f29-beta-go_no_go-meeting
MINUTES

#fedora-meeting-1: F29 Beta Go/No-Go meeting

Meeting started by bcotton at 17:00:30 UTC (full logs).

Meeting summary

  1. Roll Call (bcotton, 17:00:38)
  2. Purpose of this meeting (bcotton, 17:03:12)
    1. Purpose of this meeting is to check whether or not F29 Beta is ready for shipment, according to the release criteria. (bcotton, 17:03:14)
    2. This is determined in a few ways: (bcotton, 17:03:26)
    3. 1. No remaining blocker bugs (bcotton, 17:03:27)
    4. 2. Release candidate compose is available (bcotton, 17:03:35)
    5. 3. Test matrices for Beta are fully completed (bcotton, 17:03:36)

  3. Blockers (bcotton, 17:04:27)
    1. https://qa.fedoraproject.org/blockerbugs/milestone/29/beta/buglist (bcotton, 17:04:34)
    2. https://bugzilla.redhat.com/show_bug.cgi?id=1592686 (bcotton, 17:05:56)
    3. 7 Proposed Blockers (adamw, 17:07:09)
    4. 1 Accepted Blockers (adamw, 17:07:09)
    5. starting with proposed... (adamw, 17:07:16)

  4. (1592686) the installer will hang there if inst.vncpassword boot option is added (adamw, 17:07:21)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1592686 (adamw, 17:07:21)
    2. Proposed Blocker, anaconda, NEW (adamw, 17:07:21)
    3. AGREED: 1592686 - RejectedBlocker (Beta) - this only affects passworded VNC, and the criterion doesn't require that to work, just 'VNC'. which does work (adamw, 17:12:34)

  5. (1596540) RuntimeError: C++ std::exception: Exec failed: no such table: main.trans_cmdline (adamw, 17:12:44)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1596540 (adamw, 17:12:44)
    2. Proposed Blocker, dnf, ASSIGNED (adamw, 17:12:44)
    3. AGREED: 1596540 - RejectedBlocker (Beta) - we're all sort of worried about dnf in general and a little about this bug in particular, but as lots of people *didn't* hit it and no-one has found a clear reproducer yet, we don't feel like we can reasonably count it as a blocker at this point (adamw, 17:28:58)

  6. (1625259) RuntimeError: TransactionItem not found for key (adamw, 17:29:53)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1625259 (adamw, 17:29:53)
    2. Proposed Blocker, dnf, ASSIGNED (adamw, 17:29:53)
    3. https://bugzilla.redhat.com/show_bug.cgi?id=1627534#c1 (bowlofeggs, 17:33:32)
    4. https://bugzilla.redhat.com/show_bug.cgi?id=1627534#c3 (adamw, 17:35:37)
    5. AGREED: 1625259 - RejectedBlocker (Beta) - again this is quite worrying, but with careful analysis of zbyszek's case, we still don't believe this is affecting 'normal', criteria-covered operations, only downgrades and possibly cases where the RPM database has existing duplicate entries (adamw, 17:39:01)

  7. (1628525) F29 KDE Live contains old wallpaper from F28 (adamw, 17:39:59)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1628525 (adamw, 17:39:59)
    2. Proposed Blocker, f29-backgrounds, NEW (adamw, 17:39:59)
    3. we're going to postpone discussion on this bug and whether to propose a criterion change till after we've been through the others (adamw, 17:53:07)

  8. (1626932) Error getting active session: No data available (adamw, 17:53:19)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1626932 (adamw, 17:53:19)
    2. Proposed Blocker, gnome-shell, NEW (adamw, 17:53:19)
    3. AGREED: 1626932 - RejectedBlocker (Beta) - the state of this issue with Beta RC1 as described in comment #7 clearly doesn't violate any criteria (adamw, 17:56:27)

  9. (1628497) gnome-software shows upgrade banner for Fedora 28 instead of Fedora 29 (adamw, 17:56:45)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1628497 (adamw, 17:56:45)
    2. Proposed Blocker, gnome-software, NEW (adamw, 17:56:45)
    3. AGREED: 1628497 - RejectedBlocker (Beta), AcceptedPreviousReleaseBlocker (Final) - as there are viable workarounds (use dnf, go through F28 first, hack up the cached JSON...) we believe those are sufficient for Beta, but this counts as a violation of the upgrade criterion for Final and must be fixed in F27 prior to F29 release (adamw, 18:21:22)

  10. (1628192) Fedora 29 installation cannot see a firmware RAID device. (adamw, 18:21:45)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1628192 (adamw, 18:21:45)
    2. Proposed Blocker, mdadm, NEW (adamw, 18:21:45)
    3. AGREED: 1628192 - RejectedBlocker (Beta) AcceptedFreezeException (Beta) - this is a fairly clear violation of the criteria as written (though with a workaround), but there is substantial support for either moving the firmware RAID criterion to Final or removing it entirely. On that basis this is rejected as a blocker, and we will send some criteria proposals to the list, expecting at least that the Beta requirement will no longer exist. (adamw, 18:38:32)
    4. now doing the sole accepted blocker before circling back to the KDE backgrounds thing (adamw, 18:39:00)

  11. (1628355) F29 Beta RC1 does not boot on Raspberry Pi (adamw, 18:39:09)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1628355 (adamw, 18:39:09)
    2. Accepted Blocker, spin-kickstarts, NEW (adamw, 18:39:09)
    3. this bug meant Beta RC1 32-bit ARM images do not boot on Raspberry Pi. Pi2 and Pi3 are supported platforms. Beta RC2 should fix this, but it is still composing at present and so has not been tested at all yet. (adamw, 18:41:34)
    4. circling back to KDE backgrounds blocker (adamw, 18:41:48)

  12. (1628525) F29 KDE Live contains old wallpaper from F28 (adamw, 18:42:01)
    1. https://bugzilla.redhat.com/show_bug.cgi?id=1628525 (adamw, 18:42:01)
    2. Proposed Blocker, f29-backgrounds, NEW (adamw, 18:42:01)
    3. AGREED: 1628525 - AcceptedFreezeException (Beta) - we do not have a clear vote on blocker status for this bug (the vote was +4, -3). but we have consensus that it should at least be a freeze exception. we will proceed with the meeting with its blocker status undecided (adamw, 18:53:14)

  13. Current status - tests (bcotton, 18:54:19)
    1. validation coverage is actually pretty good for RC1 (which still has the known ARM blocker). I see only a couple of missing tests (hardware RAID, 'basic audio'). There is no validation coverage at all for RC2 yet. Difference between RC1 and RC2 is minimal so we could consider most RC1 testing valid against RC2, but we should really at least smoke-test RC2 and check the ARM images boot on rasbpi (adamw, 18:56:04)

  14. Go/No-Go decision (bcotton, 18:59:06)
    1. AGREED: Fedora 29 is NO-GO (bcotton, 19:00:46)
    2. The next F29 Beta Go/No-Go meeting will be Thursday, 2018-09-20 at 1700 UTC (bcotton, 19:00:47)
    3. ACTION: bcotton to announce decision (bcotton, 19:00:53)


Meeting ended at 19:01:30 UTC (full logs).

Action items

  1. bcotton to announce decision


Action items, by person

  1. bcotton
    1. bcotton to announce decision


People present (lines said)

  1. adamw (257)
  2. contyk (58)
  3. sgallagh (54)
  4. bcotton (52)
  5. cmurf (50)
  6. nirik (49)
  7. bowlofeggs (46)
  8. mboddu (37)
  9. mattdm (27)
  10. lruzicka3 (23)
  11. zodbot (17)
  12. lruzicka (14)
  13. frantisekz (11)
  14. langdon (6)
  15. cverna (5)
  16. x3mboy (4)
  17. Southern_Gentlem (3)
  18. jforbes (2)
  19. mkolman (1)


Generated by MeetBot 0.1.4.