Proposed validation test case: root on LVM thinp

[Adam Williamson]

On Sat, 2013-12-14 at 00:49 -0700, Chris Murphy wrote:
> On Dec 13, 2013, at 5:49 PM, Adam Williamson <awilliam at redhat.com> wrote:
> 
> > On Fri, 2013-12-13 at 15:49 -0700, Chris Murphy wrote:
> >> 
> >> Yeah I'm holding off on a major f bomb email until I understand what
> >> happened and exactly when this broke. If you have some insight on that
> >> to save me regression testing, that would be helpful.
> > 
> > Already looked into it. it was broken by
> > dracut-034-64.git20131205.fc20.x86_64 , which landed in either TC5 or
> > RC1, I don't recall which, but either way, very late.
> 
> Install from live desktop TC5 boots: kernel 3.11.10-300 and dracut 034-19.git20131021.
> 
> Updating kernel only to 3.11.10-301, still bootable.
> 
> Updating dracut to 034-64.git120131205 and rebuilding initramfs with dracut -f, fails to reboot.

Yeah, that's precisely what I'd expect. So it went into RC1.

You could say we (and mostly 'I') did the wrong thing by not just flat
refusing to take the build they offered and requiring them to provide us
with a build that only fixed the blockers and FEs, I guess. That's my
bad. But it's pretty hard trying to hold that line with devs, and I
don't want to cause bad blood with them. I think this case makes it
clear, though, that in future we'd better just be more forceful about
it.

(I also kinda expected that we'd do RC1 much quicker after TC5, and
hence have a bit more time to test stuff with the new dracut - it kept
seeming like it was a day away, then receding further...)
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net