Subject: Re: [vserver] Linux-vServer Support in libvirt
From: Edward Capriolo <edlinuxguru@gmail.com>
Date: Tue, 6 Apr 2010 23:58:01 -0400
Tue, 6 Apr 2010 23:58:01 -0400
On Thu, Apr 1, 2010 at 6:15 PM, Walter Stanish <
walter.stanish@saffrondigital.com> wrote:

>  Hi libvir-list (cc: vserver list),
>
> I am another Linux-vServer user who would like to see support in libvirt.
>
> From a very brief look back through the archives it appears that there may
> have been some architectural concerns re: libvirt's assumptions about what
> kind of network stack a virtualisation platform provides, in that
> Linux-vServer doesn't use a traditional approach to networking and thus may
> break previously valid assumptions.
>
> I found 2007 posts and a post last year (2009).
>

I remeber hearing that the patches have been submitted multiple times
http://www.paul.sladen.org/vserver/archives/200807/0003.html


> Would it be possible for someone with a good overall understanding of
> previous integration attempts and the results thereof to summarise what
> issues were encountered and how difficult it would be to overcome these, so
> that users can understand what's going on.
>
> The issue: linux-vserver is not sexy and hip like some of those other
technologies. Maybe if vserver was a huge hulking patch, had really slow
emulation, and or then more or less got killed by being aquired by a large
non-open source friendly company...

http://boycottnovell.com/2009/02/25/xen-spit-it-out-strategy/

...people would start adopting it.



On Thu, Apr 1, 2010 at 6:15 PM, Walter Stanish <walter.stanish@saffrondigital.com> wrote:

Hi libvir-list (cc: vserver list),

I am another Linux-vServer user who would like to see support in libvirt.

From a very brief look back through the archives it appears that there may have been some architectural concerns re: libvirt's assumptions about what kind of network stack a virtualisation platform provides, in that Linux-vServer doesn't use a traditional approach to networking and thus may break previously valid assumptions.

I found 2007 posts and a post last year (2009).


I remeber hearing that the patches have been submitted multiple times
http://www.paul.sladen.org/vserver/archives/200807/0003.html


Would it be possible for someone with a good overall understanding of previous integration attempts and the results thereof to summarise what issues were encountered and how difficult it would be to overcome these, so that users can understand what's going on.

The issue: linux-vserver is not sexy and hip like some of those other technologies. Maybe if vserver was a huge hulking patch, had really slow emulation, and or then more or less got killed by being aquired by a large non-open source friendly company...

http://boycottnovell.com/2009/02/25/xen-spit-it-out-strategy/

...people would start adopting it.