Fedora 21 Beta blocker bug status report

[Adam Williamson]

Greetings, Fedora-folk! It's time for the first blocker status report
for F21. Actually, past time - sorry I haven't sent one of these before,
things have been moving too fast and there's been too much else to take
care of.

The Go/No-Go meeting for Fedora 21 Beta is scheduled for Thursday, which
means we need a compose ideally by tomorrow and at the latest by
Wednesday in order to have a good chance at releasing without a delay.
We currently have seven accepted blocker bugs.

Developers - fixes required
---------------------------

Developers, we need fixes for the following accepted blockers:

* https://bugzilla.redhat.com/show_bug.cgi?id=1129629 - python-blivet -
"Anaconda doesn't recognize insufficient size of /boot partition" - we
in fact traced this one out today, but the patch has not been formally
submitted to anaconda-patches@ for review, nor included in a build.

* https://bugzilla.redhat.com/show_bug.cgi?id=1099299 - systemd - "fedup
fails to upgrade F20 to F21 or later - infinite loop when starting udev"
- the expected fix for this has been written, but testing is difficult
without it being included in a compose. QA/releng are asking systemd
devs to provide a koji build and Bodhi update with the fix, and we will
pull it into the next compose for testing.

There is also a bug in proposed blocker status (due to uncertainty over
precisely how bad the impact is) which is accepted as a freeze exception
bug and which we would definitely like a fix for:

* https://bugzilla.redhat.com/show_bug.cgi?id=1146580 - fedup -
"Mislabelled /usr/sbin, /usr/bin after fedup" - this needs to be fixed
in fedup-dracut, so it has to be in the compose package set.

* https://bugzilla.redhat.com/show_bug.cgi?id=1120964 - python-blivet -
"Windows NTFS volume corrupted beyond repair during installation" - in
fact there is a proposed fix for this with an updates.img and we need
the reporter (Chris Murphy) to test it, but we do also need a
python-blivet build containing the fix (ideally after confirmation, but
at a pinch 'flying blind' may have to do) before we can compose an RC1.

* https://bugzilla.redhat.com/show_bug.cgi?id=1153816 and
https://bugzilla.redhat.com/show_bug.cgi?id=1038413 - fedup - both of
these are 'special' blockers in that they need to be fixed in the F19
and F20 fedup packages, not in F21. So, Will, we have a bit more time
flexibility on these and they don't block RC1 compose, but we do need
them fixed in F19 and F20 fedup by release day.

So what we're expecting for an RC1 build is a build of python-blivet
with fixes for #1129629 and #1120964, a build of fedup-dracut with a fix
for #1146580, and a build of systemd with a fix for #1099299.

Testers - karma required
------------------------

QA folks, we need testing and Bodhi karma for the following updates that
fix blocker bugs:

* https://admin.fedoraproject.org/updates/livecd-tools-21.3-1.fc21 ,
fixing https://bugzilla.redhat.com/show_bug.cgi?id=1154138 - you have to
compose a live image and boot it to test this, ideally someone would
compare behaviour of a live image composed with this vs. a TC4 live
image, booting on a Mac

Testers - fix testing required
------------------------------

As mentioned above, we need cmurf to test the fix for
https://bugzilla.redhat.com/show_bug.cgi?id=1120964 : an updates.img is
available in the bug.

Thanks folks! I'm hoping we can get the necessary builds done and an RC1
spun up tomorrow.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net