ReleaseEngineering/Meetings/2008-may-05

From FedoraProject

Jump to: navigation, search

Contents

Fedora Release Engineering Meeting :: Monday 2008-05-05

F9 Release Candidate

  • A few things holding up the release candidate
  • texlive
  • X crashing on evdev removal during installer (kvms)
  • packagekit that just got built.
  • emacs has broken alternatives and is still on the blocker list
  • conflict between rsyslog and syslog-ng

Posting F9 Release Candiate

  • Once we have RC spun and are testing it we should pre-process the pending F9 updates and get them pushed into directories
  • We should also have tickets filed for releases tasks that we need to make sure happen like export approval filing, etc.

Rawhide Switch to F10

  • BugZappers want to cooridnate with bugzilla versioning and rawhide rebase
  • Most likely rawhide composes starting on May 8th or 9th will become F10
  • f13 to keep poelcat updated

IRC Transcript

-!- f13 changed the topic of #fedora-meeting to: Fedora Release Engineering - Roll Call<a href="#t01:00" class="time">01:00</a>
f13ping: notting jeremy jwb spot lmacken rdieter wwoods warren<a href="#t01:00" class="time">01:00</a>
jwbsort of here<a href="#t01:00" class="time">01:00</a>
rdieteryo<a href="#t01:00" class="time">01:00</a>
* notting is here<a href="#t01:00" class="time">01:00</a>
* poelcat here<a href="#t01:02" class="time">01:02</a>
* lmacken here<a href="#t01:02" class="time">01:02</a>
f13ok, lets get started.<a href="#t01:03" class="time">01:03</a>
-!- f13 changed the topic of #fedora-meeting to: Fedora Releng - Release Candidate F9<a href="#t01:03" class="time">01:03</a>
f13We know of a few things holding up the release candidate.  texlive, X crashing on evdev removal during installer (kvms), and a packagekit that just got built.<a href="#t01:04" class="time">01:04</a>
nottingpk is tagged<a href="#t01:04" class="time">01:04</a>
nottingf13: emacs having broken alternatives is still on the blocker list<a href="#t01:04" class="time">01:04</a>
f13ah<a href="#t01:04" class="time">01:04</a>
f13I thought that got fixed.  Guess not<a href="#t01:04" class="time">01:04</a>
warrenis the PK build related to the GPG problems?<a href="#t01:04" class="time">01:04</a>
nottingno<a href="#t01:04" class="time">01:04</a>
nottingit's a traceback in the yum backend<a href="#t01:05" class="time">01:05</a>
spoti can't reproduce that emacs failure<a href="#t01:05" class="time">01:05</a>
spotat all<a href="#t01:05" class="time">01:05</a>
spotjrb says he can, but he's not sure how he got his system into that state<a href="#t01:05" class="time">01:05</a>
* wwoods testing stuff<a href="#t01:05" class="time">01:05</a>
spotsomething about starting at F8 and updating<a href="#t01:05" class="time">01:05</a>
wwoodsX-crashing-on-evdev-removal is not all machines<a href="#t01:05" class="time">01:05</a>
wwoodsI can't produce that bug<a href="#t01:06" class="time">01:06</a>
f13wwoods: you're likely not getting detected as evdev<a href="#t01:06" class="time">01:06</a>
wwoodsright<a href="#t01:06" class="time">01:06</a>
nottingf13: have you successfully cornered ajax?<a href="#t01:06" class="time">01:06</a>
f13notting: was waiting for him to finish lunch<a href="#t01:06" class="time">01:06</a>
f13and now I need to go find where he actually sits.<a href="#t01:06" class="time">01:06</a>
wwoodsit should only happen on machines with devices that are keyboard *and* mouse combos, *and* you have to insert/remove a keyboard/mouse<a href="#t01:06" class="time">01:06</a>
nottingwwoods: hm, would a ps/2 converter count?<a href="#t01:07" class="time">01:07</a>
f13notting: that may pop as evdev<a href="#t01:07" class="time">01:07</a>
warrenI have a two PS/2 to USB converter at home<a href="#t01:07" class="time">01:07</a>
f13wwoods: this isn't a combo, it's a real mouse, usb, plugged into the keyboard, also usb<a href="#t01:07" class="time">01:07</a>
nottingand is this anaconda only or post-install too?<a href="#t01:07" class="time">01:07</a>
f13notting: I've only see it during anaconda<a href="#t01:08" class="time">01:08</a>
f13but I don't spend much time out of anaconda in this setup<a href="#t01:08" class="time">01:08</a>
f13and from earlier fun with evdev, we know htat the combo detection is a little bad, there are lots of systems that get picked up as combos when they aren't<a href="#t01:08" class="time">01:08</a>
f13this can also lead to bad keyboard shortcuts post-install if the detection is wrong<a href="#t01:09" class="time">01:09</a>
f13ajax had said something about neutering evdev to not claim combo devices<a href="#t01:09" class="time">01:09</a>
nottingyes... but... not a blocker. x crashing may be a blocker<a href="#t01:09" class="time">01:09</a>
* spot will try to reproduce that emacs blocker again today<a href="#t01:09" class="time">01:09</a>
f13other than those, I think we're in OK shape.  We're working our way through the testing matrix<a href="#t01:11" class="time">01:11</a>
f13wwoods: would you consider the syslog-ng and rsyslog conflict a blocker?<a href="#t01:11" class="time">01:11</a>
warrenf13: is the fix simple?<a href="#t01:11" class="time">01:11</a>
notting#?<a href="#t01:12" class="time">01:12</a>
f13I don't know if wwoods filed a bug on it.<a href="#t01:12" class="time">01:12</a>
f13warren: don't know.<a href="#t01:12" class="time">01:12</a>
* notting says 'probably not, as there are certainly multilib conflicts somewhere we haven't noticed'<a href="#t01:12" class="time">01:12</a>
f13Ok.<a href="#t01:13" class="time">01:13</a>
f13so ideally we'll get the fixes in today for the things that are blockers, and spin the real RC from tomorrow's rawhide.<a href="#t01:14" class="time">01:14</a>
wwoodsI'm doing a small-e "everything" install from the x86_64 DVD right now<a href="#t01:14" class="time">01:14</a>
f13also I can start pre-syncing content to PHX since everything is signed and there are going to be minimal changes<a href="#t01:14" class="time">01:14</a>
wwoodsit's <a href="https://bugzilla.redhat.com/show_bug.cgi?id=441664">https://bugzilla.redhat.com/show_bug.cgi?id=441664</a> essentially<a href="#t01:14" class="time">01:14</a>
buggbotBug 441664: low, low, ---, Peter Vrabec, NEW , rsyslog should explicitly conflict with syslog-ng<a href="#t01:14" class="time">01:14</a>
wwoodsthey both ship /etc/logrotate.d/syslog<a href="#t01:14" class="time">01:14</a>
wwoodswhich, wtf<a href="#t01:14" class="time">01:14</a>
f13explicitly conflict is verboten<a href="#t01:14" class="time">01:14</a>
f13maybe they need to be taught about alternatives<a href="#t01:14" class="time">01:14</a>
wwoodswhy doesn't rsyslog just ship /etc/logrotate.d/rsyslog<a href="#t01:15" class="time">01:15</a>
f13wwoods: or that<a href="#t01:15" class="time">01:15</a>
nottingwwoods: i'm inclined to leave it, it's not a regression, right?<a href="#t01:16" class="time">01:16</a>
wwoodsno, don't think it's a regression<a href="#t01:16" class="time">01:16</a>
wwoodsstupid, yes. probably a simple fix, yes. but not a huge problem either<a href="#t01:16" class="time">01:16</a>
wwoodssolution: don't install both packages<a href="#t01:17" class="time">01:17</a>
wwoodsI guess we could also fix it by dropping syslog-ng from the media<a href="#t01:17" class="time">01:17</a>
nottingit's on the media????<a href="#t01:17" class="time">01:17</a>
warrenthat doesn't belong there<a href="#t01:17" class="time">01:17</a>
nottingit shouldn't be<a href="#t01:17" class="time">01:17</a>
wwoodsyeah, that's the whole problem<a href="#t01:17" class="time">01:17</a>
nottingf13: can you fix *that*?<a href="#t01:17" class="time">01:17</a>
wwoodsinstalling everything in comps on the media (i.e. checking every "install all optional packages" box for everything in the installer)<a href="#t01:17" class="time">01:17</a>
wwoodsgives you "file conflict" crashy badness<a href="#t01:18" class="time">01:18</a>
nottingwwoods: from reading it, the f8 packages ATM have an explicit Conflicts:. so i'd rather not rush a special fix for f9<a href="#t01:18" class="time">01:18</a>
wwoodssolution: uncheck syslog-ng<a href="#t01:18" class="time">01:18</a>
wwoodsbetter solution: don't put it on the stupid media<a href="#t01:18" class="time">01:18</a>
f13whoh<a href="#t01:18" class="time">01:18</a>
wwoodsshipping a package known to conflict with something in @base seems obviously bad<a href="#t01:19" class="time">01:19</a>
metherwhy did we put it on the media before?<a href="#t01:19" class="time">01:19</a>
f13need to find out why it's on the media<a href="#t01:19" class="time">01:19</a>
nottingf13: this is presumably b/c of pungi pulling all providers of 'syslog'?<a href="#t01:19" class="time">01:19</a>
f13mether: we didn't explicitly<a href="#t01:19" class="time">01:19</a>
f13notting: oh yeah, I bet that's it.<a href="#t01:19" class="time">01:19</a>
f13we can just drop a -rsyslog in the manifest<a href="#t01:19" class="time">01:19</a>
nottingf13: otherway<a href="#t01:19" class="time">01:19</a>
f13er yea<a href="#t01:19" class="time">01:19</a>
f13sorry<a href="#t01:19" class="time">01:19</a>
notting+1 to -syslog-ng in the manifest<a href="#t01:20" class="time">01:20</a>
warrentotally +1<a href="#t01:20" class="time">01:20</a>
f13I'll toss that into pungi today and ask for a tag request.<a href="#t01:20" class="time">01:20</a>
f13incidentally that'll require a bumped fedora-release too<a href="#t01:20" class="time">01:20</a>
f13(since we ship the config there as well)<a href="#t01:20" class="time">01:20</a>
warrennot a big loss<a href="#t01:21" class="time">01:21</a>
nottingf13: tangent: did the ppc compose box survive the move?<a href="#t01:21" class="time">01:21</a>
f13notting: hrm, I haven't seen clumens in yet<a href="#t01:21" class="time">01:21</a>
f13I could hook it up in my cube for now.<a href="#t01:21" class="time">01:21</a>
wwoodsthought he was on a plane?<a href="#t01:21" class="time">01:21</a>
f13still?<a href="#t01:22" class="time">01:22</a>
f13long plane ride<a href="#t01:22" class="time">01:22</a>
wwoods      <packagereq type="optional">syslog-ng</packagereq><a href="#t01:22" class="time">01:22</a>
wwoodsyeah it's in comps?<a href="#t01:22" class="time">01:22</a>
wwoodsso<a href="#t01:22" class="time">01:22</a>
f13wwoods: optional though<a href="#t01:22" class="time">01:22</a>
f13wwoods: we don't pull optional packages into the media by default, only specific ones/groups<a href="#t01:22" class="time">01:22</a>
nottingf13: aside from that, are there any other infrastructural issues preventing a rc compose?<a href="#t01:22" class="time">01:22</a>
wwoodsis -syslog-ng sufficient?<a href="#t01:22" class="time">01:22</a>
wwoods'kay<a href="#t01:22" class="time">01:22</a>
f13this is just a case of "give me everything that provides 'syslog'"<a href="#t01:22" class="time">01:22</a>
f13wwoods: should be.<a href="#t01:22" class="time">01:22</a>
f13notting: none that I know of.<a href="#t01:23" class="time">01:23</a>
f13jwb: ping; have you done any ppc composes lately or installs from rawhide?<a href="#t01:23" class="time">01:23</a>
warrenf13: is a newer tree coming on reducto soon?<a href="#t01:23" class="time">01:23</a>
jwbf13, i was supposed to have a machine to do so this week.  no idea if that will actually happen<a href="#t01:23" class="time">01:23</a>
jwb(so no)<a href="#t01:23" class="time">01:23</a>
jwbalso, i've been out quite a bit over the last 2 weeks.  family issues<a href="#t01:23" class="time">01:23</a>
f13warren: there has been only 1 package change since the last reducto tree.<a href="#t01:24" class="time">01:24</a>
f13jwb: ok.<a href="#t01:25" class="time">01:25</a>
warrenah, ok<a href="#t01:25" class="time">01:25</a>
f13we'll fly a bit blind with ppc then.<a href="#t01:25" class="time">01:25</a>
f13I no longer have machines to test with<a href="#t01:25" class="time">01:25</a>
jwbf13, understandable.  we fixed the dvd size issue, yes?<a href="#t01:25" class="time">01:25</a>
f13I do believe so<a href="#t01:25" class="time">01:25</a>
jwbk.  we should pester dwmw2_gone<a href="#t01:25" class="time">01:25</a>
f13those excludes are in the pungi package now.  For reals.<a href="#t01:25" class="time">01:25</a>
jwbfo shizzle<a href="#t01:25" class="time">01:25</a>
f13alright, anything else regarding the release candidate?<a href="#t01:26" class="time">01:26</a>
warrenwwoods: <a href="http://fedoraproject.org/wiki/QA/TestResults/Fedora9Install/FinalRelease">http://fedoraproject.org/wiki/QA/TestResults/Fedora9Install/FinalRelease</a>  how did people on the outside get the RC in order to test this?<a href="#t01:26" class="time">01:26</a>
nottingf13: one? there's PK and allegro<a href="#t01:26" class="time">01:26</a>
notting(at least)<a href="#t01:26" class="time">01:26</a>
f13warren: they used rawhide, which is acceptable.<a href="#t01:26" class="time">01:26</a>
warrenoh<a href="#t01:26" class="time">01:26</a>
f13notting: those didn't make today's rawhide<a href="#t01:27" class="time">01:27</a>
nottingf13: sorry, confusing last tree/rawhide/post-rawhide-today<a href="#t01:27" class="time">01:27</a>
f13notting: I meant changes in rawhide.  Pulling packages down and making a local repo is possible, manually figuring out what should be multilib.<a href="#t01:27" class="time">01:27</a>
wwoodswarren: presumably they used rawhide, or built their own images<a href="#t01:27" class="time">01:27</a>
nottingf13: ok, have box with ps/2 keyboard, usb keyboard, usb mouse, and two ps/2->usb converters attached. :)<a href="#t01:28" class="time">01:28</a>
warrenInstallSourceHardDrive being broken is a little sad<a href="#t01:28" class="time">01:28</a>
f13warren: some people do use pungi locally to create media too, which is also acceptable/encouraged.<a href="#t01:28" class="time">01:28</a>
f13notting: scary<a href="#t01:28" class="time">01:28</a>
f13warren: yeah, but without a clumens, I'm not sure we'll see a fix for that soon<a href="#t01:28" class="time">01:28</a>
-!- f13 changed the topic of #fedora-meeting to: Fedora releng - post-release candidate<a href="#t01:29" class="time">01:29</a>
warrenwhere did clumens go?<a href="#t01:29" class="time">01:29</a>
f13once we have RC spun and are testing it, we should pre-process the pending F9 updates and get them pushed into directories to make sure that's all hot to trot<a href="#t01:29" class="time">01:29</a>
f13warren: I don't know<a href="#t01:29" class="time">01:29</a>
* warren tries it<a href="#t01:30" class="time">01:30</a>
f13also, we should have a ticket somewhere for release tasks, or af ew tickets.  I probably need to look into those and make some things happen, like export approval filing and such<a href="#t01:31" class="time">01:31</a>
nottingf13: hey, i broke it!<a href="#t01:31" class="time">01:31</a>
f13notting: crash?<a href="#t01:31" class="time">01:31</a>
nottingf13: yeah<a href="#t01:31" class="time">01:31</a>
nottingnow, how to debug???<a href="#t01:31" class="time">01:31</a>
f13nod.  I'll hunt ajax down after the meeting<a href="#t01:31" class="time">01:31</a>
f13notting: good luck with that.<a href="#t01:31" class="time">01:31</a>
f13it's evdev, I seriously think the answer here is stop evdev from claiming the device in this situation<a href="#t01:32" class="time">01:32</a>
nottingf13: evdev didn't claim anything according to the log<a href="#t01:32" class="time">01:32</a>
f13notting: really?!<a href="#t01:32" class="time">01:32</a>
nottinglemme try again, but it didn't appear to<a href="#t01:32" class="time">01:32</a>
f13notting: the last thing I see in the log I attached was evdev removal.<a href="#t01:32" class="time">01:32</a>
f13er, last thing before the backtrace<a href="#t01:32" class="time">01:32</a>
warrenhm, do we have any means to make a USB installer now?<a href="#t01:33" class="time">01:33</a>
warren(not live)<a href="#t01:34" class="time">01:34</a>
wwoodsthere's a script.. somewhere<a href="#t01:34" class="time">01:34</a>
f13warren: nothing official.<a href="#t01:35" class="time">01:35</a>
warrenwhere's the unofficial?<a href="#t01:35" class="time">01:35</a>
f13the script I posted to anaconda-devel list many moons ago<a href="#t01:35" class="time">01:35</a>
-!- f13 changed the topic of #fedora-meeting to: Fedora releng - Open Floor<a href="#t01:36" class="time">01:36</a>
warrenf13: if we find a fix for the install from hard drive in the next day is it too late?<a href="#t01:37" class="time">01:37</a>
f13warren: probably not.<a href="#t01:38" class="time">01:38</a>
warrenwwoods: is a bug filed for that?<a href="#t01:38" class="time">01:38</a>
wwoods<a href="https://bugzilla.redhat.com/show_bug.cgi?id=443291">https://bugzilla.redhat.com/show_bug.cgi?id=443291</a> was the NFS iso one<a href="#t01:39" class="time">01:39</a>
buggbotBug 443291: medium, low, ---, Anaconda Maintenance Team, CLOSED RAWHIDE, ISO image via NFS, got "Missing ISO 9660 image" and failed to install<a href="#t01:39" class="time">01:39</a>
wwoods<a href="https://bugzilla.redhat.com/show_bug.cgi?id=443580">https://bugzilla.redhat.com/show_bug.cgi?id=443580</a><a href="#t01:39" class="time">01:39</a>
f13nfsiso is confirmed fixed<a href="#t01:39" class="time">01:39</a>
buggbotBug 443580: low, low, ---, Chris Lumens, CLOSED DUPLICATE, Hard disk installation fails<a href="#t01:39" class="time">01:39</a>
wwoodsis the hdiso variant<a href="#t01:39" class="time">01:39</a>
wwoodswhich was closed as a dupe of the former<a href="#t01:39" class="time">01:39</a>
f13hdiso is now failing for some other reason<a href="#t01:39" class="time">01:39</a>
warrenf13: could you throw that into source control anywhere?<a href="#t01:39" class="time">01:39</a>
wwoodsguess I'll reopen it?<a href="#t01:39" class="time">01:39</a>
poelcatf13: did you see my mail asking when rawhide will start composing from devel branch again?<a href="#t01:39" class="time">01:39</a>
f13wwoods: probably best, and throw in the latest logs which should show something else.<a href="#t01:39" class="time">01:39</a>
warrenwwoods: do we have a full anaconda log of the failure?<a href="#t01:39" class="time">01:39</a>
warrenwwoods: the new failure<a href="#t01:39" class="time">01:39</a>
f13poelcat: I did not, that'll likely be once I start syncing the GOLD bits to the mirrors<a href="#t01:40" class="time">01:40</a>
poelcatf13: want to coordinate it w/ bugzilla versioning<a href="#t01:40" class="time">01:40</a>
f13actually<a href="#t01:40" class="time">01:40</a>
* f13 thinks.<a href="#t01:40" class="time">01:40</a>
f13We kind of have to wait until F9 goes public due to mirror manager redirection<a href="#t01:40" class="time">01:40</a>
f13or else people will just get yum failures for a few days<a href="#t01:41" class="time">01:41</a>
wwoodswarren: they're not interesting<a href="#t01:41" class="time">01:41</a>
wwoodsbut yeah, I can get some<a href="#t01:41" class="time">01:41</a>
poelcatf13: so that would mean, not until 2008-05-13 ?<a href="#t01:42" class="time">01:42</a>
f13poelcat: actually, what we might be able to do, is create empty updates repos and get those synced out, and point all f9 requests to the udpates directory<a href="#t01:43" class="time">01:43</a>
f13poelcat: that way we can let rawhide move on earlier than that, this week, and not break people who are on F9.<a href="#t01:44" class="time">01:44</a>
f13poelcat: I'll start a conversation with mdomsch and co about this<a href="#t01:44" class="time">01:44</a>
f13poelcat: I'd say either the 8th or the 9th rawhide will become F10<a href="#t01:44" class="time">01:44</a>
poelcatf13: okay, that is helpful... can you keep me in the loop?<a href="#t01:44" class="time">01:44</a>
f13yes<a href="#t01:45" class="time">01:45</a>
poelcati'll re-fwd that mail<a href="#t01:45" class="time">01:45</a>
nottingf13: crashing in core server code<a href="#t01:46" class="time">01:46</a>
f13Ok, if there is nothing else, I'm going to go hunt down ajax.<a href="#t01:48" class="time">01:48</a>
f13Ok, rock and roll.<a href="#t01:49" class="time">01:49</a>

Generated by irclog2html.py 2.3 by Marius Gedminas - find it at mg.pov.lt!