On 03/19/2010 12:41 AM, Bryen M. Yunashko wrote:
If we were to set up planet.opensuse.org
(and I agree
it would be nice),
we would have to relinquish at least *some* control of the planet's
server management to Novell.
But until then, planet.opensu.se is the way to go.
I'm sorry but there seems to be much confusion about how the openSUSE
project handles its infrastructure. Let me explain:
We have a bunch of projects, a lot of infrastructure, different types of
hosting and different ways of administering something.
The landing page, the wikis and the wordpress instances are hosted in
the Provo datacenter. The DNS and the mail exchange are hosted in the
Nürnberg datacenter. The reason for this is that they are very high
traffic and the face for the project so we want them to be very
reliable. Which means you put them in a datacenter where you get 24/7
support, a lot of bandwidth and good maintenance of the services and the
underlying OS. Of course the party that delivers this to you (Novell IT
in this case) will only do so if they call the shots. So we have to go
through hoops (open a bug, create a ticket or send a mail) if we want to
have something changed, but these hoops are just the price you pay for
the very good service they deliver. And they are very minimal.
Development of the services happens in a SCM and once something is done
we push Novell IT to deploy it. Everybody, no matter if you're from
Novell or not, can participate in the development of these services and
everybody can trigger the deployment. So everybody can administer these
The Build Service and all of the services connected to it (download,
software, hermes etc), openFATE, the mailinglists, users, tube, help,
connect, retro and so on are hosted in openSUSE's cage here in Nürnberg.
Here you have someone from the specific openSUSE sub-project that cares
for maintaining these services and hosts. So for instance Adrian does
OBS administration, I administer the mailinglists, Pavol cares for
connect and so on. But basically it's the same principle. Development of
these services happens in a SCM and instead of pinging someone from
Novell IT to deploy it we do it ourself. Everybody, no matter if you're
from Novell or not, can develop these services and everybody can trigger
the deployment. So everybody can administer these hosts.
Some of our services are hosted by individuals. Like the planet, live,
opensuse-community, packman, webpin, search or the software portal. Most
of these services are also developed in some SCM and here you have to
trigger the individuals to deploy something. Everybody, no matter if
you're from Novell or not, can develop these services and everybody can
trigger the deployment. So everybody can administer these hosts.
There seems to be a widely spread misunderstanding that you have to be a
Novell employee and need to have the right to deploy software or more
in general you have to be root to be able to administer something. This
is, as you can see above, not true.
Despite of that we are currently working to make exactly this possible.
For the reason to have another task in the openSUSE project where people
can do something for the project and take responsibility. This means we
try to make it possible for everybody to take up the task of software
deployment or machine administration and things like that. We open up
for contributions from sysadmins. And by the way we already have
community.o.o where you can host a service and be a sysadmin. You just
have to talk to Darix and he will figure out with you what setup fits
best for it. So as you can see we're already very far and having a
foundation will bring other benefits like paying for services
individuals host and these money related things but it won't change
anything regarding the technical aspect of machines and service
Henne Vogelsang, openSUSE.
Everybody has a plan, until they get hit.
- Mike Tyson
To unsubscribe, e-mail: opensuse-project+unsubscribe(a)opensuse.org
For additional commands, e-mail: opensuse-project+help(a)opensuse.org