[one-users] Feedback After Upgrade

[Duverne, Cyrille]

Found something :

Working host :
$ ls -lArth /var/lib/one//datastores/0/162/disk.0
-rw-r----- 1 libvirt-qemu kvm 3.0G Sep 30 19:21
/var/lib/one//datastores/0/162/disk.0

Failing host :
$ ls -lArth /var/lib/one//datastores/0/165/disk.0
-rw-r----- 1 oneadmin oneadmin 3.0G Sep 30 19:17
/var/lib/one//datastores/0/165/disk.0

How could I solve this ?

Cyrille Duverne
Consultant
Eura Nova

GSM : +32 (0) 493 25 32 63
Blog : http://blog.cduverne.com
Twitter : @CydsWorld

"Imagination is more important than Knowledge"
Albert Einstein

Dimanche 30/09/2012 à 20:05 Duverne, Cyrille a écrit:

Hello, 
Thanks for this quick feedback.

here below my environment :

OS : Ubuntu Server 12.04 - 64bits
OpenNebula 3.6
Installed via opennebula official .deb package

let's take a concreat example : Working template, working image,
working network etc :

Host 1 : SUCCESS deployment :
#id oneadmin
uid=107(oneadmin) gid=114(oneadmin)
groups=114(oneadmin),115(kvm),116(libvirtd)

qemu.conf is filled of commented lines (no specific change)

Apparmor : nothing special

Host 2 : FAILED deployment

#id oneadmin
uid=107(oneadmin) gid=114(oneadmin)
groups=114(oneadmin),115(kvm),116(libvirtd)

qemu.conf is filled of commented lines (no specific change)

Apparmor : nothing special

The hosts are looking identical...

Thanks in advance

Cyrille
Blog : http://blog.cduverne.com
Twitter : @CydsWorld

"Imagination is more important than Knowledge"
Albert Einstein

Dimanche 30/09/2012 à 19:54 Rodolfo Conte Brufatto a écrit:

Ah, btw don know which distro or OS you're using, but if you have
upgraded from a binary package... be aware that the oneadmin group has
changed (from cloud to oneadmin as well). So again, re-check the
permissions and then give it another shot.

Regards,

On Sun, Sep 30, 2012 at 2:51 PM, Rodolfo Conte Brufatto  wrote:

Have you checked the permissions on datastores dir?
Also since by default it uses shared fs. I don't know how was your set
up before the upgrade... 
But it is worth to check the dynamic_ownership option in qemu
configuration file and also apparmor settings.
I did the same upgrade but i had an issue with the databases (mysql
backend), so I ended up reinstalling the new database again.
I haven't tried to instantiate any VMs since i need to configure all
my old images and i am between a dilema in my environment in which
datastore to use since i was using file based disks.
I will perform deep checkings in my environment tho and see if I face
the same problema as yours.

Regards,

On Sun, Sep 30, 2012 at 2:04 PM, Duverne, Cyrille  wrote:

Hello,

Following the 3.2 to 3.6 upgrade, I have the following issues :

1/ the one.db still contains a template indicating to be used by 4
VMs, except that no VMs are using this one. How could I fix it ?

2/ One of my hosts refuses to create new VMs (which was working before
the upgrade), I didn't change anything on this one and here below the
error message :
Sun Sep 30 18:57:09 2012 [DiM][I]: New VM state is ACTIVE. Sun Sep 30
18:57:11 2012 [LCM][I]: New VM state is PROLOG. Sun Sep 30 19:00:33
2012 [TM][I]: clone: Cloning ../../1/14617469c72d931764fc6cba4d3b97c1
in asgaror.local:/var/lib/one//datastores/0/162/disk.0 Sun Sep 30
19:00:33 2012 [TM][I]: ExitCode: 0 Sun Sep 30 19:00:40 2012 [TM][I]:
context: Generating context block device at
asgaror.local:/var/lib/one//datastores/0/162/disk.1 Sun Sep 30
19:00:40 2012 [TM][I]: ExitCode: 0 Sun Sep 30 19:00:40 2012 [LCM][I]:
New VM state is BOOT Sun Sep 30 19:00:40 2012 [VMM][I]: Generating
deployment file: /var/lib/one/162/deployment.0 Sun Sep 30 19:00:41
2012 [VMM][I]: ExitCode: 0 Sun Sep 30 19:00:41 2012 [VMM][I]:
Successfully execute network driver operation: pre. Sun Sep 30
19:00:46 2012 [VMM][I]: Command execution fail: cat 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20120930/6507085e/attachment-0002.htm>