How to use live-upgrade with single disk, pre-patching steps?

UNIX Solaris
1

Hi,

I have Solaris-10 x86 (running on HP hardware), with 12 non-global zones running on this. I have to install latest patch cluster/set on this server. This server is not under backup schedule, so before installing patch cluster, I want to keep a backup. In case of any issue (bad patch or application break), I should be able to revert back to pre-patching status.
Now challenge is, due to hardware RAID, I can see only one disk and one zpool. There is sufficient space.
I am get examples of create image on alternate disk or alternate pool. But how will I do it when I have single disk and single zpool ?
Suggest please.

# zpool list
NAME    SIZE  ALLOC   FREE    CAP  HEALTH  ALTROOT
rpool  1.81T  84.1G  1.73T     4%  ONLINE  -
# echo|format
Searching for disks...done


AVAILABLE DISK SELECTIONS:
       0. c0t0d0 <DEFAULT cyl 60794 alt 2 hd 255 sec 252>
          /pci@0,0/pci8086,3408@1/pci103c,3245@0/sd@0,0
Specify disk (enter its number): Specify disk (enter its number):
#

Thanks

2

On Solaris 10 you would use

lucreate

This allows the system to create an alternative boot environment; BE.

This new BE created on the disk but, the new BE can be patched. Once patched you
are able to active with new BE with cluster patch, then when you bounce the system
you'll be booted into the patched BE.

From here you'll be able to test applications and zones to see if the patch had
any negative affects. If so, simply revert back to the original BE and bounce system
to pre-patch state. Open a SR with Oracle if you have a contract.

There's tons of documentation on

lucreate
3

After re-reading your post I realized that my original response was, incorrect for what you needed.

I'm not sure about the possibility of creating another
zpool from a standalone rpool. I doubt it's even possible.

To create a zpool, you need a separate disk, you are
able to create a zfs dataset from the rpool but not recommended.

As for the alternative BE to apply patches to, I believe lucreate
makes/has a pre-defined directory which it places the alternative
BEs'.

Let me know if you where able to get this resolved as this
will help solaris admins alike.

4

I was reading few posts and I think this could work

lucreate -c oldBE -n new-zfsBE

And then I can run patch set on new-zfsBE
All my zones are residing on rpool, as there is only one pool, so I will have option to revert to old environment, including all zones.
I am yet to test this.

5

*****UPDATE*****
I was able to create new-zfsBE on same disk and install patchset on that BE. After completion, I activated that and rebooted (with init 6) and all was good.