Sunday, 2015-05-17

*** _nadya_ has quit IRC00:01
*** yfried is now known as yfried|afk00:02
*** salv-orlando has joined #openstack-infra00:04
*** ivar-lazzaro has joined #openstack-infra00:06
*** ashleighfarnham has joined #openstack-infra00:10
*** julim has quit IRC00:10
*** zz_dimtruck is now known as dimtruck00:11
*** julim has joined #openstack-infra00:12
openstackgerritJames E. Blair proposed stackforge/gertty: Fix diff display of deleted empty files
*** yfried|afk is now known as yfried00:14
*** baoli has joined #openstack-infra00:24
*** markvoelker has quit IRC00:25
*** yfried is now known as yfried|afk00:28
*** leopoldj has quit IRC00:33
*** baoli has quit IRC00:34
*** rossella_s has quit IRC00:37
*** sdague_mb has quit IRC00:37
*** boris-42 has quit IRC00:38
*** baoli has joined #openstack-infra00:43
*** julim has quit IRC00:48
*** yfried|afk is now known as yfried00:49
*** yamamoto has joined #openstack-infra00:49
*** salv-orlando has quit IRC00:50
*** ivar-laz_ has joined #openstack-infra00:50
*** ivar-lazzaro has quit IRC00:52
*** baoli has quit IRC00:58
openstackgerritJames E. Blair proposed stackforge/gertty: Fix refresh on project and change lists
*** yfried is now known as yfried|afk00:59
*** dkranz has quit IRC01:00
*** sdake_ has joined #openstack-infra01:02
*** sdake has quit IRC01:02
*** otter768 has joined #openstack-infra01:04
*** otter768 has quit IRC01:08
*** ashleighfarnham has quit IRC01:12
*** ivar-laz_ has quit IRC01:13
*** ivar-lazzaro has joined #openstack-infra01:14
*** shashankhegde has quit IRC01:22
Clintjeblair: 194, 17701:24
*** sarob has quit IRC01:27
*** achanda has quit IRC01:27
*** sdake_ has quit IRC01:30
*** sdake has joined #openstack-infra01:30
*** woodster_ has quit IRC01:30
jeblairClint: urwid has ''; is that what that's from?01:33
Clintjeblair: correct01:36
jeblairClint: for reference: my modmap: mod1        Alt_L (0x40),  Alt_R (0x6c),  Meta_L (0xcd)01:40
jeblairClint: and input_test says:01:40
jeblair'meta 1' 27, 4901:40
Clintmod1        Alt_L (0x40),  Alt_R (0x6c),  Meta_L (0xcd)01:41
Clintand ± 194, 17701:41
Clintand the same for right alt01:42
Clintjeblair: another datapoint is that alt-1 actually produces ± in an xterm01:46
jeblairClint: looking at the code, i wonder if unicode has something to do with this01:46
Clintwho doesn't love unicode?01:47
jeblairClint: maybe something about your terminal input encoding or mumble mumble01:48
Clintjeblair: so i can bind it to '!' but not '±'01:50
*** mestery has quit IRC01:50
Clintwhich is of course orthogonal to the problem01:50
*** salv-orlando has joined #openstack-infra01:50
*** yamamoto has quit IRC01:51
jeblaircan you pastebin the error when you try to bind it to the glyph?01:51
*** julim has joined #openstack-infra01:51
*** yamamoto has joined #openstack-infra01:55
*** salv-orlando has quit IRC01:55
*** salv-orlando has joined #openstack-infra01:58
*** salv-orlando has quit IRC02:03
*** dimtruck is now known as zz_dimtruck02:06
*** zz_dimtruck is now known as dimtruck02:08
*** armax has quit IRC02:14
*** dboik has joined #openstack-infra02:15
*** armax has joined #openstack-infra02:19
*** armax has quit IRC02:25
*** yamahata has quit IRC02:27
*** salv-orlando has joined #openstack-infra02:29
*** woodster_ has joined #openstack-infra02:33
*** sks has joined #openstack-infra02:39
openstackgerritJames E. Blair proposed openstack-infra/system-config: Fix opening public UDP ports
*** sdake has quit IRC02:41
jeblairfungi, clarkb, mordred: ^ we should land that asap (assuming it's correct)02:42
jeblairfungi: i did some testing earlier to make sure that i could connect to the pbx via sip+jitsi from my laptop and found that problem02:43
jeblairaft[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[C[Cer correcting it manually, it works, so assuming that fixes the puppet problem i should be all set if we need to use the bridge for folks who can't make the summit in person02:44
jeblairwow that was col02:44
*** sdake has joined #openstack-infra02:44
Clintneeds more [C02:44
fungijeblair: yep, that looks like it was definitely a bug02:44
*** sarob has joined #openstack-infra02:45
clarkbwait puppet I thought wow02:45
fungivariables need to be set before you reference them02:45
jeblaircrinkle tells me that is the case.  :)02:46
clarkbbut its declarative02:46
jeblairi mean, it's puppet -- i asked02:46
crinklepuppet parses variables from top to bottom :/02:46
fungicrinkle: with the wisdom02:46
*** btully has joined #openstack-infra02:47
crinklenot wisdom, docs02:47
jeblair(also the pbx pstn line seems to work too)02:47
clarkbI just approved02:47
jeblaircrinkle: docs == wisdom02:47
fungirecorded wisdom for future generations02:48
jeblairftr crinkle is not actual loling.  but thx anyway.  :)02:48
crinklenow i am02:48
fungii could see02:48
clarkbtrain must be going well then02:48
jeblairclarkb: do you feel like you're here? :)02:48
jeblairclarkb: imagine monty and morganfainberg yelling about howe everything is wrong with openstack and identity and stuff, the puget sound out the window, and you're there.02:50
*** jogo has quit IRC02:50
clarkbI'm sad I have to miss it02:51
fungialso pyramid ipa02:52
fungiit's not great, but it's not budweiser02:52
jeblairclarkb: me too :(02:52
* jeblair raises a glass to clarkb02:52
*** yamamoto has quit IRC02:53
clarkbgreghaynes you should remove the git clones from the dib jobs, we can go straight zuul-cloner now02:53
clarkbI am going to see if I can catch up on reviews this week with everyone else summiting02:56
openstackgerritMerged openstack-infra/system-config: Fix opening public UDP ports
jeblairclarkb: aw man.  you and AJaeger are going to trounce us.02:57
* fungi cries unfair advantage03:04
*** otter768 has joined #openstack-infra03:05
* fungi suspects clarkb may _actually_ be busy with other things besides code review next week03:05
clarkbyes likely03:06
fungiAJaeger on the other hand *will* continue to trounce us03:07
*** otter768 has quit IRC03:10
*** Sukhdev has joined #openstack-infra03:10
*** rguillebert has quit IRC03:16
*** stevemar has joined #openstack-infra03:22
*** dboik has quit IRC03:23
jeblairmordred:  let's name M03:24
jeblairclarkb, fungi: -A openstack-INPUT -m udp -p udp --dport 5060 -j ACCEPT03:24
jeblairis on pbx03:24
*** sarob has quit IRC03:25
*** sarob has joined #openstack-infra03:25
*** sarob has quit IRC03:31
*** funzo_ has quit IRC03:31
*** sarob has joined #openstack-infra03:31
*** funzo has joined #openstack-infra03:32
*** funzo has quit IRC03:32
*** dboik has joined #openstack-infra03:32
*** funzo has joined #openstack-infra03:32
*** funzo has joined #openstack-infra03:32
*** funzo has quit IRC03:32
*** sks has quit IRC03:34
fungijeblair: awesome. now clarkb can summit (can be summited? can haz summitation?)03:35
clarkbjeblair s/election/poll/ maybe? election implies a seat to occupy but maybe thats not really worth thinking too hard about03:37
jeblairclarkb: yeah, i think that's probably a good idea03:38
*** salv-orlando has quit IRC03:38
jeblairclarkb: updated03:39
*** zul has joined #openstack-infra03:43
*** fedexo has joined #openstack-infra03:44
*** patrickeast has quit IRC03:46
*** jogo has joined #openstack-infra03:46
*** dboik has quit IRC03:55
*** zul has quit IRC03:55
*** tiswanso has joined #openstack-infra04:01
*** jcoufal has joined #openstack-infra04:04
*** sarob has quit IRC04:05
*** tiswanso has quit IRC04:05
*** zul has joined #openstack-infra04:06
*** patrickeast has joined #openstack-infra04:09
*** jcoufal_ has joined #openstack-infra04:10
*** zul has quit IRC04:12
*** sdake has quit IRC04:12
*** jcoufal has quit IRC04:14
*** patrickeast has quit IRC04:19
*** stevemar has quit IRC04:22
*** ivar-lazzaro has quit IRC04:23
*** ivar-lazzaro has joined #openstack-infra04:24
*** jcoufal_ has quit IRC04:24
*** jcoufal has joined #openstack-infra04:24
*** dboik has joined #openstack-infra04:26
*** zul has joined #openstack-infra04:31
*** zul has quit IRC04:31
*** yfried has joined #openstack-infra04:32
*** yfried|afk has quit IRC04:35
*** trey has quit IRC04:35
*** trey has joined #openstack-infra04:36
*** jcoufal_ has joined #openstack-infra04:37
*** salv-orlando has joined #openstack-infra04:39
*** jcoufal has quit IRC04:41
*** salv-orlando has quit IRC04:44
*** dboik has quit IRC04:50
*** btully has quit IRC04:59
*** patrickeast has joined #openstack-infra05:00
*** jcoufal has joined #openstack-infra05:00
*** dboik has joined #openstack-infra05:03
*** cody-somerville has joined #openstack-infra05:04
*** jcoufal_ has quit IRC05:04
*** yamahata has joined #openstack-infra05:05
*** shashankhegde has joined #openstack-infra05:05
*** otter768 has joined #openstack-infra05:06
*** abregman has joined #openstack-infra05:08
*** Sukhdev has quit IRC05:10
*** shashankhegde has quit IRC05:10
*** yfried has quit IRC05:11
*** salv-orlando has joined #openstack-infra05:11
*** otter768 has quit IRC05:11
*** ihrachyshka has joined #openstack-infra05:12
*** jcoufal_ has joined #openstack-infra05:15
*** changbl has quit IRC05:15
*** jcoufal has quit IRC05:18
*** ihrachyshka has quit IRC05:22
*** krtaylor has quit IRC05:22
*** dboik has quit IRC05:28
*** achanda has joined #openstack-infra05:40
*** jcoufal_ has quit IRC05:51
*** tteggel has quit IRC05:56
*** tteggel has joined #openstack-infra05:57
*** btully has joined #openstack-infra05:57
*** salv-orlando has quit IRC06:14
*** patrickeast has quit IRC06:16
*** _nadya_ has joined #openstack-infra06:28
*** yamahata has quit IRC06:33
*** yfried has joined #openstack-infra06:35
*** yfried is now known as yfried|afk06:35
*** patrickeast has joined #openstack-infra06:35
*** _nadya_ has quit IRC06:35
*** shashankhegde has joined #openstack-infra06:41
*** salv-orlando has joined #openstack-infra06:41
*** shashankhegde has quit IRC06:45
*** salv-orlando has quit IRC06:45
*** sarob has joined #openstack-infra06:51
*** ZZelle has quit IRC06:56
*** yfried has joined #openstack-infra06:58
*** yfried|afk has quit IRC06:59
*** fedexo has quit IRC07:02
*** otter768 has joined #openstack-infra07:07
*** ZZelle_ has quit IRC07:08
*** juggler has joined #openstack-infra07:09
*** ZZelle_ has joined #openstack-infra07:09
*** otter768 has quit IRC07:11
*** panda|gone has quit IRC07:13
*** panda|gone has joined #openstack-infra07:14
*** amotoki has joined #openstack-infra07:16
*** yfried has quit IRC07:21
electricalfungi: ah very cool :)07:24
*** sergsh has joined #openstack-infra07:25
*** amotoki has quit IRC07:25
*** boris-42 has joined #openstack-infra07:27
*** amotoki has joined #openstack-infra07:38
*** yamamoto has joined #openstack-infra07:39
*** yfried has joined #openstack-infra07:42
*** yfried is now known as yfried|afk07:42
*** amotoki has quit IRC07:49
*** yamamoto has quit IRC07:53
*** shashankhegde has joined #openstack-infra08:03
*** sks has joined #openstack-infra08:09
*** sergsh has quit IRC08:11
*** sks has quit IRC08:14
*** unicell has joined #openstack-infra08:18
*** unicell1 has quit IRC08:21
*** sergsh has joined #openstack-infra08:24
*** sergsh has quit IRC08:30
*** shashankhegde has quit IRC08:33
*** julim has quit IRC08:34
*** julim has joined #openstack-infra08:34
*** julim has quit IRC08:36
*** yfried has joined #openstack-infra08:37
*** yfried|afk has quit IRC08:39
*** ivar-lazzaro has quit IRC08:41
*** jamespage has joined #openstack-infra08:43
*** salv-orlando has joined #openstack-infra08:50
*** salv-orlando has quit IRC08:55
*** amotoki has joined #openstack-infra08:59
*** amotoki has quit IRC09:04
*** otter768 has joined #openstack-infra09:07
openstackgerritgreghaynes proposed openstack-infra/project-config: Remove zuul-cloner workaround for DIB
*** otter768 has quit IRC09:12
*** achanda has quit IRC09:15
*** krtaylor has joined #openstack-infra09:17
*** banix has joined #openstack-infra09:17
*** patrickeast has quit IRC09:19
*** sdake has joined #openstack-infra09:21
*** sergsh has joined #openstack-infra09:25
*** krtaylor has quit IRC09:39
*** banix has quit IRC09:39
*** patrickeast has joined #openstack-infra09:49
*** yfried is now known as yfried|afk09:49
*** yfried|afk is now known as yfried09:56
*** baoli has joined #openstack-infra10:07
*** baoli has quit IRC10:09
*** smccully has quit IRC10:19
*** teran has quit IRC10:24
*** sdake has quit IRC10:32
*** salv-orlando has joined #openstack-infra10:39
*** salv-orlando has quit IRC10:44
*** yfried is now known as yfried|afk11:02
*** yfried|afk is now known as yfried11:06
*** otter768 has joined #openstack-infra11:08
*** otter768 has quit IRC11:13
*** shardy_ has joined #openstack-infra11:24
*** shardy has quit IRC11:25
*** Krinkle|detached is now known as Krinkle11:28
*** shardy_ has quit IRC11:29
*** shardy has joined #openstack-infra11:30
*** dboik has joined #openstack-infra11:30
*** dboik has quit IRC11:31
*** salv-orlando has joined #openstack-infra11:40
*** sks has joined #openstack-infra11:44
*** salv-orlando has quit IRC11:44
*** rlandy has joined #openstack-infra11:51
*** charz_ is now known as charz12:10
*** yolanda has joined #openstack-infra12:23
*** Krinkle is now known as Krinkle|detached12:28
openstackgerritDarragh Bailey proposed openstack-infra/jenkins-job-builder: Fix Builder.update_job() on Python 3
openstackgerritDarragh Bailey proposed openstack-infra/jenkins-job-builder: Fix Builder.update_job() on Python 3
*** boris-42 has quit IRC12:48
*** hdd has joined #openstack-infra13:06
*** otter768 has joined #openstack-infra13:09
*** otter768 has quit IRC13:14
*** salv-orlando has joined #openstack-infra13:29
*** mase_x200 has joined #openstack-infra13:29
*** megm has quit IRC13:30
*** ajo has joined #openstack-infra13:32
*** salv-orlando has quit IRC13:34
*** ajo has quit IRC13:38
*** mase_x200 has quit IRC13:40
*** hdd has quit IRC13:41
*** liam__ has joined #openstack-infra13:42
*** mase_x200 has joined #openstack-infra13:43
*** liam__ has quit IRC13:45
*** igorbelikov has quit IRC13:46
*** igorbelikov has joined #openstack-infra13:49
*** salv-orlando has joined #openstack-infra13:49
*** salv-orlando has quit IRC13:49
*** salv-orlando has joined #openstack-infra13:49
*** salv-orlando has quit IRC13:50
*** salv-orlando has joined #openstack-infra13:50
*** annegentle has joined #openstack-infra13:54
*** salv-orl_ has joined #openstack-infra13:56
*** salv-orlando has quit IRC13:56
*** yfried is now known as yfried|afk13:56
*** annegentle has quit IRC13:57
*** yfried|afk is now known as yfried13:57
*** changbl has joined #openstack-infra14:00
*** hdd has joined #openstack-infra14:04
*** hdd has quit IRC14:05
*** mwagner_lap has quit IRC14:20
*** unicell has quit IRC14:22
*** mase_x200 has quit IRC14:22
*** shardy has quit IRC14:26
*** mase_x200 has joined #openstack-infra14:28
*** shardy has joined #openstack-infra14:31
*** abregman has quit IRC14:33
*** sks has quit IRC14:42
*** salv-orl_ has quit IRC14:44
*** mase_x200 has quit IRC14:48
*** julim has joined #openstack-infra14:49
*** liam__ has joined #openstack-infra14:51
*** liam__ has quit IRC14:51
*** julim has quit IRC14:51
*** julim has joined #openstack-infra14:55
*** annegentle has joined #openstack-infra14:58
*** sarob has quit IRC15:01
*** sarob has joined #openstack-infra15:01
*** annegentle has quit IRC15:03
*** sarob has quit IRC15:04
*** sarob has joined #openstack-infra15:05
*** otter768 has joined #openstack-infra15:10
*** Krinkle|detached is now known as Krinkle15:13
*** otter768 has quit IRC15:15
*** sdake has joined #openstack-infra15:18
*** ihrachyshka has joined #openstack-infra15:22
*** armax has joined #openstack-infra15:23
*** sks has joined #openstack-infra15:24
*** sarob has quit IRC15:27
*** sarob has joined #openstack-infra15:27
*** ihrachyshka_ has joined #openstack-infra15:30
*** alexpilotti has joined #openstack-infra15:31
*** sarob has quit IRC15:32
*** yamahata has joined #openstack-infra15:32
*** ihrachyshka has quit IRC15:34
*** ihrachyshka_ has quit IRC15:34
*** zul has joined #openstack-infra15:35
*** palar has joined #openstack-infra15:43
*** salv-orlando has joined #openstack-infra15:44
*** _nadya_ has joined #openstack-infra15:47
*** yfried has quit IRC15:47
*** yamahata has quit IRC15:48
*** armax has quit IRC15:50
*** salv-orlando has quit IRC15:50
*** yolanda has quit IRC15:56
*** yolanda has joined #openstack-infra16:00
*** jogo has quit IRC16:02
*** shashankhegde has joined #openstack-infra16:03
*** _nadya_ has quit IRC16:14
*** julim has quit IRC16:18
*** salv-orlando has joined #openstack-infra16:22
mordredjeblair: ++16:23
openstackgerritDarragh Bailey proposed openstack-infra/jenkins-job-builder: Fix Builder.update_job() on Python 3
*** mestery has joined #openstack-infra16:26
*** salv-orlando has quit IRC16:27
*** annegentle has joined #openstack-infra16:28
*** ociuhandu has joined #openstack-infra16:34
*** rossella_s has joined #openstack-infra16:35
*** ashleighfarnham has joined #openstack-infra16:40
*** mestery has quit IRC16:41
*** banix has joined #openstack-infra16:43
*** vishy has left #openstack-infra16:43
*** banix has quit IRC16:54
*** _nadya_ has joined #openstack-infra16:57
lifelessok yay we're not coinstallable again16:57
lifelessits not just my resolver having trouble :)16:57
*** banix has joined #openstack-infra17:00
fungiit's clearly a schrödinger's packageset. if you hadn't looked it wouldn't have been either coinstallable or not17:00
lifelessits os-brick17:01
lifelessits incompatible with current releases of os-*17:01
lifelesse.g. i18n17:01
fungithingee: ^ fun times!17:01
lifelessI mean its good to have my resolver heuristics worked out...17:03
*** julim has joined #openstack-infra17:08
thingeehemnafk: ^17:11
*** otter768 has joined #openstack-infra17:11
*** otter768 has quit IRC17:15
*** ociuhandu has quit IRC17:18
*** ashleighfarnham has quit IRC17:19
*** sergsh has quit IRC17:22
*** mestery has joined #openstack-infra17:22
lifelessoh reat17:23
lifelessalso flake817:23
*** ociuhandu has joined #openstack-infra17:25
lifelesshacking requires 2.2.417:28
lifelessall current releases17:28
lifelessfungi: everything is awful17:30
clarkbfungi I will get the puppet testing etherpad up and in shape soon (either today or tomorrow) so it is ready for you17:30
fungiclarkb: awesome--that's really appreciated. though i can probably find time for it if you can't. priorities, man17:31
openstackgerritlifeless proposed openstack/requirements: Fix requirements to be co-installable.
lifelessmordred: ^ since you're not at all busy right now17:35
*** salv-orlando has joined #openstack-infra17:37
*** stevemar has joined #openstack-infra17:38
mordredlifeless: I'm not sure I agree with the flake8 change - it's there to facilitate
lifelessmordred: its incompatible with hacking17:40
mordredwe have to land a change to requirements to be able to land the change to hacking17:41
lifelessmordred: its not a matter of agreement17:41
mordredlifeless: it is17:41
mordredwe can't update hacking without first updating g-r17:41
*** fallenpegasus has joined #openstack-infra17:41
lifelessmordred: this is exactly the deadlock that I have warned about showing up in practice17:41
mordredso there must be a mechanism to do that17:41
fungithat's what 183930 does?17:41
*** julim has quit IRC17:41
*** achanda has joined #openstack-infra17:42
lifeless183930 undoes it17:42
mordredright. the answer is landing the change to hacking, not reverting the change to requirements17:42
lifelessand lets change teh process so hacking lists known bad, not known good17:42
fungifair enough. does hacking get sync'd from g-r?17:42
lifelessbut thats for later in the week17:42
lifelessbut we can do this compatibly let me update my patch17:43
mordredlifeless: hacking is quite specially and very explicitly wants a hard pin - because we do not want any variation in a flake8 version in a cycle17:43
mordredfungi: it does17:43
mordredknown-bad for flake8 is a strange idea17:43
*** sdake has quit IRC17:43
lifelessmordred: yes, but transitions.17:43
mordredI do not think requirements needs to be sstrictly co-installable every second of the day17:44
lifelessmordred: I'm not sggesting variations during a cycle.17:44
lifelessmordred: we'll come back to that17:44
fungii'm wondering how we can represent that hacking should be able to pick its own dep versions independent of g-r17:44
*** sdake has joined #openstack-infra17:44
lifelessmordred: [you're wrong, but now isn't the time to debate]17:44
mordredwell, I'm not +2 on this until I understand why I'm wrong and what the sollution is17:44
lifelessthere are several confounded things17:44
fungimaybe we can ensure that the explicit deps of hacking don't overlap with explicit deps of other projects, and then remove its g-r enforcement/sync so that it can vary independently?17:45
lifelessright now, I now understand there is a transition happening for hacking's deps and I'm going to check how to represent that17:45
mordred-1 left17:45
mordredI'll be +2 when I understand the future state and how it does the thigns we want17:45
fungibasically if any project wants to participate in global-requirements and wants to use any of hackings requirements, it should do so through hacking as a proxy?17:46
mordredno project shold EVER directly depend on flake8 other than hacking17:46
mordredthan via17:46
fungiin which case we should take any of hacking's requirements out of global-requirements17:46
clarkbor pep8 or pyflakes17:46
mordredyah. I think it would be fine to special-case hacking17:46
mordredand deal with it outside of g-r17:46
lifelessno need17:47
fungibasically be able to guarantee that no project's first-order requirements conflict with hacking over its first-order requirements (and hope that the transitive requirement tree is shallow enough there that we don't run into more issues)17:47
lifelessbut I can understand why you're thinking that17:47
*** rossella_s has quit IRC17:50
*** sparkycollier has joined #openstack-infra17:50
*** woodster_ has quit IRC17:50
openstackgerritlifeless proposed openstack/requirements: Fix requirements to be co-installable.
lifelessmordred: ok so on 'should g-r be always resolvable' - yes, it must be, because if its not pip 8 will refuse to install.17:51
lifelessmordred: you need to argue why it should ever be *unresolvable*, not why it must always be resolvable.17:51
lifelessmordred: that update, with the embedded comment, shows one way of achieving a transition.17:52
lifelessmordred: An actual OR operator might be nicer (and I'm open to putting the effort into add one - I can't see one in pthe specifier code today)17:53
lifelessmordred: we need CI to be really robust and reliable; being able to specify known good for CI jobs - which is what we attempt today - is critical to avoiding flack from upstream changes breaking us.17:54
lifelessmordred: I think we agree on that17:54
*** fedexo has joined #openstack-infra17:54
pabelangero/ from Vancouver17:55
mordredlifeless: yes. we do.17:55
mordredlifeless: how does that change end up?17:55
*** ashleighfarnham has joined #openstack-infra17:56
lifelessmordred: which change?17:56
mordredlifeless: do we land that change, then land a change to hacking that matches that, then land a change to remove the lower bound in g-r, then land a change to hacking to do that?17:56
mordredlifeless: (I'm missing the context of what your plan is)17:56
lifelessfor the hacking thing. we land this patch (fixes g-r). We land the ==2.4.0 change in hacking and release. We then tighten up g-r.17:57
lifelesswhen we release hacking it will switch everything from 2.2.4 to 2.4.0 as the new release is consumed17:57
mordredlifeless: ok - so - how do we land that change in hacking since ==2.4.0 would be a different value than is in g-r?17:57
lifeless==2.4.0 is compatible with the g-r specifier.17:58
mordredand our sync code supports that now?17:58
fungicurrent g-r enforcement will require that the string matches17:58
lifelessWe'll need it to to do transitions on things where we do want a hard pin in the project17:58
lifelessOk, so thats something I can prep (or help someone prep) - but deep dive possibly needed during the week17:59
fungicount me in17:59
mordredyah. I mean, I thnik it's a 4-step for now, and potentailly the pain of that will motivate us to do the betterer17:59
lifeless(deep dive is on the conceptual design - the code is yeasy)17:59
*** stevemar has quit IRC18:00
lifelessso yeah today, g-r:range, hacking:range, hacking:release, g-r:==, hacking:==, hacking:release18:00
mordredit's "land this change, sync this change to hacking, land a new change with the ==, sync the == to hacking" - which is obviously silly, but we'll fix the silly later18:00
mordredlifeless: yah18:00
*** mestery has quit IRC18:00
mordredk. I'm on board with that18:00
mordrednow - the other thing - os-brick18:00
lifelessit needs a current release18:01
lifelesssynced with current requirements18:01
lifelessif its going to be in current requirements18:01
*** salv-orlando has quit IRC18:01
mordreddo we know hwo to avoid falling in to this hole generally?18:01
mordredbecause it sounds like g-r has moved and os-brick hasn't synced/released yet18:01
mordredwhich seems like a thing that might arise in the future as well18:01
lifelessyes: we have wider ranges in project reqs, and use gr for pinning.18:02
lifelessfor the few cases where we must not have variation (linters are a good example) we accept the cost of the dance.18:02
*** mestery has joined #openstack-infra18:03
*** sdake has quit IRC18:05
openstackgerritDarragh Bailey proposed openstack-infra/jenkins-job-builder: Wrap stdin/stdout with codecs to handle unicode
mordredlifeless: so thingee fixes os-brick by syncing and releasing, yeah?18:05
fungithat would be one solution, yes18:06
lifelessmordred: yes18:06
fungiwe could in the future consider a ci job which checks coinstallability though it sounds like pip 8's use in our ci may end up providing that as a side effect anyway18:07
thingeeI just got done with breakfast. Sounds like I need to get back to my room and grab my laptop.18:07
lifelessfungi: it will18:07
lifelessfungi: I want a job that looks for new releases and tries to see if they are coinstallable ... and pass CI, and if so proposes to g-r.18:08
fungiso what we really need to think through is how to avoid or work around deadlocks between the coupled set of repos18:08
mordredyah. I'm mostly poking because I want to make sure we keep understanding what the 'right' behavior and process will be in the pip 8 world18:08
mordredfungi: yah. that18:08
fungican't update x because of y, but can't update y because of x. catch-2218:08
lifelessshould I put something in the etherpad18:08
lifelesswrite a spec18:08
lifelessfungi: yes, I've been harping on that for a while :)18:09
clarkbthingee: I don't think this is breaking anything right now, but it is forward looking "we should fix this"18:09
*** ashleighfarnham has quit IRC18:10
ekarlsofungi: what did you end up with for the bot stuff ?18:10
thingeeclarkb: ok, I will continue to sip my coffee and fix it when I get back to the room.18:10
fungiekarlso: what bot stuff? end up where?18:11
lifelessclarkb: sorry I think you've missed some context18:11
ekarlsofungi: u where looking at alternatives to the bot framework ?18:11
lifelessclarkb: os-brick is not coinstallable with the rest of g-r18:11
lifelessclarkb: I put up a review to remove it from g-r18:11
*** mriedem has joined #openstack-infra18:11
fungiekarlso: i was not. i merely mentioned that some others had looked at a possibility to reimplement our various bots using errbot as a framework18:12
lifelessclarkb: and its entirely able to break stuff 'randomly' right now - anything triggering re-evaluation of os-brick's requirements will downgrade all of teh environment to kilo18:12
mordredekarlso: yes. I believe that errbot is the current thinking18:12
lifeless's versions of os-bricks deps.18:12
*** ociuhandu has quit IRC18:12
ekarlsomordred: heh, I was toying with an idea of doing irc logs > ES18:12
mordredekarlso: we need to do some work towards making configurability of errbot things not be crazy18:13
clarkblifeless: its not worth skipping breakfast over18:13
mordredsince errbot is currently configured via irc commands, and we like to configure things via config files18:13
lifelessclarkb: I agree, but its not 'future looking' either18:13
*** alexpilotti has quit IRC18:15
*** yfried has joined #openstack-infra18:16
mordredekarlso: but the work pabelanger is doing with grafyaml is probably a similar approach we can do with errbot18:17
*** mestery has quit IRC18:18
*** palar has quit IRC18:18
*** mestery has joined #openstack-infra18:19
*** sdake has joined #openstack-infra18:20
*** yfried_ has joined #openstack-infra18:21
*** yfried has quit IRC18:21
*** Krinkle is now known as Krinkle|detached18:26
*** yfried__ has joined #openstack-infra18:26
*** yfried_ has quit IRC18:27
*** rbradfor has joined #openstack-infra18:28
lifelessfungi: mordred: Could you re-review 18393018:33
*** sdake has quit IRC18:34
lifelessahg no18:34
lifelessno point,18:34
lifeless2015-05-17 18:01:25.972 | + python /opt/stack/new/cinder18:34
lifeless2015-05-17 18:01:26.112 | 'os-brick' is not in global-requirements.txt18:34
*** sdake has joined #openstack-infra18:34
lifelesscinder has os-brick in use already, so we need thingee to fix it, I'll narrow my patch to just flake818:34
openstackgerritlifeless proposed openstack/requirements: Fix requirement conflict with flake8.
*** shashankhegde has quit IRC18:38
*** yfried__ has quit IRC18:39
*** rbradfor has quit IRC18:40
*** _nadya_ has quit IRC18:42
clarkbjenkins06 has leaked threds going to put it in shutdown mode18:44
clarkb(just a general FYI)18:44
*** mestery has quit IRC18:46
*** dboik has joined #openstack-infra18:47
electricalclarkb: could i bother you with a jjb question ? ( if you hav etime )18:48
clarkbelectrical: sure18:48
electricalFor some jobs i have a very long exclusion list for the combination-filter part. i would love to store them in jjb as an array and then compile them later as a string so combination-filter can work with them. not sure if jjb has such functionality18:49
electrical < -- example of an exclusion list18:49
clarkbelectrical: JJB in theory knows how to pass around native yaml objects like lists and dicts, but last time I helped someone with it they couldn't get it working18:49
clarkbelectrical: so I think it is likely buggy (and we don't use it)18:50
*** yfried__ has joined #openstack-infra18:50
electricalin the example i got them on separate lines but combination-filter expects it to be a single string.18:50
clarkbelectrical: is the docs for it, you should see if it works for you18:51
electricalAh, will give that a try. thanks!18:51
*** dboik has quit IRC18:52
*** yfried has joined #openstack-infra18:54
*** yfried__ has quit IRC18:55
greghayneslets see if it all breaks again...18:56
*** yamamoto has joined #openstack-infra18:56
clarkbelectrical: curious to know what you find too, mroe info may help in the fixing of potential bugs there18:57
*** rguillebert has joined #openstack-infra18:57
electricalclarkb: okay, will give it a shot and see what happens.18:57
electricalfrom what i understand ( and please correct me if im wrong ) i have to define a variable under my job-group and use {obj:varname} in the job-template?18:58
electricalso far i'm only getting a " expected <block end>, but found '<block mapping start>'" error.18:59
*** _nadya_ has joined #openstack-infra18:59
electricalsorry, that was my fault :-)19:00
electricaltrying again19:00
*** yamamoto has quit IRC19:01
electricaldocumentation is a bit confusing clarkb  :p19:03
clarkbya thats the basic idea of it, you define the var then use it19:03
openstackgerritMerged openstack-infra/project-config: Remove zuul-cloner workaround for DIB
*** fedexo has quit IRC19:04
*** banix has quit IRC19:05
electricalclarkb: okay, got it partially working so far. its able to move the object over but the list i'm giving is parsing the array in a string form. let me make a gist.19:05
electricalclarkb: <--19:06
electricalfirst part is the yaml list i have. second part is the xml it generated19:06
clarkbelectrical: combfilters is just a list of strings there so I think that is expected19:06
clarkbelectrical: distributions: !!python/tuple [precise, jessie] is the definition in the example19:07
clarkbelectrical: that creates a python tuple object with precise and jessie in it19:07
*** enikanorov2 has quit IRC19:07
electricalah okay.19:08
electricalchanged it in the python tuple still messes it up a bit. updating gist19:10
*** zaneb has joined #openstack-infra19:10
electricalclarkb:  gist updated19:11
electricalWould be cool if it just could do an array.join or something in the back :-)19:11
*** otter768 has joined #openstack-infra19:12
clarkbelectrical: its supposed to be simple19:12
electricali could of course store the combination filter as a single string but doesn't make it really manageable.19:12
clarkbelectrical: I am not really sure what you are trying to achieve there but it looks like the combination filter is wrapping your stuff in ()s or {}s19:12
clarkber []s19:13
*** cody-somerville has quit IRC19:13
electricalclarkb: let me try to explain :-)19:13
*** panda|gone has quit IRC19:13
*** sdake has quit IRC19:14
electricali got a whole list of combinations that should be excluded. i would love to have them as a list so i its very clear what is in the combination filter instead of having it as one large string.19:14
*** yamamoto has joined #openstack-infra19:14
*** sdake has joined #openstack-infra19:14
*** enikanorov2 has joined #openstack-infra19:14
*** panda|gone has joined #openstack-infra19:14
electricalso i would love to have an array that i can put in the combination-filter directive and have that array put in there as a string.19:15
clarkbelectrical: so for the second thing I think the module is working against you19:16
clarkbelectrical: beacuse its wrapping your string in []s19:16
clarkbelectrical: that is possibly a bug in the module19:16
electricalYeah indeed19:16
electricalah okay19:16
electricalwill have a look at the module it self of it19:16
*** otter768 has quit IRC19:17
clarkbfor the first thing you should be able to define an array like in your gist and if the module doesn't counter what it is doing have that work19:17
clarkbhowever I would just use a yaml string instead19:18
clarkblet me paste an example for your case19:18
electricalthe module is doing this: str(ex_d.get('combination-filter', '')).rstrip()19:19
clarkb something like that for the string19:19
electricalAh i see, okay19:20
electricaland that would output it as a single string?19:20
clarkbelectrical: yup the yaml > means collapse the newlines in this string19:21
electricalah i see, okay.19:21
*** yamahata has joined #openstack-infra19:21
electricalI'll give that a shot.19:21
clarkbif you use | instead you don't get collapsed newlines19:22
clarkb(two different methods of block quoting based on what you need)19:22
electricalAhh, i see okay. it works now indeed :-)19:24
electrical <-- have this now19:24
electricalatleast its way better readable now whats going on19:25
electricalwithout having to guess or split it up manually19:25
electricalthis will help with showing off the ease of managing our jenkins jobs via jjb :-)19:26
*** mriedem1 has joined #openstack-infra19:28
*** mriedem has quit IRC19:28
*** ZZelle has joined #openstack-infra19:29
*** zb has joined #openstack-infra19:35
*** armax has joined #openstack-infra19:38
*** zaneb has quit IRC19:38
*** achanda_ has joined #openstack-infra19:41
*** achanda has quit IRC19:41
*** zb has quit IRC19:44
*** yamamoto has quit IRC19:47
*** yamamoto has joined #openstack-infra19:48
*** isviridov_away is now known as isviridov19:49
*** enikanorov2 has quit IRC19:52
*** mriedem1 has quit IRC19:57
*** annegentle has quit IRC19:58
*** enikanorov2 has joined #openstack-infra19:58
*** sparkycollier has quit IRC19:59
*** achanda_ has quit IRC20:02
*** woodster_ has joined #openstack-infra20:02
*** zul has quit IRC20:05
*** yamamoto has quit IRC20:06
thingeelifeless, fungi hey I'm at my laptop now. I'm confused about the incompatibility os-brick is having.20:08
thingeealready read through the backlog, but I didn't understand.20:09
clarkbthingee aiui the currebt release has deps that conflict with global requirements20:09
*** yamahata has quit IRC20:09
clarkbthingee so we need a reqs sync and a release20:09
thingeeclarkb: doh right. I see it now by checking out the previous tag release. thanks.20:12
thingeeclarkb: ok so I just need to do a release for you all?20:12
fungiwell, "conflict with global requirements" is really more "conflict with requirements of other released openstack component packages"20:15
fungiso current version of os-brick is not coinstallable with current releases of a number of other openstack libs20:16
fungiand yes, tagging a recent release with a current sync from global-requirements applied should resolve that20:16
openstackgerritMerged openstack-infra/shade: Catch client exceptions during list ops
funginote that this isn't something that was discovered because of ci test failures, but rather discovered through lifeless's work on writing a dependency solver within pip20:17
thingeefungi: how are we going to resolve this issue going forward? Is updating global-requirements going to result in tagging a new release each time?20:17
thingeefungi: just trying to understand better.20:17
fungithingee: i don't think so, i think longer-term one of the things we should be discussing this week is upcoming changes in pip and what openstack can do to accommodate those20:18
*** _david_ has joined #openstack-infra20:18
thingeefungi: ok, is there a session already set for that?20:18
fungiand how we can avoid pointless makework/lockstep updates while also not introducing deadlocks between packages20:18
* thingee is bad.. hasn't really looked at the schedule20:19
*** patrickeast has quit IRC20:19
fungithingee: the global requirements/dependencies session is where we should be discussing that. i haven't memorized the schedule, but we do have at least one ;)20:19
thingeefungi: thanks :)20:19
fungiam trying to pay attention in the board meeting or i'd get you more details now20:20
thingeefungi: sorry, yup ignore me. I'm pushing some other things through and then I'll tag a release for brick.20:20
openstackgerritMerged openstack-dev/cookiecutter: Improve CONTRIBUTING.rst for clarity
*** e0ne has joined #openstack-infra20:23
openstackgerritMerged openstack-dev/hacking: Drop py33 as a target
openstackgerritMerged openstack-dev/hacking: whitelist bash in integration test
*** annegentle has joined #openstack-infra20:29
*** sputnik13 has joined #openstack-infra20:30
electricalclarkb: thanks for the help today. managed to improve the things further. hopefully an other week or so and i have a very nice proof of concept to show some people at Elastic and try to convince them to use it.20:30
_david_fungi, in mail to Gerrit development list you are saying that the version of Gerrit was 2.10.3, with JGit 3.7.2. The JGit version that contained in official 2.10.3 is:
fungizaro: ^ do you have any additional details there?20:32
_david_fungi, zaro looking at downgrade change:
electricalfungi: is there a way to re-trigger the tests ? even if jenkins-python is not released?20:33
_david_fungi, zaro i am downloading the WAR to check, what is inside, ...20:33
*** ashleighfarnham has joined #openstack-infra20:33
fungi_david_: afiak we were running a fork of 2.10.2 where we had selectively backported patches roughly equivalent to 2.10.3 but zaro should be able to clarify once he's around20:33
_david_fungi, someone else reported exactly the same issue, so you guys are not alone:
fungielectrical: i think we just need to go ahead and release a newer jenkins-python with those fixes20:35
electricalfungi: ah okay.20:35
electricalwould love to get the logstash publisher in :-)20:35
_david_fungi, Negative, as i said, mail to dev ML was wrong. JGit 3.7.2 doesn't exist. In war, you have deployed is broken JGit version:20:37
_david_davido@wizball:~/projects/openstack$ jar -tf gerrit-v2. | grep -i jgit-320:37
_david_fungi, This is the regression: bug 465509 in JGit "Massive performance problem in Gerrit caused by ObjectWalk.markUninteresting marking the root tree as uninteresting" [Normal,Resolved: fixed] - Assigned to jgit.core-inbox20:38
fungi_david_: right, we didn't see any performance issues20:38
fungi_david_: i saw no evidence in that bug report that the "performance regression" involved corrupt packfiles20:39
*** _nadya_ has quit IRC20:39
fungior jgit misidentifying correct packfiles as corrupt20:39
_david_fungi, One assumption that when system is under very heavy load, caused by this JGit regression (and in mail to dev ML you are using this wording: "significant load") that then zlib problem occurs.20:40
fungi_david_: that sounds like a reasonable theory20:42
_david_fungi, Sony, Google and we, have seen dramatic performance regression and resource consumption explosion on both: git push and replication operations.20:42
fungi_david_: anyway, since we were under a time crunch and had only just seen the bug report about that performance regression, and the gerrit upstream answer to that was to upgrade to jgit 4 which we hadn't tested at all, we were slightly hesitant (that's why we also wanted to try to run gerrit 2.10.x with jgit 3.2.x20:43
_david_fungi, we hadn't emergency downgrade option, as mordred didn't provide us the SQL downgrade option, so we did emergency revert and replaced broken JGit custom patched, without this regression: all problems gone.20:44
*** ashleighfarnham has quit IRC20:47
_david_fungi, Well, yes, but this version is needed, to void JGit downgrade, as explained in this change:
*** ashleighfarnham has joined #openstack-infra20:47
*** ashleighfarnham has quit IRC20:47
fungiright, we saw that20:47
*** ashleighfarnham has joined #openstack-infra20:48
*** ashleighfarnham has quit IRC20:48
fungiit nor any of the bugs/discussions linked from it mentioned the specific error messages we were seeing20:48
*** julim has joined #openstack-infra20:48
fungiso it wasn't immediately obvious to us that any of what we experienced would have been solved in jgit 420:49
fungii still haven't heard any definitive proof that it would20:49
fungiexcept for the theory that zlib ends up with buffer problems under performance-regressed load20:49
_david_One interesting question, which patch level the other guys are running that reported exactly the same problem20:50
fungii agree it's worth experimenting to see if we can reproduce our problem in isolation and then test to see if that solves it20:50
fungi_david_: absolutely--thank you for seeing that thread and asking them. i'd love to know too20:50
_david_fungi, But you wasn't able to repro the problem on review-dev?20:51
*** Krinkle|detached is now known as Krinkle20:51
fungi_david_: it didn't show up during our light-load tests on review-dev but we need to reengineer how review-dev is set up and come up with some representative load testing mechanisms. likely it'll be a few weeks before we're in a good position to reproduce the failure we saw that way20:52
fungiespecially since most of us are busy with our conference all this week20:53
*** e0ne is now known as e0ne_20:53
*** isviridov is now known as isviridov_away20:54
_david_fungi, I see, and no other attempts to upgrade to 2.10.x is going to happen, before you identified the root cause for the failure i guess?20:54
*** banix has joined #openstack-infra20:54
*** mestery has joined #openstack-infra20:55
*** e0ne_ is now known as e0ne20:55
*** reed_ has joined #openstack-infra20:57
_david_fungi, I am opening JGit bug as well, i guess you don't have any problem when I link in the report?20:58
*** teran has joined #openstack-infra20:59
fungi_david_: feel free to link it wherever. we wouldn't have put it on a public webserver if we were concerned about its contents being sensitive20:59
fungi_david_: and correct, we aren't currently planning to try upgrading gerrit again until we can reproduce the errors we saw and be able to test what does or doesn't fix it21:00
fungioutages of our code review system are very expensive to our projects, in terms of wasted developer time21:01
*** sks has quit IRC21:02
*** achanda has joined #openstack-infra21:02
*** rossella_s has joined #openstack-infra21:03
*** armax has quit IRC21:04
*** mestery has quit IRC21:04
*** ddieterl_ has quit IRC21:05
*** achanda has quit IRC21:08
*** julim has quit IRC21:08
*** otter768 has joined #openstack-infra21:13
*** annegentle has quit IRC21:14
*** otter768 has quit IRC21:18
*** dboik has joined #openstack-infra21:23
*** rossella_s has quit IRC21:23
*** dboik_ has joined #openstack-infra21:24
*** rlandy has quit IRC21:24
*** e0ne has quit IRC21:26
*** ajo has joined #openstack-infra21:26
*** dboik has quit IRC21:28
*** _david_ has quit IRC21:34
*** julim has joined #openstack-infra21:36
*** ajo has quit IRC21:38
*** sputnik13 has quit IRC21:40
*** sarob has joined #openstack-infra21:42
*** jaypipes has joined #openstack-infra21:44
*** annegentle has joined #openstack-infra21:45
lifelessthingee: fungi: yes, g-r changes can cause this. Most of the requirements are not pins but ranges though.21:48
lifelessthingee: fungi: so the condition is that there must be *a* version with in g-r's specs that overlaps the specs of everything else at every point in time.21:48
*** fedexo has joined #openstack-infra21:48
lifelessthingee: so with todays methodoloy: g-r expand, sync to projects, release in projects, g-r contract, sync to projects, release in projects.21:49
*** breton has left #openstack-infra21:51
*** salv-orlando has joined #openstack-infra21:57
*** julim has quit IRC22:19
*** ZZelle has quit IRC22:21
*** fedexo has quit IRC22:24
*** fallenpe_ has joined #openstack-infra22:28
*** palar has joined #openstack-infra22:30
*** woodster_ has quit IRC22:30
*** Krinkle is now known as Krinkle|detached22:41
*** banix has quit IRC22:45
*** dvorak has quit IRC22:47
*** achanda has joined #openstack-infra22:47
*** jaypipes has quit IRC22:50
*** yamahata has joined #openstack-infra22:50
*** jaypipes has joined #openstack-infra22:53
*** boris-42 has joined #openstack-infra22:56
*** yamahata has quit IRC23:03
*** dvorak has joined #openstack-infra23:04
*** rbradfor has joined #openstack-infra23:07
*** sdake has quit IRC23:08
*** sdake has joined #openstack-infra23:08
*** julim has joined #openstack-infra23:11
*** Krinkle|detached is now known as Krinkle23:12
*** julim has quit IRC23:13
*** otter768 has joined #openstack-infra23:14
*** panda|gone has quit IRC23:14
*** panda|gone has joined #openstack-infra23:14
*** reed_ has quit IRC23:15
*** otter768 has quit IRC23:18
*** jaypipes has quit IRC23:18
*** julim has joined #openstack-infra23:21
Kennanit seems the infra still not fix the auth_url issue for cliutils.env('OS_AUTH_URL') in jenkins run env ?23:28
KennanI remembered someone talked about this serveral days ago?23:28
clarkbKennan I am completely unaware of anything related to that. have any more info23:30
Kennanthis is our jenkins job failure
clarkbthat doesnt look like ab infra issue23:33
clarkbif magnum tests need to set up auth they should do that23:33
Kennanclarkb: the code just get it from cliutils.env('OS_AUTH_URL')23:35
clarkbthats a devstack setting23:36
clarkbbut it looks like tge test needs admin and isnt getting it?23:37
*** yamamoto has joined #openstack-infra23:38
Kennanclarkb if I understood it correctly python code cliutils.env('OS_AUTH_URL') not need any admin23:38
clarkboh tge admin warning isnt importabt23:39
clarkbyou should check with devstack, OS_AUTH_URL is not an infra thing23:39
Kennanok clarkb I not found devstack channel in iRC23:40
clarkbopenstack-qa is probably the best place for that23:41
*** chlong has joined #openstack-infra23:49
*** chlong has quit IRC23:52
*** dalgaaf has joined #openstack-infra23:54
*** rbradfor42 has joined #openstack-infra23:54
*** fallenpegasus has quit IRC23:56
*** julim has quit IRC23:57
*** annegentle has quit IRC23:57
*** sarob has quit IRC23:58
*** rbradfor42 has quit IRC23:58
*** sarob has joined #openstack-infra23:59

Generated by 2.14.0 by Marius Gedminas - find it at!