News:

Rule #1 - Be Patient - Rule #2 - Don't ask when, if you don't contribute - Rule #3 - You have coding skills - LinuxMCE's small brother is available: http://www.agocontrol.com

Main Menu

Update 10.04 to 12.04 = success (at least to me)

Started by polly, March 13, 2013, 11:42:43 PM

Previous topic - Next topic

polly

sorry, i planned to test asterisk.
asterisk starts but i havent had time to setup and test.

Current focus is on getting the update script issues solved im aware of. ;-)


polly

asterisk does not work for me atm! ....
tried to setup a phone line and it always saves something like:

Enabled    Prefix    Name    Phone number    Host    Protocol    Username    Status    Action
[  ]             default    SIP          Edit Incoming routing Delete

maybe its a lmce-admin related issue, cant tell today ...

phenigma

For anyone wondering what is working and what is not in 1204:  
There is a lot that is not working completely in 1204 at this point.  Very little has been tested.  

What users can do to help:
We need testers to identify the current issues and file trac tickets at svn.linuxmce.org so the issues can be looked at.  Be sure to search for an existing ticket identifying the same issue before creating a new one and then check the ticket regularly to see if there has been any action or questions.  This will help devs to trac the issues properly.

And, thank you to anyone who tests and files tickets for us!

J.

ps, Polly: you're doing a great job on the upgrade script.
My setup: [url="http://wiki.linuxmce.org/index.php/User:Phenigma"]http://wiki.linuxmce.org/index.php/User:Phenigma[/url]

polly

thanks phenigma :-) ... 

current state of the script:
the day befor yesterday it happend again! Something went horribly wrong on upgrading the MD.
It run roughly 14 houes (toooo long!), something during "apt-get -y --force-yes install pluto-confirm-dependencies" went wrong.
it wrote brackets at the very top of . /usr/pluto/bin/Config_Ops.sh.

Cheers!

phenigma

Have you tried upgrading the MDs through chroot on the core?  Upgrading the MD on the MD itself will be extremely slow due to the network file system.

The other option, and likely the best, is re-running Diskless_CreateTBZ.sh on the core, setting the needs configure flag in the db for each of the Generic PC as MD devices, and then running Diskless_Setup.sh to recreate the MD filesystems.

iirc, nothing in pluto-confirm-dependencies will alter /usr/pluto/bin/Config_Ops.sh

J.
My setup: [url="http://wiki.linuxmce.org/index.php/User:Phenigma"]http://wiki.linuxmce.org/index.php/User:Phenigma[/url]

polly

hm ... havent tried chroot yet ...
i will give it a try ....

thank you!


polly

hey ...

chroot into MD seems faster... BUT ...
i get " WARNING:root:estimate_kernel_in_boot() returned '0' " when the script is running do-release-upgrade.

any hints are greatly appreciated! :-)

Thanks.

phenigma

The chroot method will be MUCH faster.

Look at Diskless_CreateTBZ.sh and find the mount/unmount commands in it.  You will need to use the same mounts / umounts to run in chroot.

J.
My setup: [url="http://wiki.linuxmce.org/index.php/User:Phenigma"]http://wiki.linuxmce.org/index.php/User:Phenigma[/url]

polly

mh ...  :) ...
ok.... but not today....

ugh.... by the way, can't remember if i ran Diskless_CreateTBZ.sh after upgrading Core. Is it required prior to upgrading MDs?

Cheers,
ochorocho

phenigma

No, it is required prior to creating new MDs under 1204, but not for upgrading.

J.
My setup: [url="http://wiki.linuxmce.org/index.php/User:Phenigma"]http://wiki.linuxmce.org/index.php/User:Phenigma[/url]

polly

i had a look at Diskless_CreateTBZ.sh...

i reckon $TEMP_DIR is /usr/pluto/diskless/<MD number>  ?

This is what i did:
mount --bind /var/cache/apt $TEMP_DIR/var/cache/apt
mount --bind /usr/pluto/deb-cache $TEMP_DIR/usr/pluto/deb-cache
mount --bind /dev/pts $TEMP_DIR/dev/pts
mount none -t sysfs $TEMP_DIR/sys
mount none -t proc $TEMP_DIR/proc
chroot /usr/pluto/diskless/<MD number>


it gave the same error as before:
WARNING:root:estimate_kernel_in_boot() returned '0' ........

i got really no idea what to do now ... :-(

polly

sooooo .... MD Upgrade worked .... and it was really fast!

how it was done:
On the Core (10.04) i chroot'ed into the the MD (see attached file).
Executed the lmceUpgrade.sh as usual.
After all MDs are upgrade, Upgrade ur Core....

I was told to upgrade Core first, but this did not work for me! ....

I'll try to include the chroot thingie into the script.

Cheers,
ochorocho

polly

so far the upgrade on the MD worked ... except the MD shows errors on boot:

....
could not log bootup: address already in use
mount: according to mtab, proc is already mounted on /proc
....

maybe this is caused by chroot'ed method ...

see image for details.