17:01:31 <mboddu> #startmeeting RELENG (2018-09-20)
17:01:31 <zodbot> Meeting started Thu Sep 20 17:01:31 2018 UTC.
17:01:31 <zodbot> This meeting is logged and archived in a public location.
17:01:31 <zodbot> The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:31 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:01:31 <zodbot> The meeting name has been set to 'releng_(2018-09-20)'
17:01:31 <mboddu> #meetingname releng
17:01:31 <zodbot> The meeting name has been set to 'releng'
17:01:31 <mboddu> #chair nirik tyll sharkcz masta pbrobinson pingou puiterwijk maxamillion mboddu Kellin dustymabe
17:01:31 <zodbot> Current chairs: Kellin dustymabe masta maxamillion mboddu nirik pbrobinson pingou puiterwijk sharkcz tyll
17:01:31 <mboddu> #topic init process
17:02:04 <puiterwijk> hi
17:02:17 <nirik> morning
17:02:19 <mboddu> puiterwijk: Hello, how is it going?
17:02:22 <puiterwijk> mboddu: I have one point I'd like to bring up.
17:02:28 * dustymabe waves
17:02:38 <mboddu> puiterwijk: Sure, we can start with open floor
17:02:44 <puiterwijk> Oh, I can also wait
17:02:58 <puiterwijk> Just wanted to point out I've got something for open floo
17:04:30 <mboddu> puiterwijk: I dont have anything, you can start
17:04:34 <mboddu> #topic Open Floor
17:04:48 * mboddu might have something, still reading the ticket...
17:04:57 <puiterwijk> Okay. So, it looked like for whatever reason, somewhere yesterday, rawhide was totally emptied out in /pub
17:05:31 <puiterwijk> I noticed this because the new-updates-sync script was taking way too long for normal, it spent very long at rawhide sync, and when I looked, most directories until where it was with syncing where empty
17:06:03 <dustymabe> weird
17:06:12 <mboddu> puiterwijk: Huh, weird
17:06:27 <puiterwijk> n-u-s brought it all back, but I just wanted to bring it to the attention in case any scripts were changed yesterday
17:07:07 <dustymabe> so we don't know why it got emptied ?
17:07:12 <puiterwijk> Nope. I have no clue
17:07:26 <dustymabe> does n-u-s somehow empty it all before it gets added back ?
17:07:31 <puiterwijk> dustymabe: nope
17:07:34 <mboddu> Not that I know off, esp with freeze, we should know if things are changing with FBR's
17:07:35 <puiterwijk> It uses rsync with --delete-after
17:07:38 <nirik> note that we have seen it get stuck.
17:07:47 <dustymabe> does the nightly.sh script somehow still run and cause issues ?
17:07:48 <nirik> (but thats possibly our nfs issues)
17:08:14 <puiterwijk> dustymabe: that might possibly be, yes. Though I'd find that weird, since it also just does rsync
17:08:16 <mboddu> dustymabe: Nope, nightly.sh also does the samething, rsync with --delete-after
17:08:46 <dustymabe> do they both run today? i.e. are we doing two rsyncs when we should only be doing it in n-u-s ?
17:08:54 <puiterwijk> Anyway, I'll see whether it happens again on my pushduty, and maybe the person next week can also make sure.
17:09:04 <nirik> note that putting them all in this one place also means updates can be slower to go out.
17:09:31 <puiterwijk> nirik: yep. Maybe it's an idea to add more cron jobs, where each cron job is for a single release-request combo.
17:09:54 <puiterwijk> The n-u-s script should be able to accept arguments for it (if not, it's simple enough to add), and then we just create multiple cron entries
17:09:59 <puiterwijk> Though it will hit NFS harder
17:10:01 <nirik> but that might overload nfs... but yeah
17:10:28 <puiterwijk> nirik: or we could do one for request=stable and one for request=testing and one for rawhide. Or something like that
17:10:29 <nirik> or just split out rawhide/branched
17:10:37 <puiterwijk> Yeah
17:11:16 <puiterwijk> Anyway, I do not have any further information right now. Just wanted to bring it up so people on pushduty can watch for it
17:11:22 * masta looks in
17:11:31 <mboddu> Thanks pui
17:11:37 <puiterwijk> you're welcome, mbo
17:11:48 <puiterwijk> :P
17:12:21 <mboddu> I think I like request=testing,stable and rawhide, that way it will be easier during freezes
17:12:28 <puiterwijk> Also, just realized: if it's n-u-s or nightly.sh that deletes it all, the fedmsg on #fedora-releng should say it.
17:12:36 <puiterwijk> mboddu: the sync script isn't applied for freeze.
17:12:58 <puiterwijk> Or rather, it will sync whether there's freeze or not, because it only syncs out when a new updates push goes out, which there isn't while pre-GA
17:13:48 <mboddu> puiterwijk: True
17:14:51 <mboddu> puiterwijk: While you are here, I want to bring up https://pagure.io/releng/issue/7827
17:15:05 <mboddu> Nothing is decided yet, but just want to give you a heads up
17:15:15 * mboddu not sure atm what and how this should be handled
17:16:23 <puiterwijk> hum.
17:16:46 <puiterwijk> That's fun
17:17:18 <mboddu> puiterwijk: Indeed, I had a little talk with qa about it today morning to check if its a actual blocker or not
17:17:36 <puiterwijk> And?
17:17:44 <mboddu> Currently they are going to add it to commonbugs for f29 beta, but want to fix by final
17:17:50 <puiterwijk> ok
17:20:09 <mboddu> puiterwijk: If you have any comments/suggestions, please add them to the ticket when you got time
17:20:40 <puiterwijk> Will try to look later
17:20:41 <mboddu> Its going to be crazy to change the tagging structure based on some criteria
17:21:38 <puiterwijk> Yep. This is going to be somewhat really complicated
17:22:00 <nirik> well, mbs just needs to know for flatpaks, don't tag them into fNN-modular
17:22:11 <puiterwijk> Right
17:22:16 <puiterwijk> fNN-flatpaks
17:22:18 <puiterwijk> ?
17:22:30 <nirik> sure, wfm
17:23:27 <mboddu> nirik: Yeah, but I am not sure how crazy that logic going to be
17:24:11 <mboddu> Anyway, other than that, I dont have much for this meeting
17:24:32 <mboddu> I just wanted to bring that issue under the radar
17:25:09 <nirik> so I had something to just note also:
17:25:42 <nirik> #info nirik has synced production src.fedoraproject.org git repos and pagure db to staging, also pdc database.
17:25:49 <nirik> let me know if you hit any issues with it.
17:26:46 <mboddu> Okay, thanks nirik
17:27:03 <mboddu> nirik: Just wondering, how long did it take?
17:28:07 <nirik> not too long. The git repos I synced over night... been doing the databases this morning
17:29:59 <mboddu> nirik: Okay
17:30:24 <nirik> mboddu: could you re-add that module now for f29 and f30?
17:30:41 <nirik> modules/minimal-runtime?
17:31:11 <mboddu> nirik: Okay, sure
17:32:10 <mboddu> nirik: But the ticket says only f29 https://pagure.io/releng/issue/7732
17:48:13 * mboddu knocks on nirik's door :)
17:48:58 <nirik> oh, ok, 29 is ok then
17:49:14 <nirik> I'm in 2 meetings and debugging a networking problem with IT. ;)
17:50:07 <mboddu> nirik: Sure, I can understand, I just wanted to check if I missed the f30 check
17:50:17 <mboddu> Okay, that is all then
17:50:32 <mboddu> I will wait for 5 more min before closing this meeting
18:02:58 <mboddu> Thanks for joining
18:03:00 <mboddu> #endmeeting