From Fedora Project Wiki

Fedora Release Engineering Meeting :: Monday 2009-01-26

Fedora 11 Alpha Status

  • Refreshing freeze tag
  • two open issues:
    1. nash mounting by label/uuid bug
    2. ppc not working
  • gpg signing is a bit up in the air right
    • we need bigger bit sigs to comply with the sha-256 mandate which will require some buildsystem updates
    • Alpha will not be signed

Fedora 11 Alpha Refresh

  • we're considering refreshing the alpha tag
  • koji clone-tag action to bring f11-alpha up to par with dist-f11 late tonight
    • Tuesday's rawhide == F11 alpha

IRC Transcript

f13 ping: jwb rdieter warren wwoods lmacken jeremy notting poelcat spot 10:03
* jwb is here 10:03
* spot is hera 10:03
* poelcat here 10:03
* notting is here 10:03
rdieter yo 10:03
* warren here 10:03
f13 wow! 10:04
warren ? 10:05
jwb lots of people here 10:06
f13 and fast responses 10:06
f13 so we're just missing jeremy and wwoods 10:06
f13 kind of important to have will for the next part 10:07
warren we need will power? 10:07
* lmacken is here 10:08
f13 sending sms 10:09
wwoods boink 10:11
-!- f13 changed the topic of #fedora-meeting to: Fedora releng - Fedora 11 Alpha 10:11
wwoods sorry 10:11
f13 Alright, alpha status. 10:11
wwoods writing status reports and junk 10:11
f13 Things were pretty OK, minus the two big ticket items, the nash mounting by label/uuid bug, and ppc being DOA 10:12
f13 since things were so OK, we talked about a refresh of the freeze tag early this week 10:12
f13 (provided rawhide stayed sane) 10:12
f13 thus far, rawhide is pretty sane 10:12
wwoods https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=480842 10:12
buggbot Bug 480842: high, low, ---, dwmw2@infradead.org, CLOSED RAWHIDE, F11 Rawhide kernel fails to boot on ppc - /vmlinuz: Not a valid ELF image 10:12
f13 and we may have actually fixed ppc 10:12
wwoods and https://bugzilla.redhat.com/show_bug.cgi?id=480667 10:13
buggbot Bug 480667: medium, low, ---, pjones@redhat.com, NEW, nash unable to find dm devs by uuid or label causing boot to fail 10:13
f13 there were 2 problems with ppc, ppc64 had a bad elf image, and ppc32 had non-working squashfs-utils 10:13
f13 jwb got the ppc64 issue squared away, and kylem fixed squashfs for us just today 10:13
f13 I've got a test ppc32 tree that jwb and I are going to throw some machines at to see if it's working, tomorrow's rawhide shoudl be an even better test 10:13
wwoods nice 10:16
wwoods any progress on why ppc kernel builds take for-effing-ever? 10:16
f13 not sure. 10:16
f13 THere is one more bug that has a patch attached to it, and that's the bitmap fonts bug 10:16
f13 we should probably poke ajax today and see if we can get a build out for it 10:16
jwb wwoods, i did som 10:17
jwb wwoods, the answer is "rpm debuginfo sucks" 10:17
jwb the actual build time is on par with most of the other arches 10:17
notting f13: are we rebasing or tagging? 10:18
f13 notting: will get to that, that's the second agenda item 10:18
f13 There is one more thing of note for alpha status. It won't be gpg signed. 10:19
f13 gpg signing is a bit up in the air right now, we need bigger bit sigs to comply with the sha-256 mandate, and that's goign to require some buildsystem updates 10:19
warren f13: ajax might not be back from AU 10:19
f13 also, I'd like a chance to get sigul running so that we coudl explore reducing our gpg key set 10:19
f13 F10 alpha wasn't gpg signed either, so there is some precidence here. 10:20
jwb raise your hand if you care 10:20
wwoods yeah... nope 10:20
f13 ok, anything else on current status before we go and change that status with our next topic? 10:21
-!- f13 changed the topic of #fedora-meeting to: Fedora releng - Alpha refresh 10:23
f13 as I mentioned before, since we're in a good spot, we're considering refreshing the alpha tag 10:23
f13 essentially I'd do another koji clone-tag action to bring f11-alpha up to par with dist-f11 late tonight so that Tuesday's rawhide == F11 alpha 10:24
f13 and we'd go from there to make the bits to put on the mirrors 10:24
f13 in theory we coudl be all wrapped up by late Wed, plenty of time for mirrors to sync 10:24
warren then we all retest that to be sure it installs and boots? 10:24
notting do we have fallback trees in case it goes *bang*? 10:24
warren i'm not against this, just a lot of unknown variables 10:24
f13 warren: yes, we'd re-test everything, which we can get through pretty quickly now 10:25
* stickster adds note to Alpha release notes: Fedora_11_Alpha_release_notes#No_GPG_signatures 10:25
f13 notting: yes, we have a mash of the old f11-alpha tag as of last week. 10:25
dgilmore wwoods: FWIW sparc kernel builds take longer 10:25
dgilmore wwoods: there being build -j16 since that the max fedora will use. most of the time is in debuginfo land 10:26
f13 notting: it may be a better idea to create f11-alpha-2 and use /that/ for the refresh 10:27
f13 notting: giving us f11-alpha to fall back to in case somethign does go bang. 10:27
dgilmore f13: :) sounds sane 10:27
f13 in fact, that's what I'll do 10:27
jwb i like that 10:27
notting f13: i agree, although i may be being paranoid 10:28
f13 no, I am too 10:28
wwoods I have to admit that my testing has all been rawhide-targeted anyway, so I don't anticipate big surprises 10:28
wwoods but, yeah, better safe than sorry 10:28
jwb did most of the mysql deps get fixed? 10:29
f13 Somebody could drop a poop-bomb on us today 10:29
f13 jwb: most of it, looks like 10:29
jwb like gcc 4.4? 10:29
wwoods pretty sure we told the right people to hold off on gcc 4.4 until next week(ish) 10:29
jwb aawww 10:30
jwb no fun 10:30
f13 heh 10:33
f13 jwb: so, if ppc(32) actually looks good today/tomorrow, should we have a ppc alpha? 10:33
jwb probably, but why did you specify 32? 10:33
f13 ppc32 was the last problem we had, the squashfs 10:34
jwb well, that was ppc* 10:34
jwb but yeah 10:34
jwb i'll be testing the ppc compose on a ppc64 machine, if that clarifies anything 10:35
f13 ok 10:37
f13 So then the plan is to create f11-alpha-2 and populate it from dist-f11, at the rawhide compose time tonight (0600 UTC). We'll try to use that as our alpha release, falling back to f11-alpha if necessary 10:38
f13 if we fall back, we can still add the relevant ppc fixes to the tag 10:38
jwb aye 10:38
f13 Today I'm going to be testing the ppc item, as well as doing full x86 composes from today's rawhide to test locally 10:39
f13 I'm also going to do a compsoe in PHX from today's rawhide to prime the mirrors and the magic stage location 10:40
f13 Anything else on the refresh topic> 10:41
f13 ? 10:41
warren f13: will you fix the fedora-release package before then? 10:41
warren f13: all the broken symlinks 10:41
f13 warren: no, as we don't have a key yet to properly point them to 10:42
warren oh 10:43
-!- f13 changed the topic of #fedora-meeting to: Fedora rel-eng - Open Floor 10:45
f13 ok, open floor for any other topics 10:45
f13 alright lets wrap it up and get to work! Thanks all! 10:52

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!