15:37:04 #startmeeting RELENG (2016-11-07) 15:37:04 Meeting started Mon Nov 7 15:37:04 2016 UTC. The chair is dgilmore. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:37:04 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:37:04 The meeting name has been set to 'releng_(2016-11-07)' 15:37:04 #meetingname releng 15:37:05 The meeting name has been set to 'releng' 15:37:05 #chair dgilmore nirik tyll sharkcz bochecha masta pbrobinson pingou maxamillion mboddu 15:37:05 Current chairs: bochecha dgilmore masta maxamillion mboddu nirik pbrobinson pingou sharkcz tyll 15:37:08 #topic init process 15:37:44 .hello mohanboddu 15:37:45 mboddu: mohanboddu 'Mohan Boddu' 15:40:51 lets wait a minute for others 15:41:09 * pwhalen is lurking 15:41:51 .hello maxamillion 15:41:52 maxamillion: maxamillion 'Adam Miller' 15:41:57 .hello kevin 15:41:58 nirik: kevin 'Kevin Fenzi' 15:43:17 I lets get started 15:43:24 #topic rawhide 15:43:40 #info rawhide was broken most of last week 15:43:54 the start of the week was missing ppc hosts in the compose channels 15:44:17 when that was fixed it was broken due to systenmd-232 15:44:41 #info gating builds on passing basic QA probably would have helped 15:44:53 * nirik nods. 15:44:53 +1 15:45:31 we would have been able to pick up that installing the rpm failed, that there was unexpanded macros etc 15:46:06 not 100% sure that the rpms installed into bootloader spec locations would have been picked up 15:46:54 I would like to reach out to QA and see what we can look at in regards to getting some gating in place 15:47:10 that may need to go with the flag day for the signing stuff. 15:47:18 or after it. 15:47:22 nirik: it will give us a way to do it 15:47:26 yep 15:47:40 livemedia tasks all failed saturday due to a regression in anaconda 15:48:07 #info livemedia tasks all failed saturday due to a regression in anaconda 15:48:10 https://bugzilla.redhat.com/show_bug.cgi?id=1392468 15:48:38 livemedia tasks sunday and today failed due to systemd-udev not being installed into the chroot 15:48:45 #info livemedia tasks sunday and today failed due to systemd-udev not being installed into the chroot 15:48:58 is there a bug on that one? 15:49:18 not yet 15:49:25 I do not know why it was not installed 15:49:34 so do not know where to file the bug 15:49:36 yeah. hard to tell 15:49:54 it comes back to a long standing dnf bug of not giving us proper output 15:50:01 digging up the bug number 15:50:15 which is somewhat fixed in 2.0, but that won't help until builders are f26 most likely 15:52:07 nirik: its not close to fixed 15:52:22 the 2.0 output is vastly better 15:52:40 nirik: 2.0 only gives you extra info when it fails to install a package failing the transaction 15:53:10 well, you better reopen and speak up in the bug then 15:53:12 https://bugzilla.redhat.com/show_bug.cgi?id=1148627 15:53:23 everyone else thought it was good enough 15:55:29 https://bugzilla.redhat.com/show_bug.cgi?id=1303978 15:55:40 nirik: there is a bug open for ages 15:55:48 its not close to good enough 15:55:54 in this case 15:56:10 we need to know all the depsolving all the time 15:56:27 ok. There is a depsolving log... might have the info 15:56:37 then when systemd-udev dropped out we can see why it was previously pulled in and figure out what changed to make it not be 15:57:39 nirik: what depsolving log? 15:57:44 looks like no, it's useless. 15:58:02 /var/log/dnf.librepo.log 15:58:24 i have never found that useful 15:58:48 so the end result is that all livecds are currently broken due to two bugs 15:58:54 one in anaconda 15:59:08 and one in some unkown change that needs to be figured out 16:00:09 always the puzzles. ;) 16:02:37 nirik: indeed 16:02:52 really do not need to spend another day just tracking things down 16:03:16 thats my rawhide update 16:03:20 other than that 16:03:38 #info rawhide is now aproaching 180,000 rpms 16:04:05 fun times. 16:04:18 awesome 16:04:26 #topic Alternative Architectures updates 16:04:56 sharkcz: so people are being noisy about the breakage due to ppc not being in very good shape at import 16:05:02 there is not a lot broken 16:05:24 * sharkcz is here now 16:05:31 * nirik was going to bring up the broken deps in f25 thing. we should have retired those packages at freeze. ;( 16:05:40 I did file a bug on eclipse-cdt 16:05:41 https://bugzilla.redhat.com/show_bug.cgi?id=1392264 16:06:20 dgilmore: I'm scanning the ppc related failures in koji and trying to react on them ... 16:06:23 nirik: yeah, normally till does it. but I have not seen him in awhile 16:06:39 hes been busy I guess. 16:06:41 pygame is ftbfs 16:06:59 eclipse-cdt is ftbfs also 16:07:26 it seems that the extra broken deps on ppc are mostly due to ftbfs 16:07:27 also noarch stuff that's not fully noarch can have problems 16:07:40 sure 16:07:49 I have not seen much or any of that 16:09:11 #info some unexpected breakage in rawhide on ppc64 and ppc64le 16:09:24 sharkcz: would be good to try get them dealt with asap 16:09:38 otherwise how is f25 on alt arches looking? 16:09:46 dgilmore: yes, it's our priority 16:09:58 f25 still looks good 16:10:40 #info f25 looking okay 16:11:28 sharkcz: any idea if atomic host on power has been attempted yet? 16:12:03 dgilmore: afaik it was attempted, but failed 16:12:34 pbrobinson planned to spend Friday on the ppc composes 16:15:21 sharkcz: okay 16:15:40 #info ppc atomic host attempted but failed 16:15:54 #topic Open Floor 16:16:04 does anyone have anything? 16:16:39 yeah... 16:16:46 we have a bunch of broken deps in f25... 16:16:53 do we want to retire them asap? or just not ? 16:18:17 I think at this point work with QA to determine if anything will break 16:18:29 https://fedorahosted.org/rel-eng/ticket/6488 16:18:38 we have to have a RC by tomorrow or we will have to slip 16:18:44 ok. 16:18:50 yep 16:19:08 but retireing them introduce more breakage that causes a slip 16:19:24 it is something that should have been done over a week ago 16:20:15 yep 16:20:22 https://fedorapeople.org/groups/schedule/f-25/f-25-releng-tasks.html 16:20:28 that says november 1st though 16:20:54 it really needs to be done prior to freeze 16:21:10 yeah, well, here we are. ;) 16:21:24 I am fine with asking qa and deciding something today 16:21:45 sure 16:22:08 #info ask QA to help figure out which broken packages are safe to retire 16:22:41 I would rather us not just do it, then have breakage causing a slip 16:23:01 if there is breakage causing a slip I ant QA with us :D 16:23:05 sorry adamw 16:23:50 anything else? 16:24:12 nope 16:24:23 #endmeeting