Just for clarity on that bug report; are you suggesting that the guest must go into a paused state to take a snapshot and that the admin shouldn't be aware of this state?

Although not horribly concerning for most tasks, wouldn't this lead to a 10 second clock skew every time there is a snapshot? Or does the guest sync from host hw clock on resume?

Thanks,

Steve Dainard 
IT Infrastructure Manager
Miovision | Rethink Traffic
519-513-2407 ex.250
877-646-8476 (toll-free)

Blog  |  LinkedIn  |  Twitter  |  Facebook

Miovision Technologies Inc. | 148 Manitou Drive, Suite 101, Kitchener, ON, Canada | N2C 1L3
This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify us immediately.


On Tue, Feb 4, 2014 at 3:44 AM, Dafna Ron <dron@redhat.com> wrote:
On 02/04/2014 08:20 AM, Elad Ben Aharon wrote:
 From what I saw in the thread, libvirt pauses the VM, which effects the continuity of its operation.
I checked it also in one of the latest builds of 3.3 and I observed the same behaviour:
2014-02-04 10:18:29,441 INFO  [org.ovirt.engine.core.vdsbroker.VdsUpdateRunTimeInfo] (DefaultQuartzScheduler_Worker-22) [5cbc7ab5] VM nfs2-1 86728f5c-3583-420f-9536-bfabaf11b235 moved from Up --> Paused

Dafna, I saw you've already opened a bug on that:
https://bugzilla.redhat.com/show_bug.cgi?id=1057587

just proves I'm getting senile :)


----- Original Message -----
From: "Dafna Ron" <dron@redhat.com>
To: "Maor Lipchuk" <mlipchuk@redhat.com>
Cc: "Steve Dainard" <sdainard@miovision.com>, "Elad Ben Aharon" <ebenahar@redhat.com>, "Karli Sjöberg" <Karli.Sjoberg@slu.se>, users@ovirt.org
Sent: Tuesday, February 4, 2014 12:15:42 AM
Subject: Re: [Users] oVirt 3.3.3 RC EL6 Live Snapshot

Maor I am not saying that we are not doing a live snapshot :) I am
saying that we need a print in the log that states live snapshot command
was called i.e: Print in the log: LiveSnapshotCommand -> this can call
to the rest of snapshotVDSCreateCommand.


On 02/03/2014 07:38 PM, Maor Lipchuk wrote:
On 02/03/2014 07:46 PM, Dafna Ron wrote:
On 02/03/2014 05:34 PM, Maor Lipchuk wrote:
On 02/03/2014 07:18 PM, Dafna Ron wrote:
Maor,

If snapshotVDSCommand is for live snapshot, what is the offline create
snapshot command?
It is the CreateSnapshotVdsCommand which calls createVolume in VDSM
but we need to be able to know that a live snapshot was sent and not an
offline snapshot.
Yes, at the logs we can see the all process :

First a request to create a snapshot (new volume) sent to VDSM:
2014-02-02 09:41:09,557 INFO
[org.ovirt.engine.core.vdsbroker.irsbroker.CreateSnapshotVDSCommand]
(pool-6-thread-49) [67ea047a] START, CreateSnapshotVDSCommand(
storagePoolId = fcb89071-6cdb-4972-94d1-c9324cebf814,
ignoreFailoverLimit = false, storageDomainId =
a52938f7-2cf4-4771-acb2-0c78d14999e5, imageGroupId =
c1cb6b66-655e-48c3-8568-4975295eb037, imageSizeInBytes = 21474836480,
volumeFormat = COW, newImageId = 6d8c80a4-328f-4a53-86a2-a4080a2662ce,
newImageDescription = , imageId = 5085422e-6592-415a-9da3-9e43dac9374b,
sourceImageGroupId = c1cb6b66-655e-48c3-8568-4975295eb037), log id: 7875f3f5

after the snapshot gets created :
2014-02-02 09:41:20,553 INFO
[org.ovirt.engine.core.bll.CreateAllSnapshotsFromVmCommand]
(pool-6-thread-49) Ending command successfully:
org.ovirt.engine.core.bll.CreateAllSnapshotsFromVmCommand

then the engine calls the live snapshot (see also [1])
2014-02-02 09:41:30,234 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.SnapshotVDSCommand]
(pool-6-thread-49) FINISH, SnapshotVDSCommand, log id: 7e0d7872

Elad, somewhere in this flow we need to know that the snapshot was taken
on a running vm :) this seems like a bug to me.
we did not say that live snapshot did not succeed :)  we said that the
vm is paused and restarted - which is something that should not happen
for live snapshot (or at least never did before).
It's not sure that the restart is related to the live snapshot. but that
should be observed in the libvirt/vdsm logs.
yes, I am sure because the user is reporting it and the logs show it...
as I wrote before, we know that vdsm is reporting the vm as paused, that
is because libvirt is reporting the vm as paused and I think that its
happening because libvirt is not doing a live snapshot and so pauses the
vm while taking the snapshot.
That sounds logic to me, it's need to be checked with libvirt, if that
kind of behaviour could happen.
Elad, can you please try to reproduce and open a bug to libvirt?

Dafna


On 02/03/2014 05:08 PM, Maor Lipchuk wrote:
    From the engine logs it seems that indeed live snapshot is called
(The
command is snapshotVDSCommand see [1]).
This is done right after the snapshot has been created in the VM and it
signals the qemu process to start using the new volume created.

When live snapshot does not succeed we should see in the log something
like "Wasn't able to live snapshot due to error:...", but it does not
appear so it seems that this worked out fine.

At some point I can see in the logs that VDSM reports to the engine
that
the VM is paused.


[1]
2014-02-02 09:41:20,564 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.SnapshotVDSCommand]
(pool-6-thread-49) START, SnapshotVDSCommand(HostName = ovirt002,
HostId
= 3080fb61-2d03-4008-b47f-9b66276a4257,
vmId=e261e707-a21f-4ae8-9cff-f535f4430446), log id: 7e0d7872
2014-02-02 09:41:21,119 INFO
[org.ovirt.engine.core.vdsbroker.VdsUpdateRunTimeInfo]
(DefaultQuartzScheduler_Worker-93) VM snapshot-test
e261e707-a21f-4ae8-9cff-f535f4430446 moved from Up --> Paused
2014-02-02 09:41:30,234 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.SnapshotVDSCommand]
(pool-6-thread-49) FINISH, SnapshotVDSCommand, log id: 7e0d7872
2014-02-02 09:41:30,238 INFO
[org.ovirt.engine.core.bll.CreateSnapshotCommand] (pool-6-thread-49)
[67ea047a] Ending command successfully:
org.ovirt.engine.core.bll.CreateSnapshotCommand
...

Regards,
Maor

On 02/03/2014 06:24 PM, Dafna Ron wrote:
Thanks Steve.

from the logs I can see that the create snapshot succeeds and that the
vm is resumed.
the vm moves to pause as part of libvirt flows:

2014-02-02 14:41:20.872+0000: 5843: debug :
qemuProcessHandleStop:728 :
Transitioned guest snapshot-test to paused state
2014-02-02 14:41:30.031+0000: 5843: debug :
qemuProcessHandleResume:776
: Transitioned guest snapshot-test out of paused into resumed state

There are bugs here but I am not sure yet if this is libvirt
regression
or engine.

I'm adding Elad and Maor since in engine logs I can't see anything
calling for live snapshot (only for snapshot) - Maor, shouldn't live
snapshot command be logged somewhere in the logs?
Is it possible that engine is calling to create snapshot and not
create
live snapshot which is why the vm pauses?

Elad, if engine is not logging live snapshot anywhere I would open
a bug
for engine (to print that in the logs).
Also, there is a bug in vdsm log for sdc where the below is logged as
ERROR and not INFO:

Thread-23::ERROR::2014-02-02
09:51:19,497::sdc::137::Storage.StorageDomainCache::(_findDomain)
looking for unfetched domain a52938f7-2cf4-4771-acb2-0c78d14999e5
Thread-23::ERROR::2014-02-02
09:51:19,497::sdc::154::Storage.StorageDomainCache::(_findUnfetchedDomain)


looking for domain a52938f7-2cf4-4771-acb2-0c78d14999e5

If the engine was sending live snapshot or if there is no
difference in
the two commands in engine side than I would open a bug for libvirt
for
pausing the vm during live snapshot.

Dafna

On 02/03/2014 02:41 PM, Steve Dainard wrote:
[root@ovirt002 ~]# vdsClient -s 0 getStorageDomainInfo
a52938f7-2cf4-4771-acb2-0c78d14999e5
uuid = a52938f7-2cf4-4771-acb2-0c78d14999e5
pool = ['fcb89071-6cdb-4972-94d1-c9324cebf814']
lver = 5
version = 3
role = Master
remotePath = gluster-store-vip:/rep1
spm_id = 2
type = NFS
class = Data
master_ver = 1
name = gluster-store-rep1


*Steve Dainard *
IT Infrastructure Manager
Miovision <http://miovision.com/> | /Rethink Traffic/
519-513-2407 ex.250
877-646-8476 (toll-free)

*Blog <http://miovision.com/blog> | **LinkedIn
<https://www.linkedin.com/company/miovision-technologies>  | Twitter
<https://twitter.com/miovision>  | Facebook
<https://www.facebook.com/miovision>*
------------------------------------------------------------------------


Miovision Technologies Inc. | 148 Manitou Drive, Suite 101,
Kitchener,
ON, Canada | N2C 1L3
This e-mail may contain information that is privileged or
confidential. If you are not the intended recipient, please delete
the
e-mail and any attachments and notify us immediately.


On Sun, Feb 2, 2014 at 2:55 PM, Dafna Ron <dron@redhat.com
<mailto:dron@redhat.com>> wrote:

        please run vdsClient -s 0 getStorageDomainInfo
        a52938f7-2cf4-4771-acb2-0c78d14999e5

        Thanks,

        Dafna



        On 02/02/2014 03:02 PM, Steve Dainard wrote:

            Logs attached with VM running on qemu-kvm-rhev packages
installed.

            *Steve Dainard *
            IT Infrastructure Manager
            Miovision <http://miovision.com/> | /Rethink Traffic/
            519-513-2407 <tel:519-513-2407> ex.250

            877-646-8476 <tel:877-646-8476> (toll-free)

            *Blog <http://miovision.com/blog> | **LinkedIn
           <https://www.linkedin.com/company/miovision-technologies>  |
            Twitter <https://twitter.com/miovision>  | Facebook
            <https://www.facebook.com/miovision>*
         ------------------------------------------------------------------------


            Miovision Technologies Inc. | 148 Manitou Drive, Suite 101,
            Kitchener, ON, Canada | N2C 1L3
            This e-mail may contain information that is privileged or
            confidential. If you are not the intended recipient, please
            delete the e-mail and any attachments and notify us
immediately.


            On Sun, Feb 2, 2014 at 5:05 AM, Dafna Ron <dron@redhat.com
            <mailto:dron@redhat.com> <mailto:dron@redhat.com
            <mailto:dron@redhat.com>>> wrote:

                can you please upload full engine, vdsm, libvirt and
vm's
            qemu logs?


                On 02/02/2014 02:08 AM, Steve Dainard wrote:

                    I have two CentOS 6.5 Ovirt hosts (ovirt001,
ovirt002)

                    I've installed the applicable qemu-kvm-rhev
packages
            from this
                    site:
http://www.dreyou.org/ovirt/vdsm32/Packages/ on
            ovirt002.

                    On ovirt001 if I take a live snapshot:

                    Snapshot 'test qemu-kvm' creation for VM
            'snapshot-test' was
                    initiated by admin@internal.
                    The VM is paused
                    Failed to create live snapshot 'test qemu-kvm'
for VM
                    'snapshot-test'. VM restart is recommended.
                    Failed to complete snapshot 'test qemu-kvm'
creation
            for VM
                    'snapshot-test'.
                    The VM is then started, and the status for the
snapshot
                    changes to OK.

                    On ovirt002 (with the packages from dreyou) I don't
            get any
                    messages about a snapshot failing, but my VM is
still
            paused
                    to complete the snapshot. Is there something else
            other than
                    the qemu-kvm-rhev packages that would enable this
            functionality?

                    I've looked for some information on when the
packages
            would be
                    built as required in the CentOS repos, but I
don't see
                    anything definitive.

          http://lists.ovirt.org/pipermail/users/2013-December/019126.html
                    Looks like one of the maintainers is waiting for
            someone to
                    tell him what flags need to be set.

                    Also, another thread here:
           http://comments.gmane.org/gmane.comp.emulators.ovirt.arch/1618
                    same maintainer, mentioning that he hasn't seen
            anything in
                    the bug tracker.

                    There is a bug here:
            https://bugzilla.redhat.com/show_bug.cgi?id=1009100 that
seems
                    to have ended in finding a way for qemu to expose
            whether it
                    supports live snapshots, rather than figuring
out how
            to get
                    the CentOS team the info they need to build the
            packages with
                    the proper flags set.

                    I have bcc'd both dreyou (packaged the
qemu-kvm-rhev
            packages
                    listed above) and Russ (CentOS maintainer
mentioned in
the
                    other threads) if they wish to chime in and perhaps
                    collaborate on which flags, if any, should be set
for the
                    qemu-kvm builds so we can get a CentOS bug report
            going and
                    hammer this out.

                    Thanks everyone.

                    **crosses fingers and hopes for live snapshots
soon**



                    *Steve Dainard *
                    IT Infrastructure Manager
                    Miovision <http://miovision.com/> | /Rethink
Traffic/
            519-513-2407 <tel:519-513-2407> <tel:519-513-2407
            <tel:519-513-2407>> <tel:519-513-2407 <tel:519-513-2407>
                    <tel:519-513-2407 <tel:519-513-2407>>> ex.250
            877-646-8476 <tel:877-646-8476> <tel:877-646-8476
            <tel:877-646-8476>> <tel:877-646-8476 <tel:877-646-8476>

                    <tel:877-646-8476 <tel:877-646-8476>>> (toll-free)

                    *Blog <http://miovision.com/blog> | **LinkedIn
                        <https://www.linkedin.com/company/miovision-technologies>  |
                    Twitter <https://twitter.com/miovision>  | Facebook
                    <https://www.facebook.com/miovision>*
                        ------------------------------------------------------------------------




                    Miovision Technologies Inc. | 148 Manitou Drive,
Suite
            101,
                    Kitchener, ON, Canada | N2C 1L3
                    This e-mail may contain information that is
privileged or
                    confidential. If you are not the intended
recipient,
            please
                    delete the e-mail and any attachments and notify us
            immediately.


                    On Fri, Jan 31, 2014 at 1:26 PM, Steve Dainard
                    <sdainard@miovision.com
            <mailto:sdainard@miovision.com>
<mailto:sdainard@miovision.com
            <mailto:sdainard@miovision.com>>
                    <mailto:sdainard@miovision.com
            <mailto:sdainard@miovision.com>

                    <mailto:sdainard@miovision.com
            <mailto:sdainard@miovision.com>>>> wrote:


                            How would you developers, speaking for the
                    oVirt-community,
                            propose to
                            solve this for CentOS _now_ ?

                            I would imagine that the easiest way is
that
            you build and
                            host this one
                            package(qemu-kvm-rhev), since you´ve
basically
            already
                    have
                            the source
                            and recipe (since you´re already
providing it
            for RHEV
                            anyway). Then,
                            once that´s in place, it´s more a
question of
            where to
                    host the
                            packages, in what repository. Be it your
own,
            or some
                    other
                            repo set up
                            for the SIG.

                            This is my view, how I as a user view this
issue.



                        I think this is a pretty valid view.

                        What would it take to get the correct qemu
package
            hosted
                    in the
                        ovirt repo?

                            --

                            Med Vänliga Hälsningar

                                       -------------------------------------------------------------------------------



                            Karli Sjöberg
                            Swedish University of Agricultural Sciences
            Box 7079
                    (Visiting
                            Address
                            Kronåsvägen 8)
                            S-750 07 Uppsala, Sweden
                            Phone: +46-(0)18-67 15 66
            <tel:%2B46-%280%2918-67%2015%2066>
                    <tel:%2B46-%280%2918-67%2015%2066>
                    <tel:%2B46-%280%2918-67%2015%2066>
            karli.sjoberg@slu.se <mailto:karli.sjoberg@slu.se>
            <mailto:karli.sjoberg@slu.se <mailto:karli.sjoberg@slu.se>>
                    <mailto:karli.sjoberg@slu.se
            <mailto:karli.sjoberg@slu.se> <mailto:karli.sjoberg@slu.se
            <mailto:karli.sjoberg@slu.se>>>

                          _______________________________________________
                            Users mailing list
            Users@ovirt.org <mailto:Users@ovirt.org>
            <mailto:Users@ovirt.org <mailto:Users@ovirt.org>>
                    <mailto:Users@ovirt.org <mailto:Users@ovirt.org>
            <mailto:Users@ovirt.org <mailto:Users@ovirt.org>>>

            http://lists.ovirt.org/mailman/listinfo/users






                    _______________________________________________
                    Users mailing list
            Users@ovirt.org <mailto:Users@ovirt.org>
            <mailto:Users@ovirt.org <mailto:Users@ovirt.org>>

            http://lists.ovirt.org/mailman/listinfo/users



                --     Dafna Ron




        --     Dafna Ron





--
Dafna Ron