2:18 a.m.
In the past we've talked about running libvirt or ovirt on alterway02.
Now that we have the server access we should really decide. I want to
reach an agreement before the next meeting (monday 18).
First let's review the benefits of each. Feel free to add here.
Benefits of libvirt:
- Simple to set up
- Uses little resources
Benefits of ovirt:
- Eating our own dogfood
- Scales beter when we want to have multiple nodes / HA features
- Has an extensive API
Both have:
- Good integration with foreman
So libvirt would be a better choice if we want to set up fast, but may
limit us in the longer run. Also may use fewer resources, but in the
long run ovirt may allows us to build a bit of monitoring on top of the
API.
I'd like to get more benefits and/downsides of each solution so we can
reach a conclusion fast.
12:24 p.m.
--=-uVr3mLq9O/sV4SkLZhH6
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Hey,
Am Dienstag, den 12.02.2013, 00:18 +0100 schrieb Ewoud Kohl van
Wijngaarden:
- Eating our own dogfood
For me this is one of the most important reasons to use our own tools.
And as libvirt as well as ovirt (node?) provide what we need I'd go with
ovirt (node?).
Greetings
fabian
--=-uVr3mLq9O/sV4SkLZhH6
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: This is a digitally signed message part
Content-Transfer-Encoding: 7bit
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)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=IvSc
-----END PGP SIGNATURE-----
--=-uVr3mLq9O/sV4SkLZhH6--
12:45 p.m.
On 02/12/2013 01:18 AM, Ewoud Kohl van Wijngaarden wrote:
In the past we've talked about running libvirt or ovirt on
alterway02.
Now that we have the server access we should really decide. I want to
reach an agreement before the next meeting (monday 18).
First let's review the benefits of each. Feel free to add here.
Benefits of libvirt:
- Simple to set up
- Uses little resources
Benefits of ovirt:
- Eating our own dogfood
- Scales beter when we want to have multiple nodes / HA features
- Has an extensive API
Both have:
- Good integration with foreman
So libvirt would be a better choice if we want to set up fast, but may
limit us in the longer run. Also may use fewer resources, but in the
long run ovirt may allows us to build a bit of monitoring on top of the
API.
I'd like to get more benefits and/downsides of each solution so we can
reach a conclusion fast.
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
I personally find it hard to understand the base point here.
how do we intend other people to use oVirt - if we are saying it's too
complicated/high-maintenance to use.
we aren't looking for the easiest way here, but the way we can make this
project be better (on this aspect as well).
if we find issues let's tackle them and solve it, working our way around
would bring no added value in the long term.
Moran.
1:05 p.m.
On Tue, Feb 12, 2013 at 11:45:04AM +0200, Moran Goldboim wrote:
On 02/12/2013 01:18 AM, Ewoud Kohl van Wijngaarden wrote:
>In the past we've talked about running libvirt or ovirt on alterway02.
>Now that we have the server access we should really decide. I want to
>reach an agreement before the next meeting (monday 18).
>
>First let's review the benefits of each. Feel free to add here.
>
>Benefits of libvirt:
>- Simple to set up
>- Uses little resources
>
>Benefits of ovirt:
>- Eating our own dogfood
>- Scales beter when we want to have multiple nodes / HA features
>- Has an extensive API
>
>Both have:
>- Good integration with foreman
>
>So libvirt would be a better choice if we want to set up fast, but may
>limit us in the longer run. Also may use fewer resources, but in the
>long run ovirt may allows us to build a bit of monitoring on top of the
>API.
>
>I'd like to get more benefits and/downsides of each solution so we can
>reach a conclusion fast.
I personally find it hard to understand the base point here. how do we
intend other people to use oVirt - if we are saying it's too
complicated/high-maintenance to use. we aren't looking for the easiest
way here, but the way we can make this project be better (on this
aspect as well). if we find issues let's tackle them and solve it,
working our way around would bring no added value in the long term.
While I agree with your point there is a slight nuance here. First of
all we're in need of a solution that we can implement in a short
timeframe because it's blocking other progress. For me personally I know
how to configure libvirt and would have that up and running in no time.
Then when we do have time we can start setting up oVirt. Something
we've said is that we have two rackspace machines as well. We can set up
oVirt there and convert the libvirt solution later.
Of course if someone with more experience can help set it up we'd be
very happy to accept help.
7:17 p.m.
This is an OpenPGP/MIME signed message (RFC 4880 and 3156)
------enig2WNWFRDTHVGASARGGWRFT
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
On 02/12/2013 02:05 AM, Ewoud Kohl van Wijngaarden wrote:
While I agree with your point there is a slight nuance here. First
of
all we're in need of a solution that we can implement in a short
timeframe because it's blocking other progress. For me personally I kno=
w
how to configure libvirt and would have that up and running in no
time.=
Then when we do have time we can start setting up oVirt. Something
we've said is that we have two rackspace machines as well. We can set u=
p
oVirt there and convert the libvirt solution later.
One point against using libvirt to start came from DV in the Infra
meeting the other day - Daniel Veillard is a libvirt maintainer, so we
can trust his point. He figured there would be some complications
involved in converting VMs from libvirt to oVirt if we didn't set things
up most efficiently from the start.
So it sounds as if we can work around that potential issue, but it's
going to be more work in the end to start with libvirt and move to oVirt.=
If we want an end-state of oVirt, maybe we need to go there first.
Perhaps scheduling some time for people-who-know-already to be present
for the oVirt Infra team when we do the work? Can we schedule some
sprints together?
- Karsten
--=20
Karsten 'quaid' Wade, Sr. Analyst - Community Growth
http://TheOpenSourceWay.org .^\ http://community.redhat.com
@quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
------enig2WNWFRDTHVGASARGGWRFT
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iD8DBQFRGmsJ2ZIOBq0ODEERAjyxAJ4/+y347dBIcW0ZkgUkv3wtQGjqegCgvTqZ
C/uEFJQo4ruo8St2VfjkrO0=
=YHrE
-----END PGP SIGNATURE-----
------enig2WNWFRDTHVGASARGGWRFT--
12:57 p.m.
Hi,
On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
Benefits of ovirt:
- Eating our own dogfood
- Scales beter when we want to have multiple nodes / HA features
- Has an extensive API
I'm in favour of using oVirt. One of the things that I liked about
AlterWay's offer was the idea of running oVirt project infrastructure on
oVirt managed VMs.
There are two choices inside the oVirt decision: do we want a separate
oVirt manager, or are we happy having the host managed as a node in
AlterWay's oVirt managed infrastructure? I personally am fine with using
AlterWay's management console for the VMs and the host.
So libvirt would be a better choice if we want to set up fast, but
may
limit us in the longer run. Also may use fewer resources, but in the
long run ovirt may allows us to build a bit of monitoring on top of the
API.
Using AlterWay's oVirt set-up negates the argument about it being longer
to set up too - everything is ready to go right now.
Cheers,
Dave.
--
Dave Neary - Community Action and Impact
Open Source and Standards, Red Hat - http://community.redhat.com
Ph: +33 9 50 71 55 62 / Cell: +33 6 77 01 92 13
1:05 p.m.
--=-OXoXGhuf5hcG1AE6Rr/q
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Am Dienstag, den 12.02.2013, 10:57 +0100 schrieb Dave Neary:
Hi,
=20
On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
> Benefits of ovirt:
> - Eating our own dogfood
> - Scales beter when we want to have multiple nodes / HA features
> - Has an extensive API
=20
I'm in favour of using oVirt. One of the things that I liked about=20
AlterWay's offer was the idea of running oVirt project infrastructure on=
=20
oVirt managed VMs.
=20
There are two choices inside the oVirt decision: do we want a separate=
=20
oVirt manager, or are we happy having the host managed as a node
in=20
AlterWay's oVirt managed infrastructure? I personally am fine with using=
=20
AlterWay's management console for the VMs and the host.
Using AlterWay's management console is a charming idea.
- fabian
> So libvirt would be a better choice if we want to set up fast,
but may
> limit us in the longer run. Also may use fewer resources, but in the
> long run ovirt may allows us to build a bit of monitoring on top of the
> API.
=20
Using AlterWay's oVirt set-up negates the argument about it being longer=
=20
to set up too - everything is ready to go right now.
=20
Cheers,
Dave.
=20
--=-OXoXGhuf5hcG1AE6Rr/q
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: This is a digitally signed message part
Content-Transfer-Encoding: 7bit
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)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=PNg1
-----END PGP SIGNATURE-----
--=-OXoXGhuf5hcG1AE6Rr/q--
1:57 p.m.
On Tue, Feb 12, 2013 at 10:57:58AM +0100, Dave Neary wrote:
On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
>Benefits of ovirt:
>- Eating our own dogfood
>- Scales beter when we want to have multiple nodes / HA features
>- Has an extensive API
I'm in favour of using oVirt. One of the things that I liked about
AlterWay's offer was the idea of running oVirt project
infrastructure on oVirt managed VMs.
There are two choices inside the oVirt decision: do we want a
separate oVirt manager, or are we happy having the host managed as a
node in AlterWay's oVirt managed infrastructure? I personally am
fine with using AlterWay's management console for the VMs and the
host.
Can we use the standard oVirt API to manage our VMs? User level API is a
3.2 feature[1].
>So libvirt would be a better choice if we want to set up fast,
but may
>limit us in the longer run. Also may use fewer resources, but in the
>long run ovirt may allows us to build a bit of monitoring on top of the
>API.
Using AlterWay's oVirt set-up negates the argument about it being
longer to set up too - everything is ready to go right now.
Agreed.
[1]: http://www.ovirt.org/OVirt_3.2_release-management#Features
7:19 p.m.
This is an OpenPGP/MIME signed message (RFC 4880 and 3156)
------enig2XHCCAGFROAQKFCLJEGPO
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
On 02/12/2013 01:57 AM, Dave Neary wrote:
Hi,
=20
On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
> Benefits of ovirt:
> - Eating our own dogfood
> - Scales beter when we want to have multiple nodes / HA features
> - Has an extensive API
=20
I'm in favour of using oVirt. One of the things that I liked about
AlterWay's offer was the idea of running oVirt project infrastructure o=
n
oVirt managed VMs.
=20
There are two choices inside the oVirt decision: do we want a separate
oVirt manager, or are we happy having the host managed as a node in
AlterWay's oVirt managed infrastructure? I personally am fine with usin=
g
AlterWay's management console for the VMs and the host.
=20
> So libvirt would be a better choice if we want to set up fast, but may=
> limit us in the longer run. Also may use fewer resources, but in
the
> long run ovirt may allows us to build a bit of monitoring on top of th=
e
> API.
=20
Using AlterWay's oVirt set-up negates the argument about it being longe=
r
to set up too - everything is ready to go right now.
And doesn't stop us from putting up our own management console/oVirt
manager in the future, right? It should be simple to migrate VMs from
one manager to another, right?
- Karsten
--=20
Karsten 'quaid' Wade, Sr. Analyst - Community Growth
http://TheOpenSourceWay.org .^\ http://community.redhat.com
@quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
------enig2XHCCAGFROAQKFCLJEGPO
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iD8DBQFRGmuA2ZIOBq0ODEERAlNnAKCsmiDCsREZCySKHGfm6Arlmea/5wCgxbsm
4H989oxQJutASHD3TTbvCcI=
=yA+z
-----END PGP SIGNATURE-----
------enig2XHCCAGFROAQKFCLJEGPO--
10:08 p.m.
----- Original Message -----
From: "Karsten 'quaid' Wade"
<kwade(a)redhat.com>
To: infra(a)ovirt.org
Sent: Tuesday, February 12, 2013 6:19:12 PM
Subject: Re: libvirt or ovirt
On 02/12/2013 01:57 AM, Dave Neary wrote:
> Hi,
>
> On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
>> Benefits of ovirt:
>> - Eating our own dogfood
>> - Scales beter when we want to have multiple nodes / HA features
>> - Has an extensive API
>
> I'm in favour of using oVirt. One of the things that I liked about
> AlterWay's offer was the idea of running oVirt project
> infrastructure on
> oVirt managed VMs.
>
> There are two choices inside the oVirt decision: do we want a
> separate
> oVirt manager, or are we happy having the host managed as a node in
> AlterWay's oVirt managed infrastructure? I personally am fine with
> using
> AlterWay's management console for the VMs and the host.
>
>> So libvirt would be a better choice if we want to set up fast, but
>> may
>> limit us in the longer run. Also may use fewer resources, but in
>> the
>> long run ovirt may allows us to build a bit of monitoring on top
>> of the
>> API.
>
> Using AlterWay's oVirt set-up negates the argument about it being
> longer
> to set up too - everything is ready to go right now.
And doesn't stop us from putting up our own management console/oVirt
manager in the future, right? It should be simple to migrate VMs from
one manager to another, right?
Correct. Migrating VM from one oVirt to another is
quite simple.
- Karsten
--
Karsten 'quaid' Wade, Sr. Analyst - Community Growth
http://TheOpenSourceWay.org .^\ http://community.redhat.com
@quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
12:51 p.m.
I don't see any reason to choose libvirt over oVirt, other then physical
resources, if we have them, go with oVirt.
If oVirt is that hard to setup, something is wrong, and it isnt that hard
to setup. So I'm not sure why this discussion exists :-)
On Tue, Feb 12, 2013 at 8:08 PM, Kiril Nesenko <kiril(a)redhat.com> wrote:
----- Original Message -----
> From: "Karsten 'quaid' Wade" <kwade(a)redhat.com>
> To: infra(a)ovirt.org
> Sent: Tuesday, February 12, 2013 6:19:12 PM
> Subject: Re: libvirt or ovirt
>
> On 02/12/2013 01:57 AM, Dave Neary wrote:
> > Hi,
> >
> > On 02/12/2013 12:18 AM, Ewoud Kohl van Wijngaarden wrote:
> >> Benefits of ovirt:
> >> - Eating our own dogfood
> >> - Scales beter when we want to have multiple nodes / HA features
> >> - Has an extensive API
> >
> > I'm in favour of using oVirt. One of the things that I liked about
> > AlterWay's offer was the idea of running oVirt project
> > infrastructure on
> > oVirt managed VMs.
> >
> > There are two choices inside the oVirt decision: do we want a
> > separate
> > oVirt manager, or are we happy having the host managed as a node in
> > AlterWay's oVirt managed infrastructure? I personally am fine with
> > using
> > AlterWay's management console for the VMs and the host.
> >
> >> So libvirt would be a better choice if we want to set up fast, but
> >> may
> >> limit us in the longer run. Also may use fewer resources, but in
> >> the
> >> long run ovirt may allows us to build a bit of monitoring on top
> >> of the
> >> API.
> >
> > Using AlterWay's oVirt set-up negates the argument about it being
> > longer
> > to set up too - everything is ready to go right now.
>
> And doesn't stop us from putting up our own management console/oVirt
> manager in the future, right? It should be simple to migrate VMs from
> one manager to another, right?
Correct. Migrating VM from one oVirt to another is quite simple.
>
> - Karsten
> --
> Karsten 'quaid' Wade, Sr. Analyst - Community Growth
> http://TheOpenSourceWay.org .^\ http://community.redhat.com
> @quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
>
>
> _______________________________________________
> Infra mailing list
> Infra(a)ovirt.org
> http://lists.ovirt.org/mailman/listinfo/infra
>
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
--
/Alexander Rydekull
2:08 p.m.
----- Original Message -----
From: "Ewoud Kohl van Wijngaarden"
<ewoud+ovirt(a)kohlvanwijngaarden.nl>
To: infra(a)ovirt.org
Sent: Tuesday, February 12, 2013 1:18:50 AM
Subject: libvirt or ovirt
In the past we've talked about running libvirt or ovirt on
alterway02.
Now that we have the server access we should really decide. I want to
reach an agreement before the next meeting (monday 18).
First let's review the benefits of each. Feel free to add here.
Benefits of libvirt:
- Simple to set up
- Uses little resources
Benefits of ovirt:
- Eating our own dogfood
- Scales beter when we want to have multiple nodes / HA features
- Has an extensive API
Both have:
- Good integration with foreman
So libvirt would be a better choice if we want to set up fast, but
may
limit us in the longer run. Also may use fewer resources, but in the
long run ovirt may allows us to build a bit of monitoring on top of
the
API.
I'm in favor of using oVirt, we know foreman supports integration with it, so it
shouldn't be a problem to use it together.
I'd go with a stable version instead of a nightly version though.
It might take a bit more time, but we've waited so far, i think it's better to
wait a bit longer and do it properly,
and won't do a temporary solution because using libvirt is simpler.
Eyal.
I'd like to get more benefits and/downsides of each solution so we
can
reach a conclusion fast.
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
4 p.m.
On Tue, Feb 12, 2013 at 12:18:50AM +0100, Ewoud Kohl van Wijngaarden wrote:
In the past we've talked about running libvirt or ovirt on
alterway02.
Now that we have the server access we should really decide. I want to
reach an agreement before the next meeting (monday 18).
I think we have sufficient input (thanks for that!) to make a short
summary of the thread.
- There seems to be a very strong preference to go for oVirt.
- Lack of knowledge setting up shouldn't be a problem because we should
be able to get plenty of help if needed.
- Resource wise we think we have enough and if not we'll find that out
along the way.
- Dave suggested use alterways ovirt. Possible problem there is that the
user level API was only added in 3.2 which is still beta and I think
Alterway is still running 3.1. This could be a project in the future
since oVirt to oVirt migrations should be easy.
Can we conclude that we want to install oVirt on alterway02? If so
there's some questions I do have:
- Do we still want to put it on CentOS 6? It has received limited
testing, but EL6 is (IMHO) easier to maintain than Fedora.
- Do we pick oVirt 3.1 or 3.2? Mike Burns said 3.2 would be the better
choice because migration from 3.1 to 3.2 will not be easy. 3.1 would
also mean running on F17, 3.2 on F18 (or maybe EL6).
4:06 p.m.
Hi,
I plan to reinstall OS tomorrow on Centos6.3, let me know if I have to or
not asap :)
Kevin
2013/2/13 Ewoud Kohl van Wijngaarden <ewoud+ovirt(a)kohlvanwijngaarden.nl>
On Tue, Feb 12, 2013 at 12:18:50AM +0100, Ewoud Kohl van Wijngaarden
wrote:
> In the past we've talked about running libvirt or ovirt on alterway02.
> Now that we have the server access we should really decide. I want to
> reach an agreement before the next meeting (monday 18).
I think we have sufficient input (thanks for that!) to make a short
summary of the thread.
- There seems to be a very strong preference to go for oVirt.
- Lack of knowledge setting up shouldn't be a problem because we should
be able to get plenty of help if needed.
- Resource wise we think we have enough and if not we'll find that out
along the way.
- Dave suggested use alterways ovirt. Possible problem there is that the
user level API was only added in 3.2 which is still beta and I think
Alterway is still running 3.1. This could be a project in the future
since oVirt to oVirt migrations should be easy.
Can we conclude that we want to install oVirt on alterway02? If so
there's some questions I do have:
- Do we still want to put it on CentOS 6? It has received limited
testing, but EL6 is (IMHO) easier to maintain than Fedora.
- Do we pick oVirt 3.1 or 3.2? Mike Burns said 3.2 would be the better
choice because migration from 3.1 to 3.2 will not be easy. 3.1 would
also mean running on F17, 3.2 on F18 (or maybe EL6).
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
--
Kevin Mazière
Responsable Infrastructure
Alter Way – Hosting
1 rue Royal - 227 Bureaux de la Colline
92213 Saint-Cloud Cedex
Tél : +33 (0)1 41 16 38 41
Mob : +33 (0)7 62 55 57 05
http://www.alterway.fr
5:31 p.m.
----- Original Message -----
From: "Kevin Maziere Aubry"
<kevin.maziere(a)alterway.fr>
To: "infra" <infra(a)ovirt.org>
Sent: Wednesday, February 13, 2013 3:06:21 PM
Subject: Re: libvirt or ovirt
Hi,
I plan to reinstall OS tomorrow on Centos6.3, let me know if I have
to or not asap :)
re-installing alterway1 (jenkins master) should be with centos 6.3.
as for alterway2 (ovirt instance), we still need to decide if we'll need
fedora/rhel/centos...
Kevin
2013/2/13 Ewoud Kohl van Wijngaarden <
ewoud+ovirt(a)kohlvanwijngaarden.nl >
On Tue, Feb 12, 2013 at 12:18:50AM +0100, Ewoud Kohl van Wijngaarden
wrote:
> In the past we've talked about running libvirt or ovirt on
> alterway02.
> Now that we have the server access we should really decide. I want
> to
> reach an agreement before the next meeting (monday 18).
I think we have sufficient input (thanks for that!) to make a short
summary of the thread.
- There seems to be a very strong preference to go for oVirt.
- Lack of knowledge setting up shouldn't be a problem because we
should
be able to get plenty of help if needed.
- Resource wise we think we have enough and if not we'll find that
out
along the way.
- Dave suggested use alterways ovirt. Possible problem there is that
the
user level API was only added in 3.2 which is still beta and I think
Alterway is still running 3.1. This could be a project in the future
since oVirt to oVirt migrations should be easy.
Can we conclude that we want to install oVirt on alterway02? If so
there's some questions I do have:
- Do we still want to put it on CentOS 6? It has received limited
testing, but EL6 is (IMHO) easier to maintain than Fedora.
- Do we pick oVirt 3.1 or 3.2? Mike Burns said 3.2 would be the
better
choice because migration from 3.1 to 3.2 will not be easy. 3.1 would
also mean running on F17, 3.2 on F18 (or maybe EL6).
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
--
Kevin Mazière
Responsable Infrastructure
Alter Way – Hosting
1 rue Royal - 227 Bureaux de la Colline
92213 Saint-Cloud Cedex
Tél : +33 (0)1 41 16 38 41
Mob : +33 (0)7 62 55 57 05
http://www.alterway.fr
_______________________________________________
Infra mailing list
Infra(a)ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra
1:01 a.m.
This is an OpenPGP/MIME signed message (RFC 4880 and 3156)
------enig2LCTEWHHITTKOQOTWNCGI
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
On 02/13/2013 05:00 AM, Ewoud Kohl van Wijngaarden wrote:
On Tue, Feb 12, 2013 at 12:18:50AM +0100, Ewoud Kohl van Wijngaarden
wr=
ote:
> In the past we've talked about running libvirt or ovirt on
alterway02.=
> Now that we have the server access we should really decide. I
want to
> reach an agreement before the next meeting (monday 18).
=20
I think we have sufficient input (thanks for that!) to make a short
summary of the thread.
=20
- There seems to be a very strong preference to go for oVirt.
- Lack of knowledge setting up shouldn't be a problem because we should=
be able to get plenty of help if needed.
- Resource wise we think we have enough and if not we'll find that out
along the way.
- Dave suggested use alterways ovirt. Possible problem there is that th=
e
user level API was only added in 3.2 which is still beta and I
think
Alterway is still running 3.1. This could be a project in the future
since oVirt to oVirt migrations should be easy.
=20
Can we conclude that we want to install oVirt on alterway02?=20
+1
If so
there's some questions I do have:
=20
- Do we still want to put it on CentOS 6? It has received limited
testing, but EL6 is (IMHO) easier to maintain than Fedora.
I also like the idea of using what the oVirt userbase is most likely to r=
un.
- Do we pick oVirt 3.1 or 3.2? Mike Burns said 3.2 would be the
better
choice because migration from 3.1 to 3.2 will not be easy. 3.1 would
also mean running on F17, 3.2 on F18 (or maybe EL6).
If we can do oVirt 3.2 on EL6 (CentOS, in this case), that sounds like
the best combination.
I like the idea of using our own software, and I like the idea of
running as much as we can on a stable computing platform. We are picking
from the Red Hat family of Linuxes because it's what the people here
know and trust. I do hope one day to see a mix of Linuxes in the oVirt
infrastructure.
- Karsten
--=20
Karsten 'quaid' Wade, Sr. Analyst - Community Growth
http://TheOpenSourceWay.org .^\ http://community.redhat.com
@quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
------enig2LCTEWHHITTKOQOTWNCGI
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iD8DBQFRHA042ZIOBq0ODEERAneoAJwPu+Mzeh719OypJ/OzWTNQUdryZwCgl44p
VlAhWoRiyxBurYTYM1SLoGk=
=Bh4O
-----END PGP SIGNATURE-----
------enig2LCTEWHHITTKOQOTWNCGI--
12:38 a.m.
On Wed, Feb 13, 2013 at 02:01:28PM -0800, Karsten 'quaid' Wade wrote:
On 02/13/2013 05:00 AM, Ewoud Kohl van Wijngaarden wrote:
> On Tue, Feb 12, 2013 at 12:18:50AM +0100, Ewoud Kohl van Wijngaarden wrote:
>> In the past we've talked about running libvirt or ovirt on alterway02.
>> Now that we have the server access we should really decide. I want to
>> reach an agreement before the next meeting (monday 18).
>
> I think we have sufficient input (thanks for that!) to make a short
> summary of the thread.
>
> - There seems to be a very strong preference to go for oVirt.
> - Lack of knowledge setting up shouldn't be a problem because we should
> be able to get plenty of help if needed.
> - Resource wise we think we have enough and if not we'll find that out
> along the way.
> - Dave suggested use alterways ovirt. Possible problem there is that the
> user level API was only added in 3.2 which is still beta and I think
> Alterway is still running 3.1. This could be a project in the future
> since oVirt to oVirt migrations should be easy.
>
> Can we conclude that we want to install oVirt on alterway02?
+1
> If so
> there's some questions I do have:
>
> - Do we still want to put it on CentOS 6? It has received limited
> testing, but EL6 is (IMHO) easier to maintain than Fedora.
I also like the idea of using what the oVirt userbase is most likely to run.
> - Do we pick oVirt 3.1 or 3.2? Mike Burns said 3.2 would be the better
> choice because migration from 3.1 to 3.2 will not be easy. 3.1 would
> also mean running on F17, 3.2 on F18 (or maybe EL6).
If we can do oVirt 3.2 on EL6 (CentOS, in this case), that sounds like
the best combination.
I like the idea of using our own software, and I like the idea of
running as much as we can on a stable computing platform. We are picking
from the Red Hat family of Linuxes because it's what the people here
know and trust. I do hope one day to see a mix of Linuxes in the oVirt
infrastructure.
And I think that's the conclusion of this thread. Thanks all.
Now to to spin up a VM to try out oVirt 3.2 on CentOS 6 and see if it
works.
4298
days inactive
4302
days old
16 comments
9 participants
participants (9)
-
Alexander Rydekull -
Dave Neary -
Ewoud Kohl van Wijngaarden -
Eyal Edri -
Fabian Deutsch -
Karsten 'quaid' Wade -
Kevin Maziere Aubry -
Kiril Nesenko -
Moran Goldboim