2:10 a.m.
Looking for a design advise on oVirt provisioning. I'm running a PoC lab on
single bare-metal host (suddenly it was setup with just Local Storage
domain) and
no I'd like to rebuild the setup by making a cluster of 2 physical servers,
no external storage array available. That are the options here? is there
any options to build cheap HA cluster with just 2 servers?
Thanks in advance!
Artem
2:24 a.m.
This is a multi-part message in MIME format.
--------------7FE4F8F8D0FA672540A59BDE
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
For HA you will need some kind of storage available to all the compute
nodes in the cluster. If you have no external storage and few nodes, I
think your best option for storage is gluster , and the minimum number
of nodes you will need for HA is 3 (the third gluster node can be
metadata-only, but you still need that third node to give you quorum,
avoid split-brains and have something that you can call "HA" with a
straight face.
Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153
On 03/11/17 08:10, Artem Tambovskiy wrote:
Looking for a design advise on oVirt provisioning. I'm running a
PoC
lab on single bare-metal host (suddenly it was setup with just Local
Storage domain) and
no I'd like to rebuild the setup by making a cluster of 2 physical
servers, no external storage array available. That are the options
here? is there any options to build cheap HA cluster with just 2 servers?
Thanks in advance!
Artem
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
--------------7FE4F8F8D0FA672540A59BDE
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>For HA you will need some kind of storage available to all the
compute nodes in the cluster. If you have no external storage and
few nodes, I think your best option for storage is gluster , and
the minimum number of nodes you will need for HA is 3 (the third
gluster node can be metadata-only, but you still need that third
node to give you quorum, avoid split-brains and have something
that you can call "HA" with a straight face.<br>
</p>
<pre class="moz-signature" cols="72">Eduardo Mayoral Jimeno
(<a class="moz-txt-link-abbreviated"
href="mailto:emayoral@arsys.es">emayoral@arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153</pre>
<div class="moz-cite-prefix">On 03/11/17 08:10, Artem Tambovskiy
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAAq+Rfab8fvhNV+uP2hEceJywaezEqyRSDwdBx60333SMR7k4A@mail.gmail.com">
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
<div dir="ltr">
<div>Looking for a design advise on oVirt provisioning. I'm
running a PoC lab on single bare-metal host (suddenly it was
setup with just Local Storage domain) and </div>
<div>no I'd like to rebuild the setup by making a cluster of 2
physical servers, no external storage array available. That
are the options here? is there any options to build cheap HA
cluster with just 2 servers? </div>
<div><br>
</div>
<div>Thanks in advance!</div>
<div><br>
</div>
<div>Artem</div>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
Users mailing list
<a class="moz-txt-link-abbreviated"
href="mailto:Users@ovirt.org">Users@ovirt.org</a>
<a class="moz-txt-link-freetext"
href="http://lists.ovirt.org/mailman/listinfo/users">http://...
</pre>
</blockquote>
<br>
</body>
</html>
--------------7FE4F8F8D0FA672540A59BDE--
2:39 a.m.
Thanks Eduardo!
I think I can find a third server to build a glusterFS storage. So the
first step will be to install a self-hosted engine on the new server and
start building a glusterFS storage. IS there any easy way to migrate
existing 5 VM's running on the second bare-metal oVirt host, right? I found
a little bit tricky moving oVirt backups between the hosts (at least I
failed to replicate the existing VM's on the second server).
Regards,
Artem
On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es> wrote:
For HA you will need some kind of storage available to all the
compute
nodes in the cluster. If you have no external storage and few nodes, I
think your best option for storage is gluster , and the minimum number of
nodes you will need for HA is 3 (the third gluster node can be
metadata-only, but you still need that third node to give you quorum, avoid
split-brains and have something that you can call "HA" with a straight face.
Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.+34 941 620 145
ext. 5153 <+34%20941%2062%2001%2045>
On 03/11/17 08:10, Artem Tambovskiy wrote:
Looking for a design advise on oVirt provisioning. I'm running a PoC lab
on single bare-metal host (suddenly it was setup with just Local Storage
domain) and
no I'd like to rebuild the setup by making a cluster of 2 physical
servers, no external storage array available. That are the options here? is
there any options to build cheap HA cluster with just 2 servers?
Thanks in advance!
Artem
_______________________________________________
Users mailing listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users
2:50 a.m.
On fre, 2017-11-03 at 10:39 +0300, Artem Tambovskiy wrote:
Thanks Eduardo!
I think I can find a third server to build a glusterFS storage. So
the first step will be to install a self-hosted engine on the new
server and start building a glusterFS storage.
You´ll need to build the Gluster storage first, as you´ll want to
install the Hosted Engine _in_ the HA Gluster storage, right?
/K
IS there any easy way to migrate existing 5 VM's running on the
second bare-metal oVirt host, right? I found a little bit tricky
moving oVirt backups between the hosts (at least I failed to
replicate the existing VM's on the second server).
Regards,
Artem
On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es>
wrote:
> For HA you will need some kind of storage available to all the
> compute nodes in the cluster. If you have no external storage and
> few nodes, I think your best option for storage is gluster , and
> the minimum number of nodes you will need for HA is 3 (the third
> gluster node can be metadata-only, but you still need that third
> node to give you quorum, avoid split-brains and have something that
> you can call "HA" with a straight face.
> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
> Administrador de sistemas. Departamento de Plataformas. Arsys
> internet.
> +34 941 620 145 ext. 5153
> On 03/11/17 08:10, Artem Tambovskiy wrote:
> > Looking for a design advise on oVirt provisioning. I'm running a
> > PoC lab on single bare-metal host (suddenly it was setup with
> > just Local Storage domain) and
> > no I'd like to rebuild the setup by making a cluster of 2
> > physical servers, no external storage array available. That are
> > the options here? is there any options to build cheap HA cluster
> > with just 2 servers?
> >
> > Thanks in advance!
> >
> > Artem
> >
> >
> > _______________________________________________
> > Users mailing list
> > Users(a)ovirt.org
> > http://lists.ovirt.org/mailman/listinfo/users
>
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
3:11 a.m.
Good point! Need to focus on this first
Thanks,
Artem
On Fri, Nov 3, 2017 at 10:50 AM, Karli Sjöberg <karli(a)inparadise.se> wrote:
On fre, 2017-11-03 at 10:39 +0300, Artem Tambovskiy wrote:
> Thanks Eduardo!
>
> I think I can find a third server to build a glusterFS storage. So
> the first step will be to install a self-hosted engine on the new
> server and start building a glusterFS storage.
You´ll need to build the Gluster storage first, as you´ll want to
install the Hosted Engine _in_ the HA Gluster storage, right?
/K
> IS there any easy way to migrate existing 5 VM's running on the
> second bare-metal oVirt host, right? I found a little bit tricky
> moving oVirt backups between the hosts (at least I failed to
> replicate the existing VM's on the second server).
>
> Regards,
> Artem
>
>
>
> On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es>
> wrote:
> > For HA you will need some kind of storage available to all the
> > compute nodes in the cluster. If you have no external storage and
> > few nodes, I think your best option for storage is gluster , and
> > the minimum number of nodes you will need for HA is 3 (the third
> > gluster node can be metadata-only, but you still need that third
> > node to give you quorum, avoid split-brains and have something that
> > you can call "HA" with a straight face.
> > Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
> > Administrador de sistemas. Departamento de Plataformas. Arsys
> > internet.
> > +34 941 620 145 ext. 5153
> > On 03/11/17 08:10, Artem Tambovskiy wrote:
> > > Looking for a design advise on oVirt provisioning. I'm running a
> > > PoC lab on single bare-metal host (suddenly it was setup with
> > > just Local Storage domain) and
> > > no I'd like to rebuild the setup by making a cluster of 2
> > > physical servers, no external storage array available. That are
> > > the options here? is there any options to build cheap HA cluster
> > > with just 2 servers?
> > >
> > > Thanks in advance!
> > >
> > > Artem
> > >
> > >
> > > _______________________________________________
> > > Users mailing list
> > > Users(a)ovirt.org
> > > http://lists.ovirt.org/mailman/listinfo/users
> >
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
3:15 a.m.
This is a multi-part message in MIME format.
--------------54D2274BE68ED5ED9B86230D
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Exporting the VMs to an export storage domain on the first server,
copying the files to another export storage domain on the second server
and importing it should work. What kind of problem did you find?
As for the order of doing things, my first oVirt lab was a 3-node setup
with gluster (oVirt version 4.0). I found this guide helpful:
https://www.ovirt.org/blog/2016/08/up-and-running-with-ovirt-4-0-and-glus...
I see there is a new version of it:
https://www.ovirt.org/blog/2017/04/up-and-running-with-ovirt-4.1-and-glus...
If I understand correctly you plan to re-use the server you already
have. I think you will probably find it much easier have to export your
VMs, re-deploy and import your VMs back. It means some downtime, but you
seem very constrained on the hardware available and there is not much
room to manouver...
Good luck!
Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153
On 03/11/17 08:39, Artem Tambovskiy wrote:
Thanks Eduardo!
I think I can find a third server to build a glusterFS storage. So the
first step will be to install a self-hosted engine on the new server
and start building a glusterFS storage. IS there any easy way to
migrate existing 5 VM's running on the second bare-metal oVirt host,
right? I found a little bit tricky moving oVirt backups between the
hosts (at least I failed to replicate the existing VM's on the second
server).
Regards,
Artem
On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es
<mailto:emayoral@arsys.es>> wrote:
For HA you will need some kind of storage available to all the
compute nodes in the cluster. If you have no external storage and
few nodes, I think your best option for storage is gluster , and
the minimum number of nodes you will need for HA is 3 (the third
gluster node can be metadata-only, but you still need that third
node to give you quorum, avoid split-brains and have something
that you can call "HA" with a straight face.
Eduardo Mayoral Jimeno (emayoral(a)arsys.es <mailto:emayoral@arsys.es>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153 <tel:+34%20941%2062%2001%2045>
On 03/11/17 08:10, Artem Tambovskiy wrote:
> Looking for a design advise on oVirt provisioning. I'm running a
> PoC lab on single bare-metal host (suddenly it was setup with
> just Local Storage domain) and
> no I'd like to rebuild the setup by making a cluster of 2
> physical servers, no external storage array available. That are
> the options here? is there any options to build cheap HA cluster
> with just 2 servers?
>
> Thanks in advance!
>
> Artem
>
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org <mailto:Users@ovirt.org>
> http://lists.ovirt.org/mailman/listinfo/users
> <http://lists.ovirt.org/mailman/listinfo/users>
--------------54D2274BE68ED5ED9B86230D
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Exporting the VMs to an export storage domain on the first
server, copying the files to another export storage domain on the
second server and importing it should work. What kind of problem
did you find?</p>
<p>As for the order of doing things, my first oVirt lab was a 3-node
setup with gluster (oVirt version 4.0). I found this guide
helpful:</p>
<p><a class="moz-txt-link-freetext"
href="https://www.ovirt.org/blog/2016/08/up-and-running-with-ovirt-4...
<p>I see there is a new version of it:</p>
<p><a class="moz-txt-link-freetext"
href="https://www.ovirt.org/blog/2017/04/up-and-running-with-ovirt-4...
<p>If I understand correctly you plan to re-use the server you
already have. I think you will probably find it much easier have
to export your VMs, re-deploy and import your VMs back. It means
some downtime, but you seem very constrained on the hardware
available and there is not much room to manouver...<br>
</p>
<p>Good luck!<br>
</p>
<pre class="moz-signature" cols="72">Eduardo Mayoral Jimeno
(<a class="moz-txt-link-abbreviated"
href="mailto:emayoral@arsys.es">emayoral@arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153</pre>
<div class="moz-cite-prefix">On 03/11/17 08:39, Artem Tambovskiy
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAAq+RfYZT9nYvPh3NttwrRu1JRb+1qhzihGi+WETrO8+vFcEng@mail.gmail.com">
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
<div dir="ltr">Thanks Eduardo!<br>
<br>
I think I can find a third server to build a glusterFS storage.
So the first step will be to install a self-hosted engine on the
new server and start building a glusterFS storage. IS there any
easy way to migrate existing 5 VM's running on the second
bare-metal oVirt host, right? I found a little bit tricky moving
oVirt backups between the hosts (at least I failed to replicate
the existing VM's on the second server).
<div><br>
</div>
<div>Regards,</div>
<div>Artem<br>
<br>
<br>
</div>
</div>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Fri, Nov 3, 2017 at 10:24 AM,
Eduardo Mayoral <span dir="ltr"><<a
href="mailto:emayoral@arsys.es" target="_blank"
moz-do-not-send="true">emayoral(a)arsys.es</a>&gt;</span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<p>For HA you will need some kind of storage available to
all the compute nodes in the cluster. If you have no
external storage and few nodes, I think your best option
for storage is gluster , and the minimum number of nodes
you will need for HA is 3 (the third gluster node can be
metadata-only, but you still need that third node to
give you quorum, avoid split-brains and have something
that you can call "HA" with a straight face.<br>
</p>
<pre class="m_-1929299242340323219moz-signature"
cols="72">Eduardo Mayoral Jimeno (<a
class="m_-1929299242340323219moz-txt-link-abbreviated"
href="mailto:emayoral@arsys.es" target="_blank"
moz-do-not-send="true">emayoral(a)arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
<a href="tel:+34%20941%2062%2001%2045" value="+34941620145"
target="_blank" moz-do-not-send="true">+34 941 620 145 ext.
5153</a></pre>
<div>
<div class="h5">
<div class="m_-1929299242340323219moz-cite-prefix">On
03/11/17 08:10, Artem Tambovskiy wrote:<br>
</div>
</div>
</div>
<blockquote type="cite">
<div>
<div class="h5">
<div dir="ltr">
<div>Looking for a design advise on oVirt
provisioning. I'm running a PoC lab on single
bare-metal host (suddenly it was setup with just
Local Storage domain) and </div>
<div>no I'd like to rebuild the setup by making a
cluster of 2 physical servers, no external
storage array available. That are the options
here? is there any options to build cheap HA
cluster with just 2 servers? </div>
<div><br>
</div>
<div>Thanks in advance!</div>
<div><br>
</div>
<div>Artem</div>
</div>
<br>
<fieldset
class="m_-1929299242340323219mimeAttachmentHeader"></fieldset>
<br>
</div>
</div>
<pre>______________________________<wbr>_________________
Users mailing list
<a class="m_-1929299242340323219moz-txt-link-abbreviated"
href="mailto:Users@ovirt.org" target="_blank"
moz-do-not-send="true">Users(a)ovirt.org</a>
<a class="m_-1929299242340323219moz-txt-link-freetext"
href="http://lists.ovirt.org/mailman/listinfo/users" target="_blank"
moz-do-not-send="true">http://lists.ovirt.org/<wbr>mai...
</pre>
</blockquote>
<br>
</div>
</blockquote>
</div>
<br>
</div>
</blockquote>
<br>
</body>
</html>
--------------54D2274BE68ED5ED9B86230D--
4:49 a.m.
Hi,
you should take a look at the hyper converged way of installing oVirt.
We have a cockpit wizard that does almost everything for you:
https://www.ovirt.org/documentation/gluster-hyperconverged/chap-Deploying...
It uses three hosts and collocates the VMs together with Gluster storage.
Best regards
--
Martin Sivak
SLA /oVirt
On Fri, Nov 3, 2017 at 8:39 AM, Artem Tambovskiy
<artem.tambovskiy(a)gmail.com> wrote:
Thanks Eduardo!
I think I can find a third server to build a glusterFS storage. So the first
step will be to install a self-hosted engine on the new server and start
building a glusterFS storage. IS there any easy way to migrate existing 5
VM's running on the second bare-metal oVirt host, right? I found a little
bit tricky moving oVirt backups between the hosts (at least I failed to
replicate the existing VM's on the second server).
Regards,
Artem
On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es> wrote:
>
> For HA you will need some kind of storage available to all the compute
> nodes in the cluster. If you have no external storage and few nodes, I think
> your best option for storage is gluster , and the minimum number of nodes
> you will need for HA is 3 (the third gluster node can be metadata-only, but
> you still need that third node to give you quorum, avoid split-brains and
> have something that you can call "HA" with a straight face.
>
> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
> Administrador de sistemas. Departamento de Plataformas. Arsys internet.
> +34 941 620 145 ext. 5153
>
> On 03/11/17 08:10, Artem Tambovskiy wrote:
>
> Looking for a design advise on oVirt provisioning. I'm running a PoC lab
> on single bare-metal host (suddenly it was setup with just Local Storage
> domain) and
> no I'd like to rebuild the setup by making a cluster of 2 physical
> servers, no external storage array available. That are the options here? is
> there any options to build cheap HA cluster with just 2 servers?
>
> Thanks in advance!
>
> Artem
>
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
>
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
5:29 a.m.
Thanks for an article, Martin!
Any chance to configure a third cost to act as GlusterFS Arbitr only using
this wizard?
And stupid question - how to make this wizard up and running? I've
everything installed and nothing is runnin on port 9090 :)
Regards,
Artem
On Fri, Nov 3, 2017 at 12:49 PM, Martin Sivak <msivak(a)redhat.com> wrote:
Hi,
you should take a look at the hyper converged way of installing oVirt.
We have a cockpit wizard that does almost everything for you:
https://www.ovirt.org/documentation/gluster-hyperconverged/chap-Deploying_
Hyperconverged/
It uses three hosts and collocates the VMs together with Gluster storage.
Best regards
--
Martin Sivak
SLA /oVirt
On Fri, Nov 3, 2017 at 8:39 AM, Artem Tambovskiy
<artem.tambovskiy(a)gmail.com> wrote:
> Thanks Eduardo!
>
> I think I can find a third server to build a glusterFS storage. So the
first
> step will be to install a self-hosted engine on the new server and start
> building a glusterFS storage. IS there any easy way to migrate existing 5
> VM's running on the second bare-metal oVirt host, right? I found a little
> bit tricky moving oVirt backups between the hosts (at least I failed to
> replicate the existing VM's on the second server).
>
> Regards,
> Artem
>
>
>
> On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es>
wrote:
>>
>> For HA you will need some kind of storage available to all the compute
>> nodes in the cluster. If you have no external storage and few nodes, I
think
>> your best option for storage is gluster , and the minimum number of
nodes
>> you will need for HA is 3 (the third gluster node can be metadata-only,
but
>> you still need that third node to give you quorum, avoid split-brains
and
>> have something that you can call "HA" with a straight face.
>>
>> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
>> Administrador de sistemas. Departamento de Plataformas. Arsys internet.
>> +34 941 620 145 ext. 5153
>>
>> On 03/11/17 08:10, Artem Tambovskiy wrote:
>>
>> Looking for a design advise on oVirt provisioning. I'm running a PoC lab
>> on single bare-metal host (suddenly it was setup with just Local Storage
>> domain) and
>> no I'd like to rebuild the setup by making a cluster of 2 physical
>> servers, no external storage array available. That are the options
here? is
>> there any options to build cheap HA cluster with just 2 servers?
>>
>> Thanks in advance!
>>
>> Artem
>>
>>
>> _______________________________________________
>> Users mailing list
>> Users(a)ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
>>
>>
>
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
6:28 a.m.
Hi,
cockpit is enabled by default when you use ovirt-node. You will
probably have to install the necessary cockpit packages yourself on
pure CentOS - you will need cockpit and ovirt + gdeploy cockpit
plugins (sadly I do not recall the exact package names).
With regards to arbiter and the wizard.. I really do not know, but I
will alert my colleagues who might have more detailed knowledge of the
gluster part.
Denis, Sahina: can you please help me here?
Best regards
Martin Sivak
On Fri, Nov 3, 2017 at 11:29 AM, Artem Tambovskiy
<artem.tambovskiy(a)gmail.com> wrote:
Thanks for an article, Martin!
Any chance to configure a third cost to act as GlusterFS Arbitr only using
this wizard?
And stupid question - how to make this wizard up and running? I've
everything installed and nothing is runnin on port 9090 :)
Regards,
Artem
On Fri, Nov 3, 2017 at 12:49 PM, Martin Sivak <msivak(a)redhat.com> wrote:
>
> Hi,
>
> you should take a look at the hyper converged way of installing oVirt.
> We have a cockpit wizard that does almost everything for you:
>
>
>
https://www.ovirt.org/documentation/gluster-hyperconverged/chap-Deploying...
>
> It uses three hosts and collocates the VMs together with Gluster storage.
>
> Best regards
>
> --
> Martin Sivak
> SLA /oVirt
>
> On Fri, Nov 3, 2017 at 8:39 AM, Artem Tambovskiy
> <artem.tambovskiy(a)gmail.com> wrote:
> > Thanks Eduardo!
> >
> > I think I can find a third server to build a glusterFS storage. So the
> > first
> > step will be to install a self-hosted engine on the new server and start
> > building a glusterFS storage. IS there any easy way to migrate existing
> > 5
> > VM's running on the second bare-metal oVirt host, right? I found a
> > little
> > bit tricky moving oVirt backups between the hosts (at least I failed to
> > replicate the existing VM's on the second server).
> >
> > Regards,
> > Artem
> >
> >
> >
> > On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es>
> > wrote:
> >>
> >> For HA you will need some kind of storage available to all the compute
> >> nodes in the cluster. If you have no external storage and few nodes, I
> >> think
> >> your best option for storage is gluster , and the minimum number of
> >> nodes
> >> you will need for HA is 3 (the third gluster node can be metadata-only,
> >> but
> >> you still need that third node to give you quorum, avoid split-brains
> >> and
> >> have something that you can call "HA" with a straight face.
> >>
> >> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
> >> Administrador de sistemas. Departamento de Plataformas. Arsys internet.
> >> +34 941 620 145 ext. 5153
> >>
> >> On 03/11/17 08:10, Artem Tambovskiy wrote:
> >>
> >> Looking for a design advise on oVirt provisioning. I'm running a PoC
> >> lab
> >> on single bare-metal host (suddenly it was setup with just Local
> >> Storage
> >> domain) and
> >> no I'd like to rebuild the setup by making a cluster of 2 physical
> >> servers, no external storage array available. That are the options
> >> here? is
> >> there any options to build cheap HA cluster with just 2 servers?
> >>
> >> Thanks in advance!
> >>
> >> Artem
> >>
> >>
> >> _______________________________________________
> >> Users mailing list
> >> Users(a)ovirt.org
> >> http://lists.ovirt.org/mailman/listinfo/users
> >>
> >>
> >
> >
> > _______________________________________________
> > Users mailing list
> > Users(a)ovirt.org
> > http://lists.ovirt.org/mailman/listinfo/users
> >
3:52 a.m.
Hi,
Can anyone share their experience on deploying hosted-engine with GlusterFS
cluster?
I managed to setup a GlusterFS cluster and started to deploy hosted engine.
At first stage I was beaten by firewall rules - the deployment process
interrupted at GlusterFS config stage.
After fixing the rules I got engine up and running, but the host is
non-operational state still. Logged in to Web UI -> and see 2 action
points: "Gluster command failed on server" and Gluster status is
disconnected for this server"
This is a bit strange, since the cluster was properly detected during
deployment and deployment script was supposed to configure the cluster.
--== STORAGE CONFIGURATION ==--
Please specify the storage you would like to use (glusterfs,
iscsi, fc, nfs3, nfs4)[nfs3]: glusterfs
[ INFO ] Please note that Replica 3 support is required for the shared
storage.
Please specify the full shared storage connection path to use
(example: host:/path): xxxx.xxxxx.xx:/oVirt
[ INFO ] GlusterFS replica 3 Volume detected
Do you want to configure this host and its cluster for gluster?
(Yes, No) [No]: Yes
[ INFO ] GlusterFS replica 3 Volume detected
Any ideas how to fix this?
Thanks in advance!
Regards,
Artem
On Fri, Nov 3, 2017 at 2:28 PM, Martin Sivak <msivak(a)redhat.com> wrote:
Hi,
cockpit is enabled by default when you use ovirt-node. You will
probably have to install the necessary cockpit packages yourself on
pure CentOS - you will need cockpit and ovirt + gdeploy cockpit
plugins (sadly I do not recall the exact package names).
With regards to arbiter and the wizard.. I really do not know, but I
will alert my colleagues who might have more detailed knowledge of the
gluster part.
Denis, Sahina: can you please help me here?
Best regards
Martin Sivak
On Fri, Nov 3, 2017 at 11:29 AM, Artem Tambovskiy
<artem.tambovskiy(a)gmail.com> wrote:
> Thanks for an article, Martin!
> Any chance to configure a third cost to act as GlusterFS Arbitr only
using
> this wizard?
>
> And stupid question - how to make this wizard up and running? I've
> everything installed and nothing is runnin on port 9090 :)
>
> Regards,
> Artem
>
> On Fri, Nov 3, 2017 at 12:49 PM, Martin Sivak <msivak(a)redhat.com> wrote:
>>
>> Hi,
>>
>> you should take a look at the hyper converged way of installing oVirt.
>> We have a cockpit wizard that does almost everything for you:
>>
>>
>> https://www.ovirt.org/documentation/gluster-
hyperconverged/chap-Deploying_Hyperconverged/
>>
>> It uses three hosts and collocates the VMs together with Gluster
storage.
>>
>> Best regards
>>
>> --
>> Martin Sivak
>> SLA /oVirt
>>
>> On Fri, Nov 3, 2017 at 8:39 AM, Artem Tambovskiy
>> <artem.tambovskiy(a)gmail.com> wrote:
>> > Thanks Eduardo!
>> >
>> > I think I can find a third server to build a glusterFS storage. So the
>> > first
>> > step will be to install a self-hosted engine on the new server and
start
>> > building a glusterFS storage. IS there any easy way to migrate
existing
>> > 5
>> > VM's running on the second bare-metal oVirt host, right? I found a
>> > little
>> > bit tricky moving oVirt backups between the hosts (at least I failed
to
>> > replicate the existing VM's on the second server).
>> >
>> > Regards,
>> > Artem
>> >
>> >
>> >
>> > On Fri, Nov 3, 2017 at 10:24 AM, Eduardo Mayoral <emayoral(a)arsys.es>
>> > wrote:
>> >>
>> >> For HA you will need some kind of storage available to all the
compute
>> >> nodes in the cluster. If you have no external storage and few nodes,
I
>> >> think
>> >> your best option for storage is gluster , and the minimum number of
>> >> nodes
>> >> you will need for HA is 3 (the third gluster node can be
metadata-only,
>> >> but
>> >> you still need that third node to give you quorum, avoid split-brains
>> >> and
>> >> have something that you can call "HA" with a straight face.
>> >>
>> >> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
>> >> Administrador de sistemas. Departamento de Plataformas. Arsys
internet.
>> >> +34 941 620 145 ext. 5153
>> >>
>> >> On 03/11/17 08:10, Artem Tambovskiy wrote:
>> >>
>> >> Looking for a design advise on oVirt provisioning. I'm running a
PoC
>> >> lab
>> >> on single bare-metal host (suddenly it was setup with just Local
>> >> Storage
>> >> domain) and
>> >> no I'd like to rebuild the setup by making a cluster of 2 physical
>> >> servers, no external storage array available. That are the options
>> >> here? is
>> >> there any options to build cheap HA cluster with just 2 servers?
>> >>
>> >> Thanks in advance!
>> >>
>> >> Artem
>> >>
>> >>
>> >> _______________________________________________
>> >> Users mailing list
>> >> Users(a)ovirt.org
>> >> http://lists.ovirt.org/mailman/listinfo/users
>> >>
>> >>
>> >
>> >
>> > _______________________________________________
>> > Users mailing list
>> > Users(a)ovirt.org
>> > http://lists.ovirt.org/mailman/listinfo/users
>> >
>
>
2:40 a.m.
This is a multi-part message in MIME format.
------------15f80d354a578e9276895b40fa
Content-Type: text/plain; format=flowed; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Check on drbd. I have used that to build a cluster for two servers. It ne=
ed=20
some more work than a three node gluster conf but works well.
I even think they have a white paper on how to do it for virtualization.
/Johan
On November 3, 2017 08:11:04 Artem Tambovskiy <artem.tambovskiy(a)gmail.com=
=20
wrote:
Looking for a design advise on oVirt provisioning. I'm running a
PoC la=
b on
single bare-metal host (suddenly it was setup with just Local
Storage
domain) and
no I'd like to rebuild the setup by making a cluster of 2 physical serv=
ers,
no external storage array available. That are the options here? is
ther=
e
any options to build cheap HA cluster with just 2 servers?
Thanks in advance!
Artem
----------
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
------------15f80d354a578e9276895b40fa
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
<html>
<head>
</head>
<body>
<div style=3D"color: black;">
<div style=3D"color: black;">
<p style=3D"margin: 0 0 1em 0; color: black;">Check on drbd. I have used
=
that
to build a cluster for two servers. It need some more work than a three
node gluster conf but works well.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">I even think they have a
wh=
ite
paper on how to do it for virtualization.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">/Johan</p>
</div>
<div style=3D"color: black;">
<p
style=3D"color: black; font-size: 10pt; font-family: Arial, sans-serif; m=
argin: 10pt 0;">On
November 3, 2017 08:11:04 Artem Tambovskiy
&lt;artem.tambovskiy(a)gmail.com&gt; wrote:</p>
<blockquote type=3D"cite" class=3D"gmail_quote"
style=3D"margin: 0 0 0 0.75ex; border-left: 1px solid #808080; padding-le=
ft: 0.75ex;">
<div dir=3D"ltr"><div>Looking for a design advise on oVirt
provisioning.
I'm running a PoC lab on single bare-metal host (suddenly it was setu=
p
with just Local Storage domain) and=C2=A0</div><div>no I'd like to
re=
build
the setup by making a cluster of 2 physical servers, no external storage
array available. That are the options here? is there any options to build
cheap HA cluster with just 2
servers?=C2=A0</div><div><br></div><div>Than=
ks in
advance!</div><div><br></div><div>Artem</div></div>
_______________________________________________<br>
Users mailing list<br>
<a class=3D"aqm-autolink aqm-autowrap"
href=3D"mailto:Users%40ovirt.org">Users@ovirt.org</a><br>
<a class=3D"aqm-autolink aqm-autowrap"
href=3D"http://lists.ovirt.org/mailman/listinfo/users">http:...
.org/mailman/listinfo/users</a><br>
<br>
</blockquote>
</div>
</div>
</body>
</html>
------------15f80d354a578e9276895b40fa--
3:04 a.m.
This is a multi-part message in MIME format.
--------------933D75B60C3821BFBAB2AF63
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Just genuinely curious, how do you avoid split-brain situations with a
2-node setup (be it drbd, gluster or anything else)?
Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153
On 03/11/17 08:40, Johan Bernhardsson wrote:
Check on drbd. I have used that to build a cluster for two servers. It
need some more work than a three node gluster conf but works well.
I even think they have a white paper on how to do it for virtualization.
/Johan
On November 3, 2017 08:11:04 Artem Tambovskiy
<artem.tambovskiy(a)gmail.com> wrote:
> Looking for a design advise on oVirt provisioning. I'm running a PoC
> lab on single bare-metal host (suddenly it was setup with just Local
> Storage domain) and
> no I'd like to rebuild the setup by making a cluster of 2 physical
> servers, no external storage array available. That are the options
> here? is there any options to build cheap HA cluster with just 2
> servers?
>
> Thanks in advance!
>
> Artem
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org <mailto:Users%40ovirt.org>
> http://lists.ovirt.org/mailman/listinfo/users
>
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
--------------933D75B60C3821BFBAB2AF63
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Just genuinely curious, how do you avoid split-brain situations
with a 2-node setup (be it drbd, gluster or anything else)?<br>
</p>
<pre class="moz-signature" cols="72">Eduardo Mayoral Jimeno
(<a class="moz-txt-link-abbreviated"
href="mailto:emayoral@arsys.es">emayoral@arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153</pre>
<div class="moz-cite-prefix">On 03/11/17 08:40, Johan Bernhardsson
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:15f80d35370.2768.ffe2d6cdf294efc1bb86526523af9945@kafit.se">
<meta http-equiv="Content-Type" content="text/html;
charset=utf-8">
<div style="color: black;">
<div style="color: black;">
<p style="margin: 0 0 1em 0; color: black;">Check on drbd. I
have used that
to build a cluster for two servers. It need some more work
than a three
node gluster conf but works well.</p>
<p style="margin: 0 0 1em 0; color: black;">I even think they
have a white
paper on how to do it for virtualization.</p>
<p style="margin: 0 0 1em 0; color: black;">/Johan</p>
</div>
<div style="color: black;">
<p style="color: black; font-size: 10pt; font-family: Arial,
sans-serif; margin: 10pt 0;">On
November 3, 2017 08:11:04 Artem Tambovskiy
<a class="moz-txt-link-rfc2396E"
href="mailto:artem.tambovskiy@gmail.com"><artem.tambovskiy@gmail.com></a>
wrote:</p>
<blockquote type="cite" class="gmail_quote"
style="margin: 0 0
0 0.75ex; border-left: 1px solid #808080; padding-left:
0.75ex;">
<div dir="ltr">
<div>Looking for a design advise on oVirt provisioning.
I'm running a PoC lab on single bare-metal host
(suddenly it was setup
with just Local Storage domain) and </div>
<div>no I'd like to rebuild
the setup by making a cluster of 2 physical servers, no
external storage
array available. That are the options here? is there any
options to build
cheap HA cluster with just 2 servers? </div>
<div><br>
</div>
<div>Thanks in
advance!</div>
<div><br>
</div>
<div>Artem</div>
</div>
_______________________________________________<br>
Users mailing list<br>
<a class="aqm-autolink aqm-autowrap"
href="mailto:Users%40ovirt.org"
moz-do-not-send="true">Users(a)ovirt.org</a><br>
<a class="aqm-autolink aqm-autowrap"
href="http://lists.ovirt.org/mailman/listinfo/users"
moz-do-not-send="true">http://lists.ovirt.org/mailman/listin...
<br>
</blockquote>
</div>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
Users mailing list
<a class="moz-txt-link-abbreviated"
href="mailto:Users@ovirt.org">Users@ovirt.org</a>
<a class="moz-txt-link-freetext"
href="http://lists.ovirt.org/mailman/listinfo/users">http://...
</pre>
</blockquote>
<br>
</body>
</html>
--------------933D75B60C3821BFBAB2AF63--
3:15 a.m.
This is a multi-part message in MIME format.
------------15f80f30a1b78e9276832d9c5a
Content-Type: text/plain; format=flowed; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
On drbd you have one master and use iscsi to share the storage. If the=20
master fails it will fail over to the other.
You have one up that will act as the iscsi storage and jump between the h=
osts.
If you run drbd master master ... Then you might get split brain and that=
=20
is not fun :)
On November 3, 2017 09:04:49 Eduardo Mayoral <emayoral(a)arsys.es> wrote:
Just genuinely curious, how do you avoid split-brain situations with
a
2-node setup (be it drbd, gluster or anything else)?
Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153
On 03/11/17 08:40, Johan Bernhardsson wrote:
>
> Check on drbd. I have used that to build a cluster for two servers. It
> need some more work than a three node gluster conf but works well.
>
> I even think they have a white paper on how to do it for virtualizatio=
n.
>
> /Johan
>
> On November 3, 2017 08:11:04 Artem Tambovskiy
> <artem.tambovskiy(a)gmail.com> wrote:
>
>> Looking for a design advise on oVirt provisioning. I'm running a PoC
>> lab on single bare-metal host (suddenly it was setup with just Local
>> Storage domain) and=C2=A0
>> no I'd like to rebuild the setup by making a cluster of 2 physical
>> servers, no external storage array available. That are the options
>> here? is there any options to build cheap HA cluster with just 2
>> servers?=C2=A0
>>
>> Thanks in advance!
>>
>> Artem
>> _______________________________________________
>> Users mailing list
>> Users(a)ovirt.org <mailto:Users%40ovirt.org>
>> http://lists.ovirt.org/mailman/listinfo/users
>>
>
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
------------15f80f30a1b78e9276832d9c5a
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
<html>
<head>
=20
</head>
<body>
<div style=3D"color: black;">
<div style=3D"color: black;">
<p style=3D"margin: 0 0 1em 0; color: black;">On drbd you have one
master=
and
use iscsi to share the storage. If the master fails it will fail over to
the other.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">You have one up that will
a=
ct
as the iscsi storage and jump between the hosts.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">If you run drbd master
mast=
er
... Then you might get split brain and that is not fun :)</p>
</div>
<div style=3D"color: black;">
<p
style=3D"color: black; font-size: 10pt; font-family: Arial, sans-serif; m=
argin: 10pt 0;">On
November 3, 2017 09:04:49 Eduardo Mayoral &lt;emayoral(a)arsys.es&gt; wrote=
:</p>
<blockquote type=3D"cite" class=3D"gmail_quote"
style=3D"margin: 0 0 0 0.75ex; border-left: 1px solid #808080; padding-le=
ft: 0.75ex;">
<p>Just genuinely curious, how do you avoid split-brain situations
with a 2-node setup (be it drbd, gluster or anything else)?<br>
</p>
<pre class=3D"moz-signature" cols=3D"72">Eduardo Mayoral
Jimeno (<a
class=3D"moz-txt-link-abbreviated"
href=3D"mailto:emayoral@arsys.es">emayoral@arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153</pre>
<div class=3D"moz-cite-prefix">On 03/11/17 08:40, Johan Bernhardsson
wrote:<br>
</div>
<blockquote type=3D"cite"
cite=3D"mid:15f80d35370.2768.ffe2d6cdf294efc1bb86526523af9945@kafit=
.se">
<meta http-equiv=3D"Content-Type" content=3D"text/html;
charset=3Du=
tf-8">
<div style=3D"color: black;">
<div style=3D"color: black;">
<p style=3D"margin: 0 0 1em 0; color: black;">Check on drbd. I
have used that
to build a cluster for two servers. It need some more work
than a three
node gluster conf but works well.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">I even think they
have a white
paper on how to do it for virtualization.</p>
<p style=3D"margin: 0 0 1em 0; color: black;">/Johan</p>
</div>
<div style=3D"color: black;">
<p style=3D"color: black; font-size: 10pt; font-family: Arial,
sans-serif; margin: 10pt 0;">On
November 3, 2017 08:11:04 Artem Tambovskiy
<a class=3D"moz-txt-link-rfc2396E"
href=3D"mailto:artem.tambovskiy@gmail.com"><artem.tambovskiy@gmail.com=
></a>
wrote:</p>
<blockquote type=3D"cite" class=3D"gmail_quote"
style=3D"margin=
: 0 0
0 0.75ex; border-left: 1px solid #808080; padding-left:
0.75ex;">
<div dir=3D"ltr">
<div>Looking for a design advise on oVirt provisioning.
I'm running a PoC lab on single bare-metal host
(suddenly it was setup
with just Local Storage domain) and=C2=A0</div>
<div>no I'd like to rebuild
the setup by making a cluster of 2 physical servers, no
external storage
array available. That are the options here? is there any
options to build
cheap HA cluster with just 2 servers?=C2=A0</div>
<div<br>
</div>
<div>Thanks in
advance!</div>
<div<br>
</div>
<div>Artem</div>
</div>
_______________________________________________<br>
Users mailing list<br>
<a class=3D"aqm-autolink aqm-autowrap"
href=3D"mailto:Users%40ovirt.org"
moz-do-not-send=3D"true">Users(a)ovirt.org</a<br>
<a class=3D"aqm-autolink aqm-autowrap"
href=3D"http://lists.ovirt.org/mailman/listinfo/users"
=20
moz-do-not-send=3D"true">http://lists.ovirt.org/mailman/list...
<br>
<br>
</blockquote>
</div>
</div>
<br>
<fieldset class=3D"mimeAttachmentHeader"></fieldset>
<br>
<pre wrap=3D"">_______________________________________________
Users mailing list
<a class=3D"moz-txt-link-abbreviated"
href=3D"mailto:Users@ovirt.org">Users@ovirt.org</a>
<a class=3D"moz-txt-link-freetext"
href=3D"http://lists.ovirt.org/mailman/listinfo/users">http:...
.org/mailman/listinfo/users</a>
</pre>
</blockquote>
<br>
</blockquote>
</div>
</div>
</body>
</html>
------------15f80f30a1b78e9276832d9c5a--
4:06 a.m.
This is an OpenPGP/MIME signed message (RFC 4880 and 3156)
--LpidP5sDUwnqXvKTBqwmoV4WNgVOc3LVR
Content-Type: multipart/mixed; boundary="EuwhS6IvMgri3Uvk0TBflcE1NlHpw3QJU";
protected-headers="v1"
From: Kai Wagner <kwagner(a)suse.com To:
users(a)ovirt.org
Message-ID: <a587b36a-adbd-8f7b-02cf-17c934347ae2(a)suse.com Subject:
Re: [ovirt-users] Advise needed: building cheap HA oVirt cluster with
just 2 physical servers
References: <CAAq+Rfab8fvhNV+uP2hEceJywaezEqyRSDwdBx60333SMR7k4A(a)mail.gmail.com
<15f80d35370.2768.ffe2d6cdf294efc1bb86526523af9945(a)kafit.se <0711091c-66a5-7aa5-0241-3ceba2ba1240(a)arsys.es
<15f80f308a0.2768.ffe2d6cdf294efc1bb86526523af9945(a)kafit.se In-Reply-To:
<15f80f308a0.2768.ffe2d6cdf294efc1bb86526523af9945(a)kafit.se
--EuwhS6IvMgri3Uvk0TBflcE1NlHpw3QJU
Content-Type: multipart/alternative;
boundary="------------2D81AC2A3E3425E12F595D94"
Content-Language: en-US
This is a multi-part message in MIME format.
--------------2D81AC2A3E3425E12F595D94
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Nevertheless you could end up with a split brain situation because
you've no quorum.
So it doesn't matter which kind of cluster you want to build, but you'll
need a third node as a tiebreaker.
I know that a 2 node cluster with drbd would work, but not in all cases,
so do yourself a favor and get a third node.
On 03.11.2017 09:15, Johan Bernhardsson wrote:
<head <meta http-equiv=3D"Content-Type"
content=3D"text/html; charset=3Dutf=
-8" </head <body text=3D"#000000"
bgcolor=3D"#FFFFFF" <p>Nevertheless you could
end up with a split brain situation
because you've no quorum.</p
<p>So it doesn't matter which kind of cluster you want to build, but
you'll need a third node as a tiebreaker.</p
<p>I know that a 2 node cluster with drbd would work, but not in all
cases, so do yourself a favor and get a third node. <br </p
<br <div class=3D"moz-cite-prefix">On
03.11.2017 09:15, Johan Bernhardsso=
n
wrote:<br </div <blockquote type=3D"cite"
cite=3D"mid:15f80f308a0.2768.ffe2d6cdf294efc1bb86526523af9945@kafit=
=2Ese" <div style=3D"color:
black;" <div style=3D"color:
black;" <p
style=3D"margin: 0 0 1em 0; color: black;">On drbd you have
one master and
use iscsi to share the storage. If the master fails it will
fail over to
the other.</p <p
style=3D"margin: 0 0 1em 0; color: black;">You have one up
that will act
as the iscsi storage and jump between the hosts.</p <p style=3D"margin: 0 0 1em 0; color:
black;">If you run drbd
master master
... Then you might get split brain and that is not fun :)</p>=
</div <div style=3D"color:
black;" <p style=3D"color:
black; font-size: 10pt; font-family: Arial,
sans-serif; margin: 10pt 0;">On
November 3, 2017 09:04:49 Eduardo Mayoral
<a class=3D"moz-txt-link-rfc2396E"
href=3D"mailto:emayoral@ar=
sys.es">&lt;emayoral(a)arsys.es&gt;</a> wrote:</p <blockquote type=3D"cite"
class=3D"gmail_quote" style=3D"margin=
: 0 0
0 0.75ex; border-left: 1px solid #808080; padding-left:
0.75ex;" <p>Just genuinely
curious, how do you avoid split-brain
situations with a 2-node setup (be it drbd, gluster or
anything else)?<br
</p <pre
class=3D"moz-signature" cols=3D"72">Eduardo Mayoral Jime=
no (<a class=3D"moz-txt-link-abbreviated"
href=3D"mailto:emayoral@arsys.e=
s" moz-do-not-send=3D"true">emayoral(a)arsys.es</a>)
Administrador de sistemas. Departamento de Plataformas. Arsys internet.
+34 941 620 145 ext. 5153</pre <div
class=3D"moz-cite-prefix">On 03/11/17 08:40, Johan
Bernhardsson wrote:<br
</div <blockquote
type=3D"cite"
cite=3D"mid:15f80d35370.2768.ffe2d6cdf294efc1bb86526523af99=
45(a)kafit.se" <meta
http-equiv=3D"Content-Type" content=3D"text/html;
charset=3Dutf-8"
<div style=3D"color: black;"
<div style=3D"color: black;"
<p style=3D"margin: 0 0 1em 0; color: black;">Check on
drbd. I have used that to build a cluster for two
servers. It need some more work than a three node
gluster conf but works well.</p
<p style=3D"margin: 0 0 1em 0; color: black;">I even
think they have a white paper on how to do it for
virtualization.</p
<p style=3D"margin: 0 0 1em 0; color: black;">/Johan</p=
</div
<div style=3D"color: black;"
<p style=3D"color: black; font-size: 10pt; font-family:=
Arial, sans-serif; margin: 10pt 0;">On November 3,
2017 08:11:04 Artem Tambovskiy <a
class=3D"moz-txt-link-rfc2396E"
href=3D"mailto:artem.tambovskiy@gmail.com"
moz-do-not-send=3D"true"><artem.tambovskiy@gmail=
=2Ecom></a wrote:</p <blockquote type=3D"cite"
class=3D"gmail_quote"
style=3D"margin: 0 0 0 0.75ex; border-left: 1px solid=
#808080; padding-left: 0.75ex;"
<div dir=3D"ltr"
<div>Looking for a design advise on oVirt
provisioning. I'm running a PoC lab on single
bare-metal host (suddenly it was setup with just
Local Storage domain) and=C2=A0</div <div>no I'd like to rebuild the
setup by making a
cluster of 2 physical servers, no external
storage array available. That are the options
here? is there any options to build cheap HA
cluster with just 2 servers?=C2=A0</div <div><br </div
<div>Thanks in advance!</div
<div><br </div <div>Artem</div </div
_______________________________________________<br Users mailing list<br <a class=3D"aqm-autolink
aqm-autowrap"
href=3D"mailto:Users%40ovirt.org"
moz-do-not-send=3D"true">Users(a)ovirt.org</a><br <a class=3D"aqm-autolink
aqm-autowrap"
href=3D"http://lists.ovirt.org/mailman/listinfo/use=
rs"
moz-do-not-send=3D"true">http://lists.ovirt.org/mai=
lman/listinfo/users</a><br
<br
</blockquote </div </div
<br <fieldset
class=3D"mimeAttachmentHeader"></fieldset
<br <pre
wrap=3D"">____________________________________________=
___
Users mailing list
<a class=3D"moz-txt-link-abbreviated"
href=3D"mailto:Users@ovirt.org" moz=
-do-not-send=3D"true">Users(a)ovirt.org</a <a
class=3D"moz-txt-link-freetext" href=3D"http://lists.ovirt.org/mailman=
/listinfo/users"
moz-do-not-send=3D"true">http://lists.ovirt.org/mailman/=
listinfo/users</a </pre </blockquote
<br </blockquote </div
</div <br <fieldset
class=3D"mimeAttachmentHeader"></fieldset
<br <pre
wrap=3D"">_______________________________________________
Users mailing list
<a class=3D"moz-txt-link-abbreviated"
href=3D"mailto:Users@ovirt.org">Use=
rs(a)ovirt.org</a <a
class=3D"moz-txt-link-freetext" href=3D"http://lists.ovirt.org/mailman=
/listinfo/users">http://lists.ovirt.org/mailman/listinfo/users</a </pre
</blockquote <br <pre class=3D"moz-signature"
cols=3D"72">--=20
SUSE Linux GmbH, GF: Felix Imend=C3=B6rffer, Jane Smithard, Graham Norton=
, HRB 21284 (AG N=C3=BCrnberg)</pre
</body </html
--------------2D81AC2A3E3425E12F595D94--
--EuwhS6IvMgri3Uvk0TBflcE1NlHpw3QJU--
--LpidP5sDUwnqXvKTBqwmoV4WNgVOc3LVR
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
On drbd you have one master and use iscsi to share the storage. If the
master fails it will fail over to the other.
You have one up that will act as the iscsi storage and jump between
the hosts.
If you run drbd master master ... Then you might get split brain and
that is not fun :)
On November 3, 2017 09:04:49 Eduardo Mayoral <emayoral(a)arsys.es> wrote:=
> Just genuinely curious, how do you avoid split-brain situations with
> a 2-node setup (be it drbd, gluster or anything else)?
> Eduardo Mayoral Jimeno (emayoral(a)arsys.es)
> Administrador de sistemas. Departamento de Plataformas. Arsys internet=
=2E
> +34 941 620 145 ext. 5153
> On 03/11/17 08:40, Johan Bernhardsson wrote:
> >> Check on drbd. I have used that to build a cluster for
two servers.
>> It need some more work than a three node gluster conf but works well.=
> >> I even think they have a
white paper on how to do it for virtualizati=
on.
> >> /Johan
> >> On November 3, 2017 08:11:04 Artem Tambovskiy
>> <artem.tambovskiy(a)gmail.com> wrote:
> >>> Looking for a design advise on oVirt provisioning.
I'm running a
>>> PoC lab on single bare-metal host (suddenly it was setup with just
>>> Local Storage domain) and=C2=A0
>>> no I'd like to rebuild the setup by making a cluster of 2 physical
>>> servers, no external storage array available. That are the options
>>> here? is there any options to build cheap HA cluster with just 2
>>> servers?=C2=A0
>> >>> Thanks in advance!
>> >>> Artem
>>> _______________________________________________
>>> Users mailing list
>>> Users(a)ovirt.org <mailto:Users%40ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>> > > >> _______________________________________________
>> Users mailing list
>> Users(a)ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
_______________________________________________
Users mailing list
Users(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
--=20
SUSE Linux GmbH, GF: Felix Imend=C3=B6rffer, Jane Smithard, Graham Norton=
, HRB 21284 (AG N=C3=BCrnberg)
--------------2D81AC2A3E3425E12F595D94
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable
<html-----BEGIN PGP SIGNATURE-----
iQEcBAEBCAAGBQJZ/DGZAAoJEB6MkVYW48Ruh90H/i/6pG6WJtfRBlwM5LaQ1AcK
OvzOmoFaX+ZBjq1jUOjghspUO0UUX1FR4OuSoDXMIUk2p7EAhrUoBKSzpXNjw//2
gCLMMMGW3iNAKXsVvnHeNfm4xTMdM6uNtGGSaS6Z4w9gbvrX99HoPnQx3KdiL7Cl
3uEj3Yl5rp4clTy5MRIFFNTweei6KnWyssSkHgmaH+oeqnkTeIdkvYlfDVDkhXfo
dl1/K0FoPMlnEccHzh2mdXEItBFr5FhPxWpFob/6dINci4dSiLy3mQ0h87aTy9+q
DUOOM4efQ0wWr5+AOtmo2gPrBR5yNrnm3qMjjgB8gKytWmxKUcFjmNPXFd4ge/o=
=oCXV
-----END PGP SIGNATURE-----
--LpidP5sDUwnqXvKTBqwmoV4WNgVOc3LVR--
2358
days inactive
2363
days old
13 comments
6 participants
participants (6)
-
Artem Tambovskiy -
Eduardo Mayoral -
Johan Bernhardsson -
Kai Wagner -
Karli Sjöberg -
Martin Sivak