ReleaseEngineering/Meetings/2007-jul-26

From FedoraProject

Jump to: navigation, search

Fedora Release Engineering Meeting :: Thursday 2007-07-26

F8 Test 1 Status Meeting

Decisions

  • Pull in the kernel cebbert is building
  • this may or may not fix boot.iso not working on some subset of (dell?) systems
  • wwoods is going to test the kernel in dist-f8 which isn't in test1 yet to see if it fixes said issue
  • f13 will continue spinning trees and finding issues, hoping to get a golden package set by tomorrow.

IRC Transcript

Idea.png f13 changed the topic of #fedora-meeting to: Fedora Release Engineering - F8 Test1 Meeting<a href="#t14:07" class="time">14:07</a>
f13notting: jeremy: rdieter: spot: warren: jwb_gone: wwoods: ping<a href="#t14:08" class="time">14:08</a>
* jeremy is still here<a href="#t14:08" class="time">14:08</a>
* wwoods still here<a href="#t14:08" class="time">14:08</a>
* notting is still here<a href="#t14:08" class="time">14:08</a>
* spot hides<a href="#t14:08" class="time">14:08</a>
rdieterhere (mostly)<a href="#t14:08" class="time">14:08</a>
* rdieter moves to hide behind spot instead.<a href="#t14:08" class="time">14:08</a>
warrenf13, I haven't had a chance to do any installs of rawhide lately.  too much prepping for moving.<a href="#t14:08" class="time">14:08</a>
warrenf13, what is our f8-test1 tag policy at the moment?<a href="#t14:08" class="time">14:08</a>
wwoodss'ok, there's no images in current rawhide anyway<a href="#t14:09" class="time">14:09</a>
f13poelcat: also you still around?<a href="#t14:09" class="time">14:09</a>
f13so here's the status.<a href="#t14:09" class="time">14:09</a>
f13All packages in f8-test1 are signed with either the -test or the -gold key<a href="#t14:09" class="time">14:09</a>
f13rawhide is composing from that tag and thus should be fully signed<a href="#t14:10" class="time">14:10</a>
poelcatf13: yes<a href="#t14:10" class="time">14:10</a>
f13due to some mistakes, rawhide today isn't installable, but it is pungiable shortly<a href="#t14:10" class="time">14:10</a>
poelcatthought I need to go :(<a href="#t14:10" class="time">14:10</a>
f13(once mirrors are done syncing)<a href="#t14:10" class="time">14:10</a>
f13poelcat: do you have any report from features that need testing in test1?<a href="#t14:10" class="time">14:10</a>
wwoodscebbert: hiya - to answer your bugzilla question, I haven't tried booting with -rc1 yet 'cuz of the rawhide blowup<a href="#t14:10" class="time">14:10</a>
poelcatf13 not yet<a href="#t14:10" class="time">14:10</a>
f13poelcat: ok, cheers.<a href="#t14:10" class="time">14:10</a>
wwoodswe should have boot images with the -rc1 kernel very soon<a href="#t14:11" class="time">14:11</a>
f13anywho, I've been doing some testing with my pungi spins and we're not looking too bad.<a href="#t14:11" class="time">14:11</a>
poelcatplease summarize "decisions" and I'll scrollback and do recap later :)<a href="#t14:11" class="time">14:11</a>
f13what I haven't been able to test yet is ppc, but hopefully my workstation is in good shape to do a compose now, and once rawhide finishes landing I'll kick off a pungi<a href="#t14:11" class="time">14:11</a>
f13poelcat: will do.<a href="#t14:11" class="time">14:11</a>
jeremya quick and dirty rawhide livecd test looked okay.  I was thinking of putting them up on the torrent<a href="#t14:11" class="time">14:11</a>
jeremy(from yesterday)<a href="#t14:11" class="time">14:11</a>
f13I've tossed up a working compose tree to <a href="http://koji.fedoraproject.org/rel-eng/trees/">http://koji.fedoraproject.org/rel-eng/trees/</a><a href="#t14:12" class="time">14:12</a>
f13once I finish today's pungi spins I'll sync those in place.<a href="#t14:12" class="time">14:12</a>
jeremyor we can put the live images there<a href="#t14:12" class="time">14:12</a>
wwoodsf13: oh the rawhide respin is in already in progress? yay<a href="#t14:12" class="time">14:12</a>
f13I'm not so sure we want to spread that url too far, as that could overload the colo/box quickly<a href="#t14:12" class="time">14:12</a>
f13wwoods: it's almost done.  It was rsyncing the output.<a href="#t14:12" class="time">14:12</a>
f13wwoods: it's /still/ not installable though<a href="#t14:12" class="time">14:12</a>
wwoodsbah<a href="#t14:12" class="time">14:12</a>
f13wwoods: so we have to rely upon pungi output for testing, which is somewhat good.<a href="#t14:13" class="time">14:13</a>
wwoodswell, I can do the pungi -GB myself<a href="#t14:13" class="time">14:13</a>
f13nod<a href="#t14:13" class="time">14:13</a>
f13I've started filling out some of the Tree Testing matrix<a href="#t14:13" class="time">14:13</a>
wwoodsbut yeah, I need valid input for that. could you ping me when it's ready? or just send me a boot.iso when you make one?<a href="#t14:13" class="time">14:13</a>
f13obviously I'm going to want help with that (:<a href="#t14:13" class="time">14:13</a>
f13wwoods: sure, lets talk after this meeting<a href="#t14:13" class="time">14:13</a>
wwoodsf13: sure<a href="#t14:14" class="time">14:14</a>
f13As for the f8-test1 tag, lets only tag things that we know ar efixing something with test1, and making sure we're vocal about it on rel-eng@fp.o<a href="#t14:14" class="time">14:14</a>
jeremyf13: I'll do rawhide live on the torrent then<a href="#t14:14" class="time">14:14</a>
f13I know there was a request this morning for a tag, maybe 2, it's OK to tag those, just remind me you're doing it so that I can sign them.<a href="#t14:14" class="time">14:14</a>
f13jeremy: that would rock very hard.<a href="#t14:14" class="time">14:14</a>
* notting curses at pungi again<a href="#t14:15" class="time">14:15</a>
wwoodsdid the New Yum-Updatesd Hotness get tagged?<a href="#t14:15" class="time">14:15</a>
f13cebbert: I asked you to join so that we can get an update on the kernel status.  We're not using the latest built kernel righ tnow.<a href="#t14:15" class="time">14:15</a>
jeremywwoods: yeah, was built before we started tagging<a href="#t14:15" class="time">14:15</a>
f13$ koji latest-pkg --quiet f8-test1 kernel<a href="#t14:15" class="time">14:15</a>
f13kernel-2.6.23-0.45.rc0.git16.fc8          f8-test1              linville<a href="#t14:15" class="time">14:15</a>
f13cebbert: is there any compelling reason to move off this kernel?<a href="#t14:15" class="time">14:15</a>
cebbertlatest kernel looks pretty good, at least there aren't major complaints on linux-kernel<a href="#t14:16" class="time">14:16</a>
wwoodsyarg. looks like I need to build my own boot.iso after all<a href="#t14:16" class="time">14:16</a>
f13wwoods: oh?  whyzat?<a href="#t14:16" class="time">14:16</a>
wwoodsI can't get that kernel to boot from boot.iso on my test machines<a href="#t14:16" class="time">14:16</a>
cebbertyeah, and we need to solve that one<a href="#t14:16" class="time">14:16</a>
wwoodssame with rwilliams, also reported on -test-list<a href="#t14:16" class="time">14:16</a>
f13wwoods: newer kernels don't help either though right?<a href="#t14:16" class="time">14:16</a>
wwoodscebbert was thinking -rc1 might fix it<a href="#t14:17" class="time">14:17</a>
f13oh.<a href="#t14:17" class="time">14:17</a>
wwoodshaven't gotten a boot.iso built with -rc1 yet.<a href="#t14:17" class="time">14:17</a>
f13wwoods: if you could verify, that would rock the socks.<a href="#t14:17" class="time">14:17</a>
f13wwoods: then we'd have compelling reason to move to a new kernel.<a href="#t14:17" class="time">14:17</a>
cebbertthe x86 setup code was completely rewritten early in 2.6.23 cycle<a href="#t14:17" class="time">14:17</a>
wwoodsI thought that'd be in today's rawhide, but I was wrong. So yeah, I'll be testing that shortly<a href="#t14:17" class="time">14:17</a>
warrenHow are we on actually deciding to stop tagging things and shipping test1?<a href="#t14:17" class="time">14:17</a>
f13the thing is, with the current kernel I haven't heard too many complaints, other than wwoods boot.iso.  And going to a new kernel is always a risk, one I'd like to avoid given that it's Thursday already.<a href="#t14:17" class="time">14:17</a>
wwoodsf13: unlike the last Dell No Boot problem, this one appears to have no workaround :/<a href="#t14:18" class="time">14:18</a>
f13warren: I've always gone with the 'we froze, only tag things that will get us to a working compose'<a href="#t14:18" class="time">14:18</a>
f13warren: it doesn't seem as far spread though<a href="#t14:18" class="time">14:18</a>
cebbertyou probably want to wait for the next kernel, APM was completely broken before today<a href="#t14:18" class="time">14:18</a>
f13APM?<a href="#t14:18" class="time">14:18</a>
cebbertyeah, some people still use that<a href="#t14:18" class="time">14:18</a>
nottingcebbert: bah, let's find out who still does<a href="#t14:19" class="time">14:19</a>
f13heh<a href="#t14:19" class="time">14:19</a>
f13cebbert: eta to 'next kernel' ?<a href="#t14:19" class="time">14:19</a>
cebbertI was going to start the build soon, can start now<a href="#t14:19" class="time">14:19</a>
f13k, knowing if the current latest koji kernel fixes wwoods' issue would be really good info too.<a href="#t14:20" class="time">14:20</a>
f13Ideally we'd be able to pick up this kernel, the new selinux-policy and be good with a spin tomorrow<a href="#t14:20" class="time">14:20</a>
cebbertnext kernel is rc1-git3<a href="#t14:21" class="time">14:21</a>
f13that i can hand off to IS and we might have a shot at mirrors being able to get to it by Tuesday.<a href="#t14:21" class="time">14:21</a>
f13an added wrinkle is that jeremy is going ot be out next week<a href="#t14:21" class="time">14:21</a>
f13so I'd /really/ like to get at least the golden package set by tomorrow<a href="#t14:21" class="time">14:21</a>
f13(since jeremy is our current well of LiveCD information.  I can get a crash course but...)<a href="#t14:22" class="time">14:22</a>
jeremyf13: you just don't want to have to make live cds :)<a href="#t14:22" class="time">14:22</a>
f13jeremy: I don't mind making them if you're around to bail me out<a href="#t14:22" class="time">14:22</a>
f13so that's where we stand.  More pungi output landing on koji.fp.o later today, maybe even ppc.<a href="#t14:23" class="time">14:23</a>
f13wwoods: are you going to be able to hook us up with some automated testing love?<a href="#t14:23" class="time">14:23</a>
wwoodsf13: yeah, I'm gonna try to coax KATE into doing some rawhide runs<a href="#t14:25" class="time">14:25</a>
f13hrm, so that migh tnot show up until tomorrow<a href="#t14:25" class="time">14:25</a>
f13given that rawhide lacks images.<a href="#t14:25" class="time">14:25</a>
f13*sigh*<a href="#t14:25" class="time">14:25</a>
wwoodswell, I can point it at another tree if there's one local<a href="#t14:25" class="time">14:25</a>
wwoodsprobably I can hack together something to get it to recognize the tree I build<a href="#t14:25" class="time">14:25</a>
f13wwoods: if "local" == Westford, I think we can hook that up.<a href="#t14:26" class="time">14:26</a>
f13or that.<a href="#t14:26" class="time">14:26</a>
wwoodsbut KATE is really designed for testing stuff from well-known locations, and with particular setups<a href="#t14:26" class="time">14:26</a>
f13ok, in the interest of Getting Stuff Done, does anybody have anymore to talk about?<a href="#t14:26" class="time">14:26</a>
warrenI have to go, contractor is here.<a href="#t14:26" class="time">14:26</a>
f13poelcat: Decisions; We're going to pull in the kernel cebbert is building, which may or may not fix boot.iso not working on some subset of (dell?) systems.  wwoods is going ot test the kernel in dist-f8 which isn't in test1 yet to see if it fixes said issue.  I'm going to continue spinning trees and finding issues, hopeing to get a golden package set by tomorrow.<a href="#t14:28" class="time">14:28</a>
f13if there is nothing else.... go forth and prosper.<a href="#t14:28" class="time">14:28</a>
wwoodswhoa, my pungi-fu is getting stronger. added a custom local repo with the new kernel and it worked on the first try.<a href="#t14:29" class="time">14:29</a>
Idea.png f13 changed the topic of #fedora-meeting to: Channel is used by various Fedora groups and committees for their regular meetings | Note that meetings often get logged | See <a href="http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel">http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel</a> for meeting schedule<a href="#t14:29" class="time">14:29</a>

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