Comment # 42 on bug 1188896 from Sarah Kriesch

(In reply to LTC BugProxy from comment #41)
> ------- Comment From geraldsc@de.ibm.com 2022-01-24 07:10 EDT-------
> FTR (again), the mentioned upstream commit is by no means a qualified bugfix
> for this issue. It is a fix for a theoretical race, which has been present
> for many years (basically forever), and never showed in practice. There is
> no indication that it would / should help with this issue, it is just
> "grasping at straws", as mentioned earlier here, due to lack of any good
> starting points for debugging (e.g. kernel dump or log).
> 
> I would suggest that you take the opportunity, if you can still easily
> reproduce the problem, and implement the advice from an earlier comment to
> get a kernel dump at the time when the first "bad page state" error occurs:
> 
> > The "bad page state" warning causes a taint, so panic_on_taint=5 should help
> > there. Needs an OBS admin to add that option and get a dump though, or
> > alternatively someone who can reproduce it (somewhat reliably) with
> > osc build --vm-type=kvm.

I had an equal idea in the past with log files and a discussion about that with
@Adrian. Saving dumps is an additional number on top.
Do you know about the size of such a kernel dump and perhaps additionally the
log? I thought overnight about this request.
That makes only sense if these dumps can be also removed again from the SUSE
infrastructure after an automated handover to you. If you take a look at my
last comments, then you can see that "many" such dumps could be reproduced then
on a daily basis and I have received such example dumps via Adrian in the past.

There are also some requirements for IBM side then. I know IBM security rules.
It is difficult to receive data from outside to Developer side. I had such a
data exchange with Sandy via the Box in the past. Can you ask Thomas or your
People Lead Matthias about the possibility of writing access for our systems in
the Box, that you can receive these data automated and that these data can be
removed after a successful upload on our side again?

I would say, that it would be a win-win situation for you as an IBMer, the same
as for openSUSE and SUSE.
BUT IBM should know, that it is an investment from SUSE side for all of us
then.
My requirement (in the name of the Linux Distributions Working Group) for that
is, that you speak in the same tone as here with SUSE in the future (again). It
does not make sense to invest in such features if that does not have any future
in the next years.

@Adrian: What do you say about this idea to create such dumps automatically and
upload these to the Box on IBM side and remove these data afterwards again on
the SUSE infrastructure?

@Ihno Is such a handshake ok if SUSE and openSUSE are handled again in a good
way by IBM?

P.S. Yes. I am interacting pro all Linux Distributions (incl. our Enterprise
partner) via the Open Mainframe Project.