Re: Commercial support for fixing bugs

2010-01-07 Thread Ante Karamatić
On 07.01.2010 13:47, Alvin wrote:

> I don't care for booting under 10 seconds on a server, but I do care for
> consistent booting. Right now, there are so many things going wrong that I
> don't know where to begin reporting. Even error messages are wrong (504224),
> and without boot logging (328881) they are the only thing we got to know what
> is going wrong.

FWIW, we had a discussion about this problem at UDS. We also said 'we 
don't care about 10 seconds boot' :). I'm sure it will get fixed...

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam


Re: Commercial support for fixing bugs

2010-01-07 Thread Etienne Goyer
I work for Canonical Global Services and Support.  A couple of comments:


Alvin wrote:
> I contacted Canonical a while ago, and a support contract will give you the 
> right to 10 support cases per annum.

That changed recently, there's no more case limit.  However, support
contract are still sold on a per-machine basis.


> So, the bugs I mentioned in my previous post would receive higher priority. I 
> only wonder whether this specific bug would be solved in karmic, because it 
> needs another solution than the one that will be used for lucid. After all, 
> is 
> is already fixed.

If you report this issue as a support case, it will be triaged and
investigated by a support analyst.  If appropriate, it might be
escalated internally to developers.

It is important to note, though, that we cannot commit to fixing a
specific bug within the scope of a support contract.  The purpose of a
support contract is to get help in resolving problems.  That help can
take the form of a bug fix where appropriate, or it can be something
else, such as a workaround.  What we aim for is to ensure the customer's
business continuity.


> So, will taking a support contract guarantee bootable machines in the next 4 
> months?

Yes.  Canonical Support will provide a solution to your problem.  It may
or may not take the form of a fix for bug #470776, but we will make sure
the Ubuntu systems under support contract remain reliable and perform
according to expectations.


If you have further question, or need clarification, on the scope of
Canonical Support, I suggest you contact someone at Canonical directly.
   Public mailing list may not be the best place to ask.


Best regards,

-- 
Etienne Goyer
Technical Account Manager - Canonical Ltd
Ubuntu Certified Instructor-   LPIC-3


-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam


Re: Commercial support for fixing bugs

2010-01-07 Thread Alvin
On Tuesday 05 January 2010 18:12:39 Ante Karamatić wrote:
> On 05.01.2010 16:50, Alvin wrote:
> > So, the bugs I mentioned in my previous post would receive higher
> > priority. I only wonder whether this specific bug would be solved in
> > karmic, because it needs another solution than the one that will be used
> > for lucid. After all, is is already fixed.
> 
> 
> 
> I, and rest of the server people, were hit by this or some other
> upstart/mountall/whatever bug that renders Karmic unusable in NFS
> environment.
> 
> [...]
> 
> 

Let's call that a workaround. There are other possibilities to work around 
this, but I don't like doing a lot of stuff in rc.local that SHOULD be done by 
the boot mechanism. Currently there are 3 things that need to be done after 
boot.

- mount NFS shares (bug 470776)
- restart samba (bug 462169)
- restart libvirtd (bug 491273)

Only the samba bug is fixed in karmic proposed. This bug is mentioned in the 
release notes. Shouldn't there be a warning in those release notes about NFS 
too?

I don't care for booting under 10 seconds on a server, but I do care for 
consistent booting. Right now, there are so many things going wrong that I 
don't know where to begin reporting. Even error messages are wrong (504224), 
and without boot logging (328881) they are the only thing we got to know what 
is going wrong.

I've had a talk about all this with my colleague, and we will probably try a 
support contract and see how it turns out.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Commercial support for fixing bugs

2010-01-05 Thread Ante Karamatić
On 05.01.2010 16:50, Alvin wrote:

> So, the bugs I mentioned in my previous post would receive higher priority. I
> only wonder whether this specific bug would be solved in karmic, because it
> needs another solution than the one that will be used for lucid. After all, is
> is already fixed.



I, and rest of the server people, were hit by this or some other 
upstart/mountall/whatever bug that renders Karmic unusable in NFS 
environment.

However, there's an easy workaround for this situation. Mount /home 
outside of /home. For example, mount it in /media/home and then add:

mount -o bind /media/home /home

in /etc/rc.local

Machine will boot (it fails to boot cause IP/portmap/something_else is 
up after "mounting network file system process") and rc.local will run 
before GDM is up anyway.



-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam


Commercial support for fixing bugs

2010-01-05 Thread Alvin
I'd like to upgrade a lot of machines under my care to karmic, because of 
different reasons, but there is a bug that 'stops the show'.

retry remote devices when parent is ready after SIGUSR1
https://bugs.launchpad.net/bugs/470776

In short, NFS mounts can not be trusted. Sometimes, it is needed to go into a 
recovery shell, mount manually, and continue booting (with some other errors). 
This is all good if you can manage the console, but for "human beings" who 
will be using some of these systems, it's just too much to ask.

(I mentioned this and other bugs in a previous post (Earth Computing), but 
there were only personal reactions.)

The good news is, that this is fixed in lucid. The bad news is: karmic will 
remain broken.

I contacted Canonical a while ago, and a support contract will give you the 
right to 10 support cases per annum. I'm perfectly willing to pay for support, 
if only to give some people the chance to be payed for the work they do on 
Ubuntu.

The situation with bugs (I quote):
- Support customer bugs receive higher priority by Canonical developers.
- Bugs reported as support cases do not actually count against your allocation 
of support tickets for that subscription.
- Bugs that require assistance from upstream projects are also managed 
directly by Canonical, increasing the likelihood, and speed of successful 
resolution.

So, the bugs I mentioned in my previous post would receive higher priority. I 
only wonder whether this specific bug would be solved in karmic, because it 
needs another solution than the one that will be used for lucid. After all, is 
is already fixed.

So, will taking a support contract guarantee bootable machines in the next 4 
months?

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam