At the Fedora 16 Beta Go/No-Go Meeting yesterday, the Fedora 16 Beta
release was declared GOLD! Fedora 16 beta will be released on Tuesday,
October 4, 2011.
Major thanks to everyone who made this happen. There is no way that we
could have gotten all this done without everyone's hard work and
assistance.
Tim
Minutes:
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-29/f16_beta_go_no_g…
Minutes (text):
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-29/f16_beta_go_no_g…
Log:
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-29/f16_beta_go_no_g…
====================================================
#fedora-meeting: F15 Beta Go No Go Meeting Round 2.5
====================================================
Meeting summary
---------------
* Who's Here? (tflink, 21:03:42)
* LINK:
https://bugzilla.redhat.com/showdependencytree.cgi?id=713564&hide_resolved=1
(adamw, 21:09:08)
* why are we here (tflink, 21:09:50)
* We're here to see whether or not the beta release criteria have been
met (tflink, 21:10:22)
* LINK: http://fedoraproject.org/wiki/Fedora_16_Beta_Release_Criteria
(tflink, 21:10:30)
* current release blockers (tflink, 21:10:44)
* LINK:
https://bugzilla.redhat.com/showdependencytree.cgi?id=713564&hide_resolved=1
(tflink, 21:10:53)
* that list doesn't take accepted/proposed into account (tflink,
21:11:02)
* (725185) grubby doesn't add the initrd line at the kernel update
(tflink, 21:11:57)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=725185 (tflink,
21:12:04)
* AGREED: - 725185 - RejectedBlocker - This is a problem, but as long
as the next kernel update requires this version of grubby, things
should be OK and this potential issue can be fixed with an update
post-beta. (tflink, 21:20:20)
* (731529) grub making uefi calls without aligning stack pointer
(tflink, 21:20:39)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=731529 (tflink,
21:20:45)
* AcceptedBlocker, ON_QA (tflink, 21:21:53)
* AGREED: - 731529 - Move to VERIFIED - All of the EFI tests are
pretty much working, other issues would have likely surfaced by now
(tflink, 21:23:25)
* (737731) Bootloader is left in F15 configuration when preupgrading to
F16 (tflink, 21:23:49)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=737731 (tflink,
21:23:54)
* AcceptedBlocker, ASSIGNED (tflink, 21:24:02)
* AGREED: - 737731 - This needs to be fixed in F15 PreUpgrade and
doesn't affect Beta compose. It doesn't appear to be in progress,
need to revisit in the next couple of days. (tflink, 21:26:52)
* (739253) unable to shut down from gdm greeter (tflink, 21:27:13)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=739253 (tflink,
21:27:19)
* AcceptedBlocker, MODIFIED (tflink, 21:27:25)
* AGREED: - 739253 - Move to VERIFIED - this has been fixed to the
point that we expected - no power option in shell that fails. The
power option will be restored in gnome-shell updates (tflink,
21:29:26)
* (742207) No usable bootloader option during a text mode f15->f16
upgrade (tflink, 21:30:04)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=742207 (tflink,
21:30:12)
* Proposed Blocker, NEW (tflink, 21:30:22)
* ACTION: adamw or tflink - Document workaround for 742207 (tflink,
21:35:10)
* AGREED: - 742207 - RejectedBlocker, CommonBugs - This doesn't
directly hit any beta release criterion, isn't the most common
upgrade path and can be worked around. The workaround needs to be
documented (tflink, 21:35:22)
* (742226) /sbin/grub2-probe: error: cannot find a GRUB drive for
/dev/mapper/nvidia_cjfffajep2 (tflink, 21:36:37)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=742226 (tflink,
21:36:44)
* Proposed Blocker, NEW (tflink, 21:36:51)
* AGREED: - 742226 - RejectedBlocker CommonBugs- BIOS Raid does work
for some controllers but not all. Re-propose as final blocker and
document for beta. (tflink, 21:44:21)
* Test Coverage (tflink, 21:46:06)
* LINK:
https://fedoraproject.org/wiki/QA:Testcase_Kickstart_File_Path_Ks_Cfg
(adamw, 21:47:27)
* LINK:
https://fedoraproject.org/wiki/QA:Testcase_Install_to_Hardware_RAID
(adamw, 21:47:36)
* AGREED: - only one exotic kickstart delivery method is missing from
test coverage, and we provisionally suggest the stranger kickstart
delivery methods should not be beta critical; therefore deeming test
coverage effectively complete, as we would not block even if that
test failed (adamw, 21:57:21)
* Is Fedora 16 Beta Ready to Go? (tflink, 21:59:12)
* QA is go for F16 Beta (tflink, 21:59:48)
* RelEng is go (tflink, 22:00:10)
* Devel is go (tflink, 22:00:41)
* AGREED: - Fedora 16 Beta RC4 is accepted as beta (tflink, 22:01:35)
* Open Discussion (tflink, 22:03:25)
Meeting ended at 22:04:55 UTC.
Action Items
------------
* adamw or tflink - Document workaround for 742207
Action Items, by person
-----------------------
* adamw
* adamw or tflink - Document workaround for 742207
* tflink
* adamw or tflink - Document workaround for 742207
* **UNASSIGNED**
* (none)
People Present (lines said)
---------------------------
* adamw (145)
* tflink (101)
* nirik (37)
* pjones (31)
* jsmith (26)
* zodbot (5)
* jdulaney (4)
* dgilmore (4)
* Jsmith (3)
* smooge (1)
* satellit_ (1)
Join us on irc.freenode.net in #fedora-meeting for the second half of
this important meeting.
Thanks to the tireless* efforts of our outstanding QA folks, testing on
RC4, created late yesterday, has been ongoing all night long. We'll be
having the second-and-a-half round of our F16 Beta Go/No-Go TODAY. The
first half of this meeting occurred yesterday[1][2], and it was decided
that we would make a stretched effort to have another RC including fixes
by midnight last night that looked viable, which actually happened. (Had
it not been viable, we would have called it No-Go by midnight.)
Thursday, September 29, 2011 @21:00 UTC (17:00 EDT/14:00 PDT)
"Before each public release Development, QA and Release Engineering
meet to determine if the release criteria are met for a particular
release. This meeting is called the: Go/No-Go Meeting."
"Verifying that the Release criteria are met is the responsibility of
the QA Team."
For more details about this meeting see:
https://fedoraproject.org/wiki/Go_No_Go_Meeting
In the meantime, keep an eye on the Fedora 16 Beta Blocker list:
https://fedoraproject.org/wiki/Current_Release_Blockers
-Robyn
* Though I know you guys are *very* tired, and your work is MONDO
appreciated.
[1] Minutes:
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-28/f16_beta_go_no_g…
[2] Logs:
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-28/f16_beta_go_no_g…
Meeting Minutes from Go/No-Go Round 2.0:
====================================================
#fedora-meeting: F16 Beta Go No Go Meeting Round two
====================================================
Meeting started by rbergeron at 21:00:24 UTC. The full logs are
available at
http://meetbot.fedoraproject.org/fedora-meeting/2011-09-28/f16_beta_go_no_g…
.
Meeting summary
---------------
* Who is here? (rbergeron, 21:00:53)
* present: jdulaney, nirik, rbergero (rbergeron, 21:01:03)
* Why are we here? (rbergeron, 21:01:47)
* Dearly beloved, we are gathered here today to see whether or not the
beta release criteria have been met. (rbergeron, 21:02:28)
* LINK: http://fedoraproject.org/wiki/Fedora_16_Beta_Release_Criteria
(rbergeron, 21:02:43)
* Current release blockers (rbergeron, 21:05:29)
* LINK: https://fedoraproject.org/wiki/Current_Release_Blockers
(rbergeron, 21:05:38)
* http://bugzilla.redhat.com/show_bug.cgi?id=727522 (adamw, 21:08:57)
* AGREED: 727522 is really fixed, the apparent rc2/rc3 reproducers
were pilot error (adamw, 21:11:50)
* http://bugzilla.redhat.com/show_bug.cgi?id=737731 (adamw, 21:11:53)
* 737731 doesn't block f16 image creation, will get fixed by an F15
update. (rbergeron, 21:13:06)
* http://bugzilla.redhat.com/show_bug.cgi?id=741655 (adamw, 21:13:45)
* AGREED: 741655 is not a beta blocker as it's a crazy custom
partition layout, moved to final (adamw, 21:15:51)
* http://bugzilla.redhat.com/show_bug.cgi?id=741781 (adamw, 21:16:14)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=741781#c19
provides an explanation of the two bugs, if you want a quickish
catch-up (adamw, 21:18:29)
* AGREED: If we have a built anaconda and RC4 by midnight pacific
tonight, opt to do go/no-go tomorrow. Otherwise, slip one week.
(rbergeron, 21:37:28)
* if we don't have what is needed by midnight pacific, Robyn will send
out the No-Go Mail. (rbergeron, 21:39:32)
* ACTION: nirik will be coordinating the RC; he needs to be told when
anaconda/grub is ready. (rbergeron, 21:39:48)
* For status updates/questions, please check in #fedora-qa
(rbergeron, 21:41:42)
* https://bugzilla.redhat.com/show_bug.cgi?id=741817 (adamw, 21:43:10)
* AGREED: 741817 not a blocker as long as it leaves f15 install
bootable, adamw will check that (adamw, 21:53:56)
* Re-No/Go Meeting time (rbergeron, 21:54:13)
* AGREED: will meet tomorrow, same bat time, same bat channel, to do
final go/no-go decision. (rbergeron, 21:56:32)
* robyn will be out tomorrow (rbergeron, 21:56:44)
Meeting ended at 21:58:19 UTC.
Action Items
------------
* nirik will be coordinating the RC; he needs to be told when
anaconda/grub is ready.
Action Items, by person
-----------------------
* nirik
* nirik will be coordinating the RC; he needs to be told when
anaconda/grub is ready.
* **UNASSIGNED**
* (none)
People Present (lines said)
---------------------------
* rbergeron (142)
* adamw (125)
* jdulaney (28)
* nirik (27)
* dgilmore (26)
* jsmith (24)
* pjones (11)
* VileGent (4)
* zodbot (4)
* tflink (3)
* hno (3)
* brunowolff (1)
* tk009 (1)
Generated by `MeetBot`_ 0.1.4
.. _`MeetBot`: http://wiki.debian.org/MeetBot
Hey, folks. So, everyone clear their schedules for the next 18 hours!
We have, we think, fixed up all the EFI installation issues. At the
go/no-go meeting today, all groups agreed to delay the go/no-go decision
till tomorrow evening (Eastern time) to give a chance at getting RC4
built and validated and hence avoiding a further slip of the release,
which we really didn't want to do.
RC4 should be landing within the next few hours, unless something goes
terribly wrong. At that point, because the fixes involved poking
anaconda, we're going to need to try and re-do the entire set of install
validation tests, if at all possible. Desktop and base do not really
need to be re-tested, as they won't have changed at all from RC3.
I'm gonna be testing the whole time from when RC4 lands to the deadline,
but if everyone else can please pitch and complete some of the install
tests that would be really helpful. It would especially help if people
can help test the slightly odd hardware cases, like all the RAID
permutations, and SCSI and PATA installs.
We can try to co-ordinate testing through #fedora-qa, I'll be there the
whole time, to try and make sure people don't test the same things
(which is usually great, but for this we should just focus on making
sure everything gets tested at least once).
Thanks a lot everyone!
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
Join us on irc.freenode.net #fedora-meeting for this important meeting.
Wednesday, September 28, 2011 @21:00 UTC (17:00 EDT/14:00 PDT)
"Before each public release Development, QA and Release Engineering
meet to determine if the release criteria are met for a particular
release. This meeting is called the: Go/No-Go Meeting."
"Verifying that the Release criteria are met is the responsibility of
the QA Team."
For more details about this meeting see:
https://fedoraproject.org/wiki/Go_No_Go_Meeting
In the meantime, keep an eye on the Fedora 16 Beta Blocker list:
https://fedoraproject.org/wiki/Current_Release_Blockers
-Robyn
Hey, everyone, while we're still busy validating Beta, there's a Test
Day going on today/tomorrow, Monday 2011-09-26, too:
https://fedoraproject.org/wiki/Test_Day:2011-09-26_ABRT
Looking at ABRT, and particularly its integration with libreport (which
some of you may remember has caused a few problems already this cycle!)
As always, the event is in #fedora-test-day in Freenode IRC, and all the
test instructions are on the Wiki. There should be a good turnout of
developers, so it's a great chance to help shake out any remaining abrt
bugs for Fedora 16.
Please help out if you have a chance! Thanks.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
Hey, folks. So, as most of you probably know, one of the big bugs that's
been holding up the Beta is
https://bugzilla.redhat.com/show_bug.cgi?id=738964 - various issues
related to bootloader installation with the wrinkles of grub2 and GPT
disk labels.
There are now not one but *two* updates images with test fixes for the
bugs:
http://dlehman.fedorapeople.org/updates-738964.6.imghttp://dlehman.fedorapeople.org/updates-738964.6-nohammer.img
The second one improves the anaconda code which excludes the device
anaconda is being installed from (particularly if it's a USB stick) from
consideration as an install or bootloader target device, and also tries
much harder to honor the user's choice of the device where the
bootloader should be installed on all the non-custom-partitioning paths.
Just these two fixes should address the most urgent cases of this bug.
The first one also tries to create a BIOS boot partition on any GPT
labelled disk present - whether formatted by anaconda or not - which has
enough free space and which does not currently have one, if the system
is BIOS-based. This should act as a 'big hammer' to make sure you can
pretty much always install the bootloader wherever you like, but
obviously it's a pretty big and possibly problematic change.
We really need testing with both images, to establish a couple of
things:
1) if you've had problems with the bootloader installing somewhere you
didn't expect in Beta RC1 and previous builds, please test whether your
issue is fixed (it's best if you can test with each updates.img and
report on both, but if you can only test one, that's fine too)
2) if either image has bugs that don't exist in Beta RC1: as much
testing as possible, with any kind of installation, is useful here. Just
try an install with either updates.img, and if you hit a problem,
confirm that it doesn't exist in Beta RC1, and add a comment to 738964.
Particularly interesting configurations would be installing to a system
which already has one completely full GPT-labelled disk that you choose
not to reformat, installing to a system with an existing GPT-labelled
disk with a BIOS boot partition that you choose not to reformat,
installing to a system with an existing GPT-labelled disk *without* a
BIOS boot partition that you choose not to reformat, and any kind of
installation at all with multiple disks (hard disks or USB sticks)
connected during install - no matter which of the disks you choose to
use or not use, and installs from both DVD and live written to a USB
stick.
Note that we expect the -nohammer version to still be 'buggy' in the
case where you choose manual partitioning with multiple disks present
and do not create a BIOS boot partition on the target disk; it won't
create one for you or require you to create one, and when you get to the
bootloader location choice (which comes later in custom partitioning
than in other paths), the target drive's MBR will not be an available
install location. The 'hammer' version should 'fix' this case.
To use an updates.img, just pass the updates= parameter to the kernel
when booting. So boot with Beta RC1, hit 'tab' at the bootloader screen,
and type 'updates=http://dlehman.fedorapeople.org/updates-738964.6.img'
or
'updates=http://dlehman.fedorapeople.org/updates-738964.6-nohammer.img'
and then proceed to install as normal. This works for both the regular
and live installers.
Thanks everyone! We really need to get testing data in and do an
anaconda build today so we can get RC2 out tonight.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net