8:39 a.m.
Hi,
I believe it's save, but just wanted to re-check, is it save to restart
libvirtd on a HV running ~40 VMs ?
Alex
--
| RHCE | Sen Sys Engineer / Platform Architect | www.vcore.co |
www.vsearchcloud.com |
10:12 a.m.
On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
Hi,
I believe it's save, but just wanted to re-check, is it save to restart
libvirtd on a HV running ~40 VMs ?
Hi,
for libvirt questions, I'd rather use libvirt-users(a)redhat.com, but for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
Have a nice day,
Martin
5:21 p.m.
This is a multi-part message in MIME format.
--------------000503080808040106050307
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Thanks Martin,
I tested on a "almost" empty host I had to play with, and it seemed all
went well, so will hopefully be able to regain ~11 GB of memory after
restarting libvirtd :) ..
I'll sign up to the other mailing list to find out why / how it could
have grown this big :\
Thanks
Alex
On 04/15/2013 04:12 PM, Martin Kletzander wrote:
On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
> Hi,
>
> I believe it's save, but just wanted to re-check, is it save to restart
> libvirtd on a HV running ~40 VMs ?
>
Hi,
for libvirt questions, I'd rather use libvirt-users(a)redhat.com, but for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
Have a nice day,
Martin
--------------000503080808040106050307
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font size="-1"><font face="Tahoma">Thanks
Martin,<br>
<br>
I tested on a "almost" empty host I had to play with, and it
seemed all went well, so will hopefully be able to regain ~11 GB
of memory after restarting libvirtd :) .. <br>
<br>
I'll sign up to the other mailing list to find out why / how it
could have grown this big :\ <br>
<br>
Thanks<br>
Alex<br>
<br>
</font></font><br>
On 04/15/2013 04:12 PM, Martin Kletzander wrote:
<blockquote cite="mid:516C18E3.1000908@redhat.com"
type="cite">
<pre wrap="">On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi,
I believe it's save, but just wanted to re-check, is it save to restart
libvirtd on a HV running ~40 VMs ?
</pre>
</blockquote>
<pre wrap="">
Hi,
for libvirt questions, I'd rather use <a class="moz-txt-link-abbreviated"
href="mailto:libvirt-users@redhat.com">libvirt-users@redhat.com</a>,
but for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
Have a nice day,
Martin
</pre>
</blockquote>
</body>
</html>
--------------000503080808040106050307--
3:18 a.m.
On 04/16/2013 12:21 AM, Alex Leonhardt wrote:
Thanks Martin,
I tested on a "almost" empty host I had to play with, and it seemed all
went well, so will hopefully be able to regain ~11 GB of memory after
restarting libvirtd :) ..
I'll sign up to the other mailing list to find out why / how it could
have grown this big :\
There might be some memleaks, especially in older version. For this
kind of stuff, I'd maybe use the libvir-list [1]
Martin
[1] http://www.redhat.com/mailman/listinfo/libvir-list
Thanks
Alex
On 04/15/2013 04:12 PM, Martin Kletzander wrote:
> On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
>> Hi,
>>
>> I believe it's save, but just wanted to re-check, is it save to restart
>> libvirtd on a HV running ~40 VMs ?
>>
> Hi,
>
> for libvirt questions, I'd rather use libvirt-users(a)redhat.com, but for
> this particular one, I can confirm that libvirt is written in a way that
> enables it to be restarted without any impact on the machines which are
> being ran.
>
> Of course I can't say "nothing will happen" due to the fact that every
> single time something can happen, but nothing _should_ happen to any of
> your machines.
>
> Have a nice day,
> Martin
3:47 p.m.
On Mon, Apr 15, 2013 at 05:12:35PM +0200, Martin Kletzander wrote:
On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
> Hi,
>
> I believe it's save, but just wanted to re-check, is it save to restart
> libvirtd on a HV running ~40 VMs ?
>
Hi,
for libvirt questions, I'd rather use libvirt-users(a)redhat.com, but for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
Indeed. Vdsm should notice that libvirtd has died, and restart itself.
However, it would be safer to stop Vdsm explicitly before you do that
(for example http://gerrit.ovirt.org/8283 )
5:02 p.m.
This is a multi-part message in MIME format.
--------------050809010202050605030902
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
would that stop any of the running VMs ?
alex
On 04/16/2013 09:47 PM, Dan Kenigsberg wrote:
On Mon, Apr 15, 2013 at 05:12:35PM +0200, Martin Kletzander wrote:
> On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
>> Hi,
>>
>> I believe it's save, but just wanted to re-check, is it save to restart
>> libvirtd on a HV running ~40 VMs ?
>>
> Hi,
>
> for libvirt questions, I'd rather use libvirt-users(a)redhat.com, but for
> this particular one, I can confirm that libvirt is written in a way that
> enables it to be restarted without any impact on the machines which are
> being ran.
>
> Of course I can't say "nothing will happen" due to the fact that every
> single time something can happen, but nothing _should_ happen to any of
> your machines.
Indeed. Vdsm should notice that libvirtd has died, and restart itself.
However, it would be safer to stop Vdsm explicitly before you do that
(for example http://gerrit.ovirt.org/8283 )
--------------050809010202050605030902
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font size="-1"><font face="Tahoma">would that stop
any of the
running VMs ?<br>
<br>
alex<br>
</font></font><br>
On 04/16/2013 09:47 PM, Dan Kenigsberg wrote:
<blockquote cite="mid:20130416204759.GK5925@redhat.com"
type="cite">
<pre wrap="">On Mon, Apr 15, 2013 at 05:12:35PM +0200, Martin
Kletzander wrote:
</pre>
<blockquote type="cite">
<pre wrap="">On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi,
I believe it's save, but just wanted to re-check, is it save to restart
libvirtd on a HV running ~40 VMs ?
</pre>
</blockquote>
<pre wrap="">
Hi,
for libvirt questions, I'd rather use <a class="moz-txt-link-abbreviated"
href="mailto:libvirt-users@redhat.com">libvirt-users@redhat.com</a>,
but for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
</pre>
</blockquote>
<pre wrap="">
Indeed. Vdsm should notice that libvirtd has died, and restart itself.
However, it would be safer to stop Vdsm explicitly before you do that
(for example <a class="moz-txt-link-freetext"
href="http://gerrit.ovirt.org/8283">http://gerrit.ovirt.org/... )
</pre>
</blockquote>
</body>
</html>
--------------050809010202050605030902--
2:07 a.m.
------=_Part_1602476_833045475.1366182476627
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
----- Original Message -----
From: "Alex Leonhardt" <alex.tuxx(a)gmail.com>
To: "Dan Kenigsberg" <danken(a)redhat.com>
Cc: "oVirt Mailing List" <users(a)ovirt.org>
Sent: Wednesday, April 17, 2013 1:02:25 AM
Subject: Re: [Users] save to restart libvirtd ?
would that stop any of the running VMs ?
it shouldn't, but if you have ovirt engine running, make sure that power management
not active for this host,
since after timeout (usually 1 minute) when vdsm is down, the host will be fenced
(restarted) - then the vms will surely stop..
look at http://www.ovirt.org/Automatic_Fencing for more info about automatic fencing in
engine.
alex
On 04/16/2013 09:47 PM, Dan Kenigsberg wrote:
> On Mon, Apr 15, 2013 at 05:12:35PM +0200, Martin Kletzander wrote:
> > On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
>
> > > Hi,
> >
>
> > > I believe it's save, but just wanted to re-check,
is it save to restart
> >
>
> > > libvirtd on a HV running ~40 VMs ?
> >
>
> > Hi,
>
> > for libvirt questions, I'd rather use
libvirt-users(a)redhat.com , but for
>
> > this particular one, I can confirm that libvirt is written in a way that
>
> > enables it to be restarted without any impact on the machines which are
>
> > being ran.
>
> > Of course I can't say "nothing will happen"
due to the fact that every
>
> > single time something can happen, but nothing _should_ happen to any of
>
> > your machines.
>
> Indeed. Vdsm should notice that libvirtd has died, and restart
itself.
> However, it would be safer to stop Vdsm explicitly before you do that
> (for example http://gerrit.ovirt.org/8283 )
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
------=_Part_1602476_833045475.1366182476627
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable
<html><body><div style=3D"font-family: times new roman, new york,
times, se=
rif; font-size: 12pt; color:
#000000"><div><br></div><div><br></div><hr
id=
=3D"zwchr"><blockquote style=3D"border-left:2px solid
#1010FF;margin-left:5=
px;padding-left:5px;color:#000;font-weight:normal;font-style:normal;text-de=
coration:none;font-family:Helvetica,Arial,sans-serif;font-size:12pt;" data-=
mce-style=3D"border-left: 2px solid #1010FF; margin-left: 5px; padding-left=
: 5px; color: #000; font-weight: normal; font-style: normal; text-decoratio=
n: none; font-family: Helvetica,Arial,sans-serif; font-size:
12pt;"><b>From=
: </b>"Alex Leonhardt"
&lt;alex.tuxx(a)gmail.com&gt;<br><b>To: </b>"Dan Kenig=
sberg" &lt;danken(a)redhat.com&gt;<br><b>Cc: </b>"oVirt
Mailing List" <use=
rs(a)ovirt.org&gt;<br><b>Sent: </b>Wednesday, April 17, 2013 1:02:25
AM<br><b=
Subject: </b>Re: [Users] save to restart libvirtd
?<br><div><br></div><spa=
n
size=3D"-1"><span style=3D"font-family: Tahoma;"
data-mce-style=3D"font-f=
amily: Tahoma;" face=3D"Tahoma">would that stop any of the running VMs
?<br=
<br> </span></span></blockquote><div>it
shouldn't, but if you have ovirt =
engine running, make sure that power
management not active for this host,<b=
r></div><div>since after timeout (usually 1 minute) when vdsm is down, the
=
host will be fenced (restarted) - then the vms will surely stop..<br></div>=
<div>look at <a href=3D"http://www.ovirt.org/Automatic_Fencing
">http://www=
.ovirt.org/Automatic_Fencing </a>for more info about automatic fencing in e=
ngine.<br></div><blockquote style=3D"border-left:2px solid
#1010FF;margin-l=
eft:5px;padding-left:5px;color:#000;font-weight:normal;font-style:normal;te=
xt-decoration:none;font-family:Helvetica,Arial,sans-serif;font-size:12pt;" =
data-mce-style=3D"border-left: 2px solid #1010FF; margin-left: 5px; padding=
-left: 5px; color: #000; font-weight: normal; font-style: normal; text-deco=
ration: none; font-family: Helvetica,Arial,sans-serif; font-size: 12pt;"><s=
pan size=3D"-1"><span style=3D"font-family: Tahoma;"
data-mce-style=3D"font=
-family: Tahoma;" face=3D"Tahoma">alex<br>
</span></span><br>On 04/16/2013 =
09:47 PM, Dan Kenigsberg wrote:<blockquote cite=3D"mid:20130416204759.GK592=
5(a)redhat.com"><pre>On Mon, Apr 15, 2013 at 05:12:35PM +0200, Martin
Kletzan=
der wrote:
</pre><blockquote><pre>On 04/15/2013 03:39 PM, Alex Leonhardt wrote:
</pre><blockquote><pre>Hi,
I believe it's save, but just wanted to re-check, is it save to restart
libvirtd on a HV running ~40 VMs ?
</pre></blockquote><pre>Hi,
for libvirt questions, I'd rather use <a
class=3D"moz-txt-link-abbreviated"=
href=3D"mailto:libvirt-users@redhat.com" target=3D"_blank"
data-mce-href=
=3D"mailto:libvirt-users@redhat.com">libvirt-users@redhat.com</a>, but
for
this particular one, I can confirm that libvirt is written in a way that
enables it to be restarted without any impact on the machines which are
being ran.
Of course I can't say "nothing will happen" due to the fact that every
single time something can happen, but nothing _should_ happen to any of
your machines.
</pre></blockquote><pre>Indeed. Vdsm should notice that libvirtd has
died, =
and restart itself.
However, it would be safer to stop Vdsm explicitly before you do that
(for example <a class=3D"moz-txt-link-freetext"
href=3D"http://gerrit.ovirt=
.org/8283" target=3D"_blank"
data-mce-href=3D"http://gerrit.ovirt.org/8283"=
</pre></blockquote><br>_______________________________________________<br>U=
sers mailing list<br>Users@ovirt.org<br>http://lists.ovirt.org/mailman/list=
info/users<br></blockquote><div><br></div></div></body></html>
------=_Part_1602476_833045475.1366182476627--
1:40 p.m.
This is a multi-part message in MIME format.
--------------090101080002020800090303
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
ah, i dont have that configured anyway as I thought from the beginning
that i'd like to keep control over when a host is being rebooted ...
unfortunately, the VMs are running storage domains too (NFS) so it'd
take those down I believe ... i'll try w/o stopping vdsm first ... on
one that hopefully doesnt run too many critical apps at the time ..
i'll report back how it went if anyone's interested :)
On 04/17/2013 08:07 AM, Omer Frenkel wrote:
------------------------------------------------------------------------
*From: *"Alex Leonhardt" <alex.tuxx(a)gmail.com>
*To: *"Dan Kenigsberg" <danken(a)redhat.com>
*Cc: *"oVirt Mailing List" <users(a)ovirt.org>
*Sent: *Wednesday, April 17, 2013 1:02:25 AM
*Subject: *Re: [Users] save to restart libvirtd ?
would that stop any of the running VMs ?
it shouldn't, but if you have ovirt engine running, make sure that
power management not active for this host,
since after timeout (usually 1 minute) when vdsm is down, the host
will be fenced (restarted) - then the vms will surely stop..
look at http://www.ovirt.org/Automatic_Fencing
<http://www.ovirt.org/Automatic_Fencing%20>for more info about
automatic fencing in engine.
--------------090101080002020800090303
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font size="-1"><font face="Tahoma">ah, i dont have
that configured
anyway as I thought from the beginning that i'd like to keep
control over when a host is being rebooted ... <br>
<br>
unfortunately, the VMs are running storage domains too (NFS) so
it'd take those down I believe ... i'll try w/o stopping vdsm
first ... on one that hopefully doesnt run too many critical
apps at the time .. <br>
<br>
i'll report back how it went if anyone's interested :) <br>
<br>
<br>
</font></font><br>
On 04/17/2013 08:07 AM, Omer Frenkel wrote:
<blockquote
cite="mid:1549403532.1602477.1366182476628.JavaMail.root@redhat.com"
type="cite">
<div><br>
</div>
<div><br>
</div>
<hr id="zwchr">
<blockquote style="border-left:2px solid
#1010FF;margin-left:5px;padding-left:5px;color:#000;font-weight:normal;font-style:normal;text-decoration:none;font-family:Helvetica,Arial,sans-serif;font-size:12pt;"
data-mce-style="border-left: 2px solid #1010FF; margin-left:
5px; padding-left: 5px; color: #000; font-weight: normal;
font-style: normal; text-decoration: none; font-family:
Helvetica,Arial,sans-serif; font-size: 12pt;"><b>From:
</b>"Alex
Leonhardt" <a class="moz-txt-link-rfc2396E"
href="mailto:alex.tuxx@gmail.com"><alex.tuxx@gmail.com></a><br>
<b>To: </b>"Dan Kenigsberg" <a
class="moz-txt-link-rfc2396E"
href="mailto:danken@redhat.com"><danken@redhat.com></a><br>
<b>Cc: </b>"oVirt Mailing List" <a
class="moz-txt-link-rfc2396E"
href="mailto:users@ovirt.org"><users@ovirt.org></a><br>
<b>Sent: </b>Wednesday, April 17, 2013 1:02:25 AM<br>
<b>Subject: </b>Re: [Users] save to restart libvirtd ?<br>
<div><br>
</div>
<span size="-1"><span style="font-family: Tahoma;"
data-mce-style="font-family: Tahoma;"
face="Tahoma">would
that stop any of the running VMs ?<br>
<br>
</span></span></blockquote>
<div>it shouldn't, but if you have ovirt engine running, make sure
that power management not active for this host,<br>
</div>
<div>since after timeout (usually 1 minute) when vdsm is down, the
host will be fenced (restarted) - then the vms will surely
stop..<br>
</div>
<div>look at <a moz-do-not-send="true"
href="http://www.ovirt.org/Automatic_Fencing%20">http://www....
</a>for more info about automatic fencing in engine.</div>
</blockquote>
</body>
</html>
--------------090101080002020800090303--
5:10 a.m.
On Thu, Apr 18, 2013 at 07:40:41PM +0100, Alex Leonhardt wrote:
ah, i dont have that configured anyway as I thought from the
beginning that i'd like to keep control over when a host is being
rebooted ...
unfortunately, the VMs are running storage domains too (NFS) so it'd
take those down I believe ... i'll try w/o stopping vdsm first ...
on one that hopefully doesnt run too many critical apps at the time
..
i'll report back how it went if anyone's interested :)
Please do - since Vdsm should've restarted itself anyway.
On 04/17/2013 08:07 AM, Omer Frenkel wrote:
>
>
>------------------------------------------------------------------------
>
> *From: *"Alex Leonhardt" <alex.tuxx(a)gmail.com>
> *To: *"Dan Kenigsberg" <danken(a)redhat.com>
> *Cc: *"oVirt Mailing List" <users(a)ovirt.org>
> *Sent: *Wednesday, April 17, 2013 1:02:25 AM
> *Subject: *Re: [Users] save to restart libvirtd ?
>
> would that stop any of the running VMs ?
>
>it shouldn't, but if you have ovirt engine running, make sure that
>power management not active for this host,
>since after timeout (usually 1 minute) when vdsm is down, the host
>will be fenced (restarted) - then the vms will surely stop..
>look at http://www.ovirt.org/Automatic_Fencing
><http://www.ovirt.org/Automatic_Fencing%20>for more info about
>automatic fencing in engine.
4029
days inactive
4036
days old
8 comments
4 participants
participants (4)
-
Alex Leonhardt -
Dan Kenigsberg -
Martin Kletzander -
Omer Frenkel