Upgrading Solaris 11 Express b151a with support to Solaris 11

The most common problem that I am seeing on the aliases this morning is folks who are running 151a with a support update finding that their upgrade is failing.

The reason for this is that the version of pkg that you need to do the upgrade is in SRU#13. You need to update to this before switching to the release repository and upgrading.

This is an absolutely required step.

If you have an SRU older than #13 and have already switched to the release repository, you will need to switch back to the support repository, update and then go back to the release repository.

Comments:

My Solaris 11 Express test system, currently on SRU12, fails to update to SRU13, thus making it impossible to upgrade to Solaris 11 (final).

What happens is the following:

root@somehost:/# pkg publisher
PUBLISHER TYPE STATUS URI
solaris (preferred) origin online https:--pkg.oracle.com-solaris-support-

root@somehost:/# pkg update
Packages to remove: 59
Packages to update: 29
Create boot environment: Yes
Rebuild boot archive: Yes
[...]

It then proceeds to update a few and remove quite a lot of packages (e.g. the mpt_sas driver, which seems kind of a bad idea, given that the driver is currently in use by the system's LSI HBA...).

After booting into the newly created BE, it will just show the Solaris splash screen, and after a few seconds, crash and enter an endless cycle of reboots.

Is this a known issue?

(P.S.: I had to replace the slashes in the repository URL by dashes, or the comment would get flagged as spam.)

Posted by Clemens on November 10, 2011 at 04:05 PM EST #

OK this is a peculiar one.

One thing you might like to try is to boot up with the kernel debugger.

You do this at the grub screen by going hitting 'e' on the image you wish to boot.

One of the lines on the next page contains the boot arguments (I don't have a solaris x86 box in front of me right now). Hit 'e' on that one and add '-k' to the end of the line. Hit esc and then 'b' to boot it.

You can get rid of the splash screen by hitting any key after the system has started booting (or you can remove the args that do it on that line we were just editing).

Once it drops into the kernel debugger, that will stop the immediate reboot and you should see some text that will give some clues as to why it panicked.

If you want to see the kernel stack you can use the '::stack' command to kmdb.

I would seriously suggest getting a service request logged on this one.

Posted by Alan on November 11, 2011 at 12:34 PM EST #

Starting with a Solaris 11 Express (151.0.1.12) when I tried to pkg update I got an error like the following:

pkg set-publisher: The origin URIs for 'solaris' do not appear to point to a valid pkg repository.
Please verify the repository's location and the client's network configuration.
Additional details:

Unable to contact valid package repository
Encountered the following error(s):
Unable to contact any configured publishers.
This is likely a network configuration problem.
https protocol error: code: 403 reason: Forbidden
URL: 'https://pkg.oracle.com/solaris/support'.

while I am pretty sure that my Solaris support subscription ends not before mid 2012.

Then I tried to switch publishers, remove the current one (https://pkg.oracle.com/solaris/support) and add the release publisher ( http://pkg.oracle.com/solaris/release). Based on what you wrote, this wasn't the right step.

Then I tried to fall back to the initial support publisher using:

pkg set-publisher -g https://pkg.oracle.com/solaris/support/ -G http://pkg.oracle.com/solaris/release/ solaris

to start the whole process again, but it gave me the same error "https protocol error: code: 403 reason: Forbidden URL: 'https://pkg.oracle.com/solaris/support'" and I am now stuck with the release publisher.

btw: the new BE produced after the pkg update from http://pkg.oracle.com/solaris/release resulted to an un-bootable system and I had to switch to the original snv151 based BE.

Now, I am wondering what is the correct procedure to upgrade from snv_151 to snv_175 and if there is any, where I can find it.

Posted by nvoutsin on November 15, 2011 at 03:11 PM EST #

You may need to re-associate your keys as well. The last time I saw this message was when I tried to connect to that url with a browser on which I had not put the keys in.

alan

Posted by Alan on November 16, 2011 at 03:01 AM EST #

exactly the same problem as Clemens:

Here are my output from the kernel debugger after reboot with the ne created BE:
(written off the screen ;-)

Loading kmdb...
SunOS Release 5.11 Version 151.0.1.12 64-bit
Copyright (c)...
Notice: Can not read the pool label from '/pci@0,0/pci-ide@14,1/ide@0/cmdk@1,0:a'
Notice: spa_import_rootpool: error 5
Cannot mount root on /pci@0,0/pci-ide@14,1/ide@0/cmdk@1,0:a fstype zfs

panic[cpu0]/thread=fffffffffbc2f2a0: vfs_mountroot: cannot mount root

Warning - stack not written to the dump buffer
fffffffffbc71960 genunix:vfs_mountroot+33e ()
fffffffffbc71990 genunix:main+136 ()
fffffffffbc719a0 unix:_locore_start+90 ()

panic: entering debugger (no dump device, continue to reboot)

Welcome to kmdb
kmdb: dmod krtld failed to load: Error 2
Loaded modules: {zfs]
[0]> ::stack
kmdb_enter+0xb()
debug_enter+0x38(fffffffffb962fc8)
panicsys+0x544(fffffffffbbeeef8, fffffffffbc718f0, fffffffffbc79af0, 1)
vpanic+0x15c()
panic+0x94()
vfs_mountroot+0x33e()
main+0x136()
_locore_start+0x90()
[0]>

Any ideas?
BR
Jan

Posted by guest on November 17, 2011 at 11:17 PM EST #

Post a Comment:
Comments are closed for this entry.
About

* - Solaris and Network Domain, Technical Support Centre


Alan is a kernel and performance engineer based in Australia who tends to have the nasty calls gravitate towards him

Search

Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today
Links
Blogroll

No bookmarks in folder

Sun Folk

No bookmarks in folder

Non-Sun Folk
Non-Sun Folks

No bookmarks in folder