I used Opennebula from Ubuntu repository, so it is version 3.2.1-2. I will try to upgrade it to version 3.4.

SSh transfer is veeeery slooooow (that is why I don't know if it works yet ...), but definitely the issue is NFS in my case.
Which parameters has to be set?

Jan

----- Pôvodná správa -----
Predmet: Re: [one-users] Error when instantiating VM from image - next status
Dátum: Fri, 22 Jun 2012 11:58:02 +0200
Od: Jaime Melis <jme...@opennebula.org>
Pre: <jan.bena...@atos.net>
Kópia: Javier Fontan <jfon...@opennebula.org>, <users@lists.opennebula.org>


Hello Jan,

I forgot to mention that it's not enough with using the SSH transfer driver, you also have to unmount all your NFS exports in your hypervisor node, so the disk images aren't copied to an NFS filesystem.

Cheers,
Jaime

On Fri, Jun 22, 2012 at 11:21 AM, Jaime Melis <jme...@opennebula.org> wrote:
Hello Jan,

let's try without NFS just to rule it out. Can you use the SSH transfer driver:
and try launching the VM again?

By the way, after reading your logs it seems you're not using the last stable release OpenNebula 3.4. Could you upgrade to this release?

Regards,
Jaime


On Fri, Jun 22, 2012 at 8:01 AM, Jan Benadik <jan.bena...@atos.net> wrote:
Yes, it runs:
oneadmin@nebula-3:~$ ps aux |grep oned
oneadmin 10158  0.0  0.1 1172252 8020 ?        Sl   Jun21   0:22 /usr/bin/oned -f


When I changed security_driver in qemu.conf to default state
/etc/libvirt/qemu.conf:
# security_driver = "selinux"
my error message went back to previous state (but still was there) ...

When I replaced OS on host to Ubuntu 10.04 Server (with the same settings), error message is:

Thu Jun 21 16:41:17 2012 [LCM][I]: New VM state is BOOT
Thu Jun 21 16:41:17 2012 [VMM][I]: Generating deployment file: /var/lib/one/1/deployment.4
Thu Jun 21 16:41:17 2012 [VMM][I]: ExitCode: 0
Thu Jun 21 16:41:17 2012 [VMM][I]: Successfully execute network driver operation: pre.
Thu Jun 21 16:41:48 2012 [VMM][I]: Command execution fail: cat << EOT | /var/tmp/one/vmm/kvm/deploy /var/lib/one/1/images/deployment.4 tyan 1 tyan
Thu Jun 21 16:41:48 2012 [VMM][I]: error: Failed to create domain from /var/lib/one/1/images/deployment.4
Thu Jun 21 16:41:48 2012 [VMM][I]: error: cannot set ownership on /var/lib/one/1/images/disk.1: Permission denied
Thu Jun 21 16:41:48 2012 [VMM][E]: Could not create domain from /var/lib/one/1/images/deployment.4
Thu Jun 21 16:41:48 2012 [VMM][I]: ExitCode: 255
Thu Jun 21 16:41:48 2012 [VMM][I]: Failed to execute virtualization driver operation: deploy.
Thu Jun 21 16:41:48 2012 [VMM][E]: Error deploying virtual machine: Could not create domain from /var/lib/one/1/images/deployment.4
Thu Jun 21 16:41:49 2012 [DiM][I]: New VM state is FAILED


Messages in /var/log/syslog at the same time:
Jun 22 10:17:01 tyan-host CRON[12881]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Jun 22 10:22:04 tyan-host kernel: [59025.594722] type=1505 audit(1340353324.455:27):  operation="profile_load" pid=13044 name="libvirt-f42d2d5f-e5a0-3bcd-a445-1d3d876451e1"
Jun 22 10:22:04 tyan-host libvirtd: 10:22:04.470: error : qemuDomainSetFileOwnership:2222 : cannot set ownership on /var/lib/one/1/images/disk.0: Permission denied
Jun 22 10:22:34 tyan-host libvirtd: 10:22:34.481: error : qemuMonitorOpenUnix:268 : monitor socket did not show up.: No such file or directory
Jun 22 10:22:34 tyan-host libvirtd: 10:22:34.481: error : qemuConnectMonitor:822 : Failed to connect monitor for one-1#012
Jun 22 10:22:34 tyan-host libvirtd: 10:22:34.665: error : qemuDomainSetFileOwnership:2222 : cannot set ownership on /var/lib/one/1/images/disk.1: Permission denied
Jun 22 10:22:34 tyan-host libvirtd: 10:22:34.665: warning : qemudShutdownVMDaemon:2703 : Failed to restore all device ownership for one-1
Jun 22 10:22:34 tyan-host kernel: [59055.797448] type=1505 audit(1340353354.655:28):  operation="profile_remove" pid=13051 name="libvirt-f42d2d5f-e5a0-3bcd-a445-1d3d876451e1" namespace="root"


Jan



Dňa 21.06.2012 17:19, Javier Fontan  wrote / napísal(a):
Also, I supposte oned is running as oneadmin user. Just to check.

On Thu, Jun 21, 2012 at 5:19 PM, Javier Fontan <jfon...@opennebula.org> wrote:
I am checking my configuration ans the only differences are:

* oneadmin is in group oneadmin
* qemu group is oneadmin
* �/var/lib/one/** lrwk, line is in /etc/apparmor.d/local/usr.sbin.libvirtd

Can you try moving the line of apparmor to
/etc/apparmor.d/local/usr.sbin.libvirtd? Maybe there's a precedence
problem that we don't know of. Unfortunately I am not an apparmor.

On Thu, Jun 21, 2012 at 9:55 AM, Jan Benadik <jan.bena...@atos.net> wrote:
So - now I have still the same error message in oned.log:
Thu Jun 21 09:26:42 2012 [LCM][I]: New VM state is BOOT
Thu Jun 21 09:26:42 2012 [VMM][I]: Generating deployment file:
/var/lib/one/0/deployment.38
Thu Jun 21 09:26:42 2012 [VMM][I]: ExitCode: 0
Thu Jun 21 09:26:42 2012 [VMM][I]: Successfully execute network driver
operation: pre.
Thu Jun 21 09:26:44 2012 [VMM][I]: Command execution fail: cat << EOT |
/var/tmp/one/vmm/kvm/deploy /var/lib/one/0/images/deployment.38 myto 0 myto
Thu Jun 21 09:26:44 2012 [VMM][I]: error: Failed to create domain from
/var/lib/one/0/images/deployment.38
Thu Jun 21 09:26:44 2012 [VMM][I]: error: Unable to read from monitor:
Connection reset by peer
Thu Jun 21 09:26:44 2012 [VMM][E]: Could not create domain from
/var/lib/one/0/images/deployment.38
Thu Jun 21 09:26:44 2012 [VMM][I]: ExitCode: 255
Thu Jun 21 09:26:44 2012 [VMM][I]: Failed to execute virtualization driver
operation: deploy.
Thu Jun 21 09:26:44 2012 [VMM][E]: Error deploying virtual machine: Could
not create domain from /var/lib/one/0/images/deployment.38
Thu Jun 21 09:26:45 2012 [DiM][I]: New VM state is FAILED

At the same time in the /var/log/libvirt/libvirtd.log the following message
appears:
2012-06-21 09:27:43.610+0000: 1114: warning :
virDomainDiskDefForeachPath:13244 : Ignoring open failure on
/var/lib/one/0/images/disk.1: Permission denied
2012-06-21 09:27:44.296+0000: 1110: error : qemuMonitorIORead:513 : Unable
to read from monitor: Connection reset by peer

Nothing in /var/log/syslog (doesn't matter if apparmor is running, stopped,
flushed ...!).

Permissions of files and folders:
oneadmin@opennebula-host:/var/lib$ ls -ld /var/lib/one
drwxr-xr-x 10 oneadmin root 4096 Jun 21 09:49 /var/lib/one

oneadmin@opennebula-host:/var/lib/one# ls -la
total 132
drwxr-xr-x� 8 oneadmin root�� 4096 Jun 21 09:27 .
drwxr-xr-x 37 root���� root�� 4096 Jun 21 06:30 ..
-rw-------� 1 oneadmin cloud� 2261 Jun 21 08:42 .bash_history
drwx------� 2 oneadmin cloud� 4096 Jun 20 09:48 .cache
drwx------� 2 oneadmin cloud� 4096 Jun 20 09:49 .one
drwx------� 2 oneadmin root�� 4096 Jun 20 17:43 .ssh
-rw-------� 1 oneadmin cloud� 3412 Jun 20 11:06 .viminfo
drwxrwxrwx� 3 oneadmin cloud� 4096 Jun 21 09:26 0
-rw-r--r--� 1 oneadmin cloud� 1738 Jun 21 08:50 config
drwxrwx--T� 2 oneadmin root�� 4096 Jun 20 10:57 images
-rw-r--r--� 1 oneadmin cloud 67584 Jun 21 09:27 one.db
-rw-r--r--� 1 oneadmin cloud 16384 Jun 20 16:28 oneacct.db
drwxr-xr-x� 8 root���� root�� 4096 Jun 20 09:33 remotes

oneadmin@opennebula-host:/var/lib/one/0# ls -la
total 20
drwxrwxrwx� 3 oneadmin cloud 4096 Jun 21 09:36 .
drwxr-xr-x 10 oneadmin root� 4096 Jun 21 09:35 ..
-rw-r--r--� 1 oneadmin cloud� 735 Jun 21 09:26 deployment.38
drwxrwxrwx� 2 oneadmin cloud 4096 Jun 21 09:26 images
-rw-r--r--� 1 oneadmin cloud� 201 Jun 21 09:26 transfer.38.prolog

oneadmin@opennebula-host:/var/lib/one/0/images# ls -la
total 906256
drwxrwxrwx 2 oneadmin cloud����� 4096 Jun 21 09:26 .
drwxrwxrwx 3 oneadmin cloud����� 4096 Jun 21 09:36 ..
-rw-r--r-- 1 oneadmin cloud������ 736 Jun 21 09:26 deployment.38
-rw-rw-rw- 1 oneadmin cloud 927989760 Jun 21 09:26 disk.0
lrwxrwxrwx 1 oneadmin cloud������� 52 Jun 21 09:26 disk.1 ->
/var/lib/one/images/9c52b90a79dba7c26a912d05ff5190b8

oneadmin@opennebula-host:~/images$ ls -la
total 1040116
drwxrwx--T� 2 oneadmin root������� 4096 Jun 20 10:57 .
drwxr-xr-x 10 oneadmin root������� 4096 Jun 21 09:37 ..
-rw-rw----� 1 oneadmin root�� 927989760 Jun 20 10:57
46440b43448202b4ee69b4b541f5eeab
-rw-rw----� 1 oneadmin root 10737418241 Jun 20 10:57
9c52b90a79dba7c26a912d05ff5190b8


Libvirtd and Qemu settings:
/etc/libvirt/libvirtd.conf:
listen_tls = 0
listen_tcp = 1
unix_sock_group = "libvirtd"
unix_sock_ro_perms = "0777"
unix_sock_rw_perms = "0777"
unix_sock_dir = "/var/run/libvirt"
auth_unix_ro = "none"
auth_unix_rw = "none"

/etc/libvirt/qemu.conf:
security_driver = "none"
user = "oneadmin"
group = "cloud"
dynamic_ownership = 0

/etc/default/libvirt-bin:
start_libvirtd="yes"
libvirtd_opts="-d -l"

/etc/apparmor.d/usr.sbin.libvirtd:
# Last Modified: Mon Jul� 6 17:23:58 2009
#include <tunables/global>
@{LIBVIRT}="libvirt"

/usr/sbin/libvirtd {
� #include <abstractions/base>
� # Site-specific additions and overrides. See local/README for details.
� #include <local/usr.sbin.libvirtd>

� capability kill,
� capability net_admin,
� capability net_raw,
� capability setgid,
� capability sys_admin,
� capability sys_module,
� capability sys_ptrace,
� capability sys_nice,
� capability sys_chroot,
� capability setuid,
� capability dac_override,
� capability dac_read_search,
� capability fowner,
� capability chown,
� capability setpcap,
� capability mknod,
� capability fsetid,
� capability ipc_lock,

� network inet stream,
� network inet dgram,
� network inet6 stream,
� network inet6 dgram,
� network packet dgram,

� # for now, use a very lenient profile since we want to first focus on
� # confining the guests
� / r,
� /** rwmkl,

� /bin/* PUx,
� /sbin/* PUx,
� /usr/bin/* PUx,
� /usr/sbin/* PUx,
� /lib/udev/scsi_id PUx,

� # Required by nwfilter_ebiptables_driver.c:ebiptablesWriteToTempFile() to
� # write and run an ebtables script.
� /var/lib/libvirt/virtd* ixr,

� # force the use of virt-aa-helper
� audit deny /sbin/apparmor_parser rwxl,
� audit deny /etc/apparmor.d/libvirt/** wxl,
� audit deny /sys/kernel/security/apparmor/features rwxl,
� audit deny /sys/kernel/security/apparmor/matching rwxl,
� audit deny /sys/kernel/security/apparmor/.* rwxl,
� /sys/kernel/security/apparmor/profiles r,
� /usr/lib/libvirt/* PUxr,
� /etc/libvirt/hooks/** rmix,
� /var/lib/one/** lrwk,

� # allow changing to our UUID-based named profiles
� change_profile ->
@{LIBVIRT}-[0-9a-f]*-[0-9a-f]*-[0-9a-f]*-[0-9a-f]*-[0-9a-f]*,

}

User settings:
oneadmin@opennebula-host:~/images$ groups oneadmin
oneadmin : cloud root disk kvm libvirtd



My question - where is an issue?

Jan


_______________________________________________
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

--
Javier Font�n Mui�os
Project Engineer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org�|�jfon...@opennebula.org�| @OpenNebula

--

Ján Beňadik
Managed Services - Solution Design Architect
+421 46 5151 332
+421 903 691 634
jan.bena...@atos.net
Vinohradnícka 6, 971 01 Prievidza
www.sk.atos.net
__________________________________



_______________________________________________
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org




--
Jaime Melis
Project Engineer
OpenNebula - The Open Source Toolkit for Cloud Computing
www.OpenNebula.org | jme...@opennebula.org



--
Jaime Melis
Project Engineer
OpenNebula - The Open Source Toolkit for Cloud Computing
www.OpenNebula.org | jme...@opennebula.org

--

Ján Beňadik
Managed Services - Solution Design Architect
+421 46 5151 332
+421 903 691 634
jan.bena...@atos.net
Vinohradnícka 6, 971 01 Prievidza
www.sk.atos.net
__________________________________


_______________________________________________
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Reply via email to