom -s

# 4  
Old 06-11-2011
Hi, am facing same issue. My release is Solaris 10 5/09 s10s_u7wos_08. Not able to activate old BE , by booting thru net.
# 5  
Old 06-12-2011
Well, I raised a call with Oracle and received the following response:

Quote:
Anyhow, from the brief description I believe you hit known bug
CR 6923286 "luactivate fails in single user mode during fallback activation"
Fixed Version s10u10_09
The fix is not in the latest publicly available LU patch.

If the issue is causing considerable ongoing impact and preventing you using live upgrade, then we could provide you with an IDR patch which contains a number of LU test fixes including a fix for 6923286 . This will require an escalation against the bug and the SR will probably need to remain open while installed so you can report back any side effects from the IDR patch.
IDR147198

A workaround would be to instead "boot -L" and select the good BE.
e.g.
ok> boot -L
I'll be trying the boot -L command tomorrow and let you know how i get on..
# 6  
Old 07-29-2011
Just thought I'd give an update on this issue to say that i've successfully used the boot -L workaround to boot to an old BE.

Oracle have been able to reproduce this behaviour in their lab. They said that this is a known bug addressed in:

Bug 6996301 : s10u9 zfs root PBE/ABE in same pool on-screen luactivate instructions don't work

and

Bug 6923286 : luactivate fails in single user mode during fallback activation

There's no official fix available currently.
Login or Register to Ask a Question

Previous Thread | Next Thread

9 More Discussions You Might Find Interesting

1. Solaris

T5-4 Bootloop with 11.4 Boot Environment

TLDR: Patched a separate boot environment to 11.4. Activated and rebooted the system only to have it boot loop and show 11.3 on the patched boot environment. Bootloop process: The system went down, came up showing 11.3 then spewed bunch of errors and resetted. SEE BELOW FOR THE FINAL MSG... (1 Reply)
Discussion started by: samthewildone
1 Replies

2. UNIX for Beginners Questions & Answers

Lucreate Fails to Create Boot Environment;

My OS solarius 5.10 Generic_147148-2 i86 Error: please review new boots environments using options 1. Solution - show me the commands Partition is full, try to remove some unneeded files, then try to compress some other unneeded files. man command creates a temp file under... (0 Replies)
Discussion started by: zbest1966
0 Replies

3. Solaris

Automating old Boot Environment Cleanup Solaris 11

I'm trying to automate the patching process using scripts and cronjobs in Solaris 11. one of the things I'd like to do is cleanup the old boot environments. unfortunately beadm destroy requires a response :~$ pfexec beadm destroy -f solaris-13 Are you sure you want to destroy... (3 Replies)
Discussion started by: os2mac
3 Replies

4. Solaris

unable to boot afer restoring root partion manages by veritas

Hi all, After restoring the / FS on a netra 240. It is unable to boot the system. The message displayed is the following: "The file just loaded does not appear to be executable" Bellow, the content of /etc/vfstab before restoring: #more /etc/vfstab... (0 Replies)
Discussion started by: faberarnold
0 Replies

5. Solaris

Update single zone in alternate boot environment.

I 'm having a weired situation my system has 8 zones, running fine on solaris x86_u4. I installed the live upgrade bundle patch and did a live upgrade. The new BE was created but it missed one of the zone and now if i mount the new BE i do not see that zone in the environment so my question is how... (3 Replies)
Discussion started by: fugitive
3 Replies

6. UNIX for Dummies Questions & Answers

Messed up my boot environment or root profile

Ok, a couple weeks ago I was fixing a cron report about perl not happy with 'locale' info (LANG and LC not set). As a result, I was experimenting with setting the correct 'locale' in several areas (like /etc/sysconfig/i18n and who knows where). Somehow after a reboot, as soon as the OS starts... (3 Replies)
Discussion started by: Garball
3 Replies

7. Solaris

Restoring Boot Manager after Windows Install

Hi everyone, so i have a 40Gb disk with Windows XP and Solaris 10 Installed... windows has once again become unoperable for virus and stufff, and i need to reinstall it, but when i do that the MBR will be overwriten right? and i wont be able to access solaris10 anymore, is there a way, like a... (1 Reply)
Discussion started by: sx3v1l_1n51de
1 Replies

8. SCO

Wanting to boot from a previous kernel

I tried to add ppp on a serial line tty1b and I relink the kernel. Now, I would like to boot from the previous kernel because I lost the connexion for the others terminal (serial lines) too. I would appreciate any help. (3 Replies)
Discussion started by: pacctono
3 Replies

9. UNIX for Dummies Questions & Answers

Adding to boot-up environment

During the boot-up process *nix runs scripts linked into the runlevel directories rc#.d. What I'm wondering is, how do I control the environment that those scripts see? I need to set a couple environment variables, and I can NOT do it from within the scripts because it poses a maintenance nightmare... (1 Reply)
Discussion started by: DreamWarrior
1 Replies
Login or Register to Ask a Question