Wednesday, 2016-05-04

*** egafford has quit IRC00:03
*** egafford1 has joined #tripleo00:03
*** egafford1 has quit IRC00:06
*** egafford has joined #tripleo00:06
*** egafford has quit IRC00:10
*** fragatin_ has joined #tripleo00:11
*** fragatin_ has quit IRC00:14
*** fragatina has quit IRC00:14
*** fragatina has joined #tripleo00:15
*** noslzzp_ has joined #tripleo00:32
*** noslzzp has quit IRC00:35
*** dmacpher has joined #tripleo00:35
*** yamahata has joined #tripleo00:39
*** weshay has joined #tripleo00:42
*** hewbrocca is now known as hewbrocca-afk00:44
*** rbrady has quit IRC00:55
*** saneax_AFK is now known as saneax01:02
*** fragatin_ has joined #tripleo01:12
*** fragatin_ has quit IRC01:14
*** fragatina has quit IRC01:16
*** vkmc is now known as Guest987401:20
*** Guest9874 is now known as cmkv01:22
*** cmkv is now known as violenciarivas01:22
*** violenciarivas is now known as vkmc01:23
*** noslzzp_ has quit IRC01:30
*** yamahata has quit IRC01:38
*** mburned is now known as mburned_out01:45
*** thrash is now known as thrash|g0ne01:45
*** anshul has joined #tripleo01:58
*** anshul is now known as Guest3271501:59
*** Guest32715 has quit IRC02:03
*** jaosorior has joined #tripleo02:09
jaosoriorbnemec: Still around?02:12
*** morazi has quit IRC02:12
*** yamahata has joined #tripleo02:12
jaosoriorHey guys, any +A for this? https://review.openstack.org/#/c/311833/02:13
*** abehl has joined #tripleo02:19
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: DO NOT MERGE - testing  https://review.openstack.org/31230602:21
*** ebalduf has quit IRC02:22
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: DO NOT MERGE - testing  https://review.openstack.org/31230702:25
*** fragatina has joined #tripleo02:31
jaosoriorEmilienM: Hey dude, one question regarding the composable roles: Currently I have this patch https://review.openstack.org/#/c/304125/ that creates the keystone endpoints and service users in puppet, in the kesytone profile02:33
jaosoriorhowever, it depends on the hieradata passed in this patch https://review.openstack.org/#/c/244162/02:34
jaosoriorwhich passes the endpoints on the puppet/controller.yaml file... should I move those to the keystone profile itself? Seems a bit weird, since it depends on the endpoints of other services02:34
*** ebalduf has joined #tripleo02:37
EmilienMjaosorior: you have a circular dependency02:49
jaosoriorI do? damn...02:50
jaosoriorEmilienM: Well, I gotta fix that. But also, should I leave the hieradata in controller.yaml?02:51
jaosoriorEmilienM: You mean circular dependency in puppet, right?02:51
EmilienMjaosorior: no, in zuul02:51
EmilienMyou have depends-on the other patch and same for the other patch02:51
EmilienM1 needs 2 but 2 needs 102:52
EmilienMit can't work02:52
EmilienMzuul can't resolve that02:52
EmilienMfor the hieradata, I don't know02:52
EmilienMI have the same issue in my patches and I don't know how to solve that02:52
jaosoriorEmilienM: Are you sure about the circular dependency?02:52
EmilienMlet me check02:52
EmilienMah no you're right02:53
EmilienMsorry I misread02:53
jaosoriornice02:53
jaosorior:D02:53
EmilienMI'm off to bed, I'll continue tomorrow02:53
jaosoriorEmilienM: have a good one! I forget how late it must be over there02:53
EmilienMnot too bad, only 11pm but tomorrow I want to start early :)02:54
jaosoriorit's almost 6am here, I'm extremely jetlagged02:54
EmilienMahah good luck man02:54
jaosoriorcoffee is my friend02:54
jaosorioranyway, sleep well Mr.!02:54
EmilienMthx02:54
*** saneax is now known as saneax_AFK03:02
*** cwolferh has quit IRC03:05
*** saneax_AFK is now known as saneax03:10
*** TSCHAK_ has joined #tripleo03:15
*** abehl has quit IRC03:16
*** TSCHAK has quit IRC03:18
*** cwolferh has joined #tripleo03:28
*** jaosorior has quit IRC03:51
*** myoung has joined #tripleo03:54
*** fragatina has quit IRC03:56
*** fragatina has joined #tripleo03:56
*** jaosorior has joined #tripleo04:06
*** liverpooler has quit IRC04:09
*** masco has joined #tripleo04:11
*** myoung has quit IRC04:15
*** myoung has joined #tripleo04:16
*** myoung has quit IRC04:17
*** shivrao has quit IRC04:17
*** myoung has joined #tripleo04:17
*** links has joined #tripleo04:25
*** dmacpher has quit IRC04:27
*** tzumainn has quit IRC04:31
*** myoung has quit IRC04:38
*** myoung has joined #tripleo04:40
*** myoung has quit IRC04:42
*** myoung has joined #tripleo04:42
*** dmacpher has joined #tripleo04:44
*** saneax is now known as saneax_AFK05:01
*** myoung has quit IRC05:04
*** myoung has joined #tripleo05:05
*** dmacpher has quit IRC05:06
*** shivrao has joined #tripleo05:11
*** shivrao_ has joined #tripleo05:12
*** saneax_AFK is now known as saneax05:13
*** fragatina has quit IRC05:15
*** shivrao has quit IRC05:16
*** shivrao_ is now known as shivrao05:16
*** bnemec has quit IRC05:17
*** dmacpher has joined #tripleo05:19
openstackgerritColleen Murphy proposed openstack/diskimage-builder: Fix OpenSUSE support  https://review.openstack.org/31232505:28
*** ayoung has quit IRC05:31
*** mnaser has quit IRC05:41
*** florianf has joined #tripleo05:42
*** oshvartz has joined #tripleo05:47
*** liverpooler has joined #tripleo05:55
*** cwolferh has quit IRC05:55
*** shivrao_ has joined #tripleo05:57
*** shivrao has quit IRC05:58
*** shivrao_ is now known as shivrao05:58
*** florianf has quit IRC05:58
*** florianf has joined #tripleo05:59
*** mnaser has joined #tripleo06:00
*** david-lyle has quit IRC06:05
*** jrist has quit IRC06:06
*** jaosorior_ has joined #tripleo06:07
*** jaosorior has quit IRC06:07
*** jaosorior_ has quit IRC06:13
*** aufi has joined #tripleo06:13
*** jaosorior_ has joined #tripleo06:17
*** openstackgerrit has quit IRC06:17
*** openstackgerrit has joined #tripleo06:18
*** TSCHAK_ has quit IRC06:18
jaosorior_marios: Online yet?06:19
*** jaosorior_ is now known as jaosorior06:20
mariossup jaosorior06:20
jaosoriormarios: Hey dude, how's it going?06:20
jaosoriormarios: Hey dude, how acquainted are you with the upgrades gate?06:20
mariosjaosorior: not bad thanks how about yourself? Is there a ci problem06:21
jaosoriormarios: Thing is, I have this commit https://review.openstack.org/#/c/304125/ which does the creation of the keystone services, service users and endpoints via puppet. And for some reason it's passing non-ha and ha gates. But there is some weird issue in the upgrades one06:22
mariosjaosorior: jobs failing? ... looking at the review06:22
jaosoriormarios: All good here, dealing with the jetlag06:22
mariosjaosorior: actually, adding to my review list, going to get to it momentarily06:22
*** ccamacho has joined #tripleo06:22
jaosoriormarios: Yeah, the funky thing is that it fails because it says that it's missing the password for the aodh::keystone::auth module... but that should be there. It works for the other gates06:23
mariosjaosorior: yeah same man. actually last night was the first 'proper' night's sleep i've had since back so feeling better06:23
jaosoriormarios: Now looking at the hieradata... almost all the parameters for <service>::keystone::auth are missing (except neutron)06:23
jaosoriorcan't figure out what's up06:23
jaosoriorccamacho: hey dude06:24
jaosoriorccamacho: I might need to revert this CR https://review.openstack.org/#/c/309039/ :/06:24
*** TSCHAK has joined #tripleo06:25
openstackgerritJuan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Revert "Finished: Update deprecated commands in tripleo.sh (Old heat * calls)"  https://review.openstack.org/31232906:27
*** tremble has joined #tripleo06:29
*** dmacpher has quit IRC06:35
*** hewbrocca-afk is now known as hewbrocca06:35
bandinimorning *06:44
*** shivrao has quit IRC06:44
jaosoriorhey bandini, what's up dude?06:45
ccamachojaosorior, hello.06:45
ccamachofound an issue?06:45
jaosoriorccamacho yeah dude, proposed a revert here: https://review.openstack.org/#/c/312329/ and the bug attached to it has more info06:46
jaosoriortrown|outtypewww proposed this fix https://review.openstack.org/#/c/312212/ but it addressed stable/mitaka only. Seems that stable/liberty still fails with it06:47
bandinijaosorior: hola hola. not much, getting accustomed to a proper timezone ;) how about you?06:47
bandinimarios: ANJ says hello btw06:48
* marios holds back tears06:48
bandinilol06:49
jaosoriorbandini: Holy crap dude, today I woke up at 5am. jetlag sucks06:49
bandiniI know right06:49
*** ebalduf has quit IRC06:50
*** paramite has joined #tripleo06:52
openstackgerritMin Pae proposed openstack/tripleo-image-elements: adding libffi-dev to required packages  https://review.openstack.org/31233606:53
*** shivrao has joined #tripleo06:54
ccamachojaosorior, what if we run the command as "openstack stack output show tenant-stack server1_public_ip -c output_value -f value"06:54
ccamachono need for grep06:54
jaosoriorccamacho: That is what trown's fix is about: https://review.openstack.org/#/c/312212/1/scripts/tripleo.sh and that didn't work06:56
*** jprovazn has joined #tripleo06:56
jaosoriorccamacho: This is how it looks like in stable/liberty https://review.openstack.org/#/c/312307/06:56
ccamachommm I see, well I will check on that to see whats happening06:57
jaosoriorccamacho: If you can come up with a fix that would be cool :D meanwhile, I think that revert should land, and then you can re-submit with the fix included06:58
ccamachoyeahp lets keep it in that way, I will find out why is not working correctly06:59
*** devvesa has joined #tripleo06:59
*** devvesa has quit IRC06:59
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: DO NOT MERGE: Testing revert  https://review.openstack.org/31233907:00
*** tesseract has joined #tripleo07:01
*** tesseract is now known as Guest2331907:01
*** leanderthal|afk is now known as leanderthal07:04
*** devvesa has joined #tripleo07:05
openstackgerritMin Pae proposed openstack/tripleo-image-elements: making ansible version configurable  https://review.openstack.org/31234207:08
ccamachojaosorior, quick question, I think what is not working properly is the validation with the $STABLE_RELEASE as If you are executing tripleo.sh with liberty you should use the HEAT CLI instead the openstack one.   Thats the reason why the bash snippet to swap between liberty and mitaka.  Do you know how to get the value of $STABLE_RELEASE from the CI gates? to check whats there07:09
jaosoriorccamacho: Well, STABLE_RELEASE is usually the way dude07:10
jaosorioror at least I've used that before07:10
ccamachoI will check on that07:12
jaosoriorccamacho: So STABLE_RELEASE is set by toci_instack.sh , and it should be set up to something... one quick check you could do is try to output the value or output something inside the check you already have07:13
jaosoriorccamacho: Another thing might be that it's not actually getting to the function and the CI is actually still using the openstack command07:13
ccamachothanks!07:13
jaosoriorwhich seems to me like the most likely issue07:14
ccamachoyeahp that should be the problem, np, I will check it :)07:14
jaosoriorccamacho: maybe it would be better to use a name other than openstack for that function. Something like openstack_wrapper or something of the sort07:15
*** dsariel has joined #tripleo07:16
ccamachojaosorior, at first I though that, but if we keep the openstack name, in november just by removing the funcion (TODO message) no more changes should be needed07:16
ccamachoLEt me check the commands, the validations and I will cherry pick the patch to test it in liberty07:17
ccamachoto see the behaviour07:17
jaosoriorccamacho: Well, the revert hasn't landed yet07:19
*** jrist has joined #tripleo07:20
*** ifarkas has joined #tripleo07:22
*** jpena|off is now known as jpena07:23
*** apetrich has joined #tripleo07:27
*** hewbrocca is now known as hewbrocca-afk07:28
*** zoli_gone-proxy is now known as zoliXXL07:29
mariosjaosorior: wrt https://review.openstack.org/#/c/304125/6 where did you see the aodh issue07:29
mariosam poking at the logs for the failed upgrades job07:29
*** hewbrocca-afk is now known as hewbrocca07:29
mariosjaosorior: on controller i see http://paste.openstack.org/raw/496028/07:30
jaosoriormarios: Here in the postci.txt logs http://logs.openstack.org/25/304125/6/check-tripleo/gate-tripleo-ci-f22-upgrades/48f6a15/logs/postci.txt.gz07:30
jaosoriormarios: whaaa, that's weird07:32
mariosjaosorior: yeah thanks. in fact it is the same trace (I didn't copy enough of it)07:32
marioshttp://paste.openstack.org/raw/496029/07:32
jaosoriormarios: And if you go to the controller, and check /etc/puppet/hieradata/controller.yaml you will notice the password is indeed missing... but I have no idea why07:33
jaosoriormarios: Since it's included and it works in the non-ha and the ha runs07:33
*** jaosorior has quit IRC07:37
*** jaosorior has joined #tripleo07:38
*** jrist has quit IRC07:38
openstackgerritCarlos Camacho proposed openstack-infra/tripleo-ci: Fix validation for stable branch  https://review.openstack.org/31235007:39
*** mikelk has joined #tripleo07:39
jaosoriormarios: Uhm... is the upgrades job only taking into account changes in t-h-t?07:40
mariosjaosorior: i am not sure i didn't work on that at all07:41
mariosjaosorior: gfidente would be a good person to ask07:42
jaosorioralright, gotta wait for him to get online I get07:42
jaosorior*I guess07:42
jaosoriormarios: But yeah, one thing is that the aodh password is introduced in a t-h-t patch on which that puppet-tripleo patch depends on07:43
mariosjaosorior: would explain it if that is the case though ...07:43
mariosyeah07:43
jaosoriorso I guess it fails cause it first tries to apply the t-h-t patch without the password... and that's when it fails. Since it applies with still using the new puppet-tripleo code07:43
openstackgerritCarlos Camacho proposed openstack-infra/tripleo-ci: Fix validation for stable branch  https://review.openstack.org/31235007:44
mariosjaosorior: thinking about gfidente comment a bit07:45
mariosjaosorior: for moving the endpoint creation to the per service manifests07:45
*** shivrao_ has joined #tripleo07:45
jaosoriormarios: That's gonna get complicated :/07:46
jaosoriorsince all of that stuff depends on parts that are introduced by that keystone manifest07:46
jaosoriormarios: we're gonna end up introducing yet another step just to make sure that the keystone one is applied first07:47
jaosoriormarios: I do think that we should try to make that work the way it is. And once it's working, look into breaking it into the other profiles if it's possible07:47
jaosoriorbut right now, it's still conforming to the pattern of composable roles07:48
openstackgerritCarlos Camacho proposed openstack-infra/tripleo-ci: Fix validation for stable branch  https://review.openstack.org/31235007:48
*** shivrao has quit IRC07:49
*** shivrao_ is now known as shivrao07:49
*** ayoung has joined #tripleo07:49
ccamachojaosorior, I have validated the change for liberty branch, I will see the value for STABLE_RELEASE when CI executes the job to see if the issue was the validation.07:50
jaosoriorccamacho: awesome, lets hope that works07:50
ccamachonice07:50
mariosjaosorior: why can't we have a 'if $step >=6 " in each of the roles, like in cinder for example just have that +    Pacemaker::Resource::Service[$::apache::params::service_name] -> Class['::cinder::keystone::auth']07:50
ccamachoIll let you know07:50
mariosjaosorior: as you have it already07:50
mariosjaosorior: just thinking that it will greatly increase composability07:50
mariosjaosorior: ok so we'll need a manage_endpoint for each service too...07:52
mariosjaosorior:07:53
jaosoriormarios: Because we need to make sure that the keystone::roles::admin and keystone::endpoint run first. Else it breaks (or it used to at least...). And with the composable roles we don't have assured. And yeah, we also need that manage_endpoint parameter07:53
mariosjaosorior: k i think it is worth thinking about at least, but since things are still fluid around how we are structuring these I guess we can land like this as a first pass07:53
jaosoriormarios: But yeah, I think once it works properly, we can iterate to make it better according to the pattern07:54
*** dmk0202 has joined #tripleo07:54
mariosjaosorior: since we are trying to make each role definition as self contained as possible right? it makes sense for the endpoint creation for each to be pulled together from each role rather than 'hardcoded'  in the keystone role07:55
*** shardy has joined #tripleo07:56
jaosoriormarios: It does make sense indeed. But, like I said, first I would like to get it working before starting to refactor the whole thing again (note that one of the patches is in patchset 38...been through a lot of iterations and rebases)07:59
*** zoliXXL is now known as zoli|afk07:59
jaosoriormarios: So I'm more concerned about the upgrade thing... What would you think is a good approach? should I do a hiera call to check if the parameters are being passed?08:00
*** jaosorior_ has joined #tripleo08:04
jaosorior_marios: Got an idea08:04
*** mcornea has joined #tripleo08:05
jaosorior_marios: what if manage_endpoint defaults to false... and then it should be introduced to be set to true in the t-h-t patch08:05
jaosorior_either way, right now os-cloud-config is managing the endpoints, so it shouldn't have been set to 'true' in the first place08:05
*** stendulker has joined #tripleo08:05
*** jaosorior has quit IRC08:07
mariosjaosorior_: don't think it is needed since as you say the ha job is passing there08:09
*** olap has quit IRC08:10
mariosjaosorior_: (and you checked the hiera already i think you mentinoed)08:10
*** yamahata has quit IRC08:11
jaosorior_marios: yeah, but that will force the keystone profile to not run that part of the code if the parameters are not there, which is why the upgrades case fails08:11
*** jaosorior_ is now known as jaosorior08:11
*** mbound has joined #tripleo08:11
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Default keystone manage_endpoint to false  https://review.openstack.org/31237308:19
*** ohamada has joined #tripleo08:21
*** myoung has quit IRC08:21
*** jistr has joined #tripleo08:24
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Pass parameters to manage endpoints via puppet  https://review.openstack.org/24416208:27
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Create keystone endpoints in keystone profile  https://review.openstack.org/30412508:29
*** jrist has joined #tripleo08:33
*** gfidente has joined #tripleo08:33
*** gfidente has quit IRC08:33
*** gfidente has joined #tripleo08:33
*** zoli|afk is now known as zoli08:33
*** abehl has joined #tripleo08:36
*** mgould has joined #tripleo08:40
*** chem has joined #tripleo08:41
*** shivrao has quit IRC08:42
*** yamahata has joined #tripleo08:46
*** saneax is now known as saneax_AFK08:54
*** derekh has joined #tripleo08:56
derekhFound the reason last night that our periodical job was failing, we couldn't upload the instack image to the cache08:58
derekhas we were deleting it befor the upload08:58
derekhhttps://review.openstack.org/#/c/312281/108:58
derekhif anybody has a second to take a quick look at that ^08:58
* jistr looks08:58
gfidentederekh, why we wanted to delete the image btw?09:00
gfidenteor the big repos as it happens in the new submission09:00
*** olap has joined #tripleo09:00
derekhgfidente: because the jenkins slaves were running out of space09:00
gfidenteI mean, where is it that these are consuming space?09:00
gfidentethat's because it doesn't cleanup the environment after the buld is finished right?09:01
derekhgfidente: on the jenkins slave09:01
*** electrofelix has joined #tripleo09:01
gfidentebut it should be emptying it one a new build starts no?09:01
derekhgfidente: no, we were fairly close to running out of space but everything was ok until we started downloading and unpacking things in from the cache09:02
derekhgfidente: that push us over the limit09:02
derekhgfidente: so we have two choices, 1. delete a few things or 2. use a bigger flavor09:03
derekhgfidente: I'm going for the delete stuff option, so that the host machines use up less space09:03
jistrthe /opt/git/... things we're deleting are pre-seeded into the machine automatically due to some openstack-infra convention? (to make sure jobs don't have to clone by themselves, we have all openstack/ and openstack-infra/ repos pre-cloned?)09:04
jistr(i wonder why are all these git repos there in the first place...)09:06
derekhjistr: gfidente every night infra build an image on our cloud with ALL openstack repositories in /opt/git09:14
derekhjistr: gfidente and they snapshot it to use if for the next days jobs09:15
openstackgerritMerged openstack-infra/tripleo-ci: Don't remove the instack.qcow2 image  https://review.openstack.org/31228109:15
derekhjistr: gfidente then at the start of each job the projects that are relevant to us are cloned from here into /opt/stack/new09:15
derekhand refreshed09:16
gfidenteoh so we want some of them09:16
jistrderekh: cool, thanks09:17
derekhgfidente: yup09:18
derekhgfidente: I could have removed more but those were the bigest of the ones we didn't need09:18
gfidenteunderstood, thanks09:18
openstackgerritCarlos Camacho proposed openstack/puppet-tripleo: WIP: Puppet manifest to configure NTP as a composable service  https://review.openstack.org/31072509:19
derekhgfidente: no prob09:20
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Use the overcloud full images from the cache  https://review.openstack.org/31239709:21
openstackgerritCarlos Camacho proposed openstack/tripleo-heat-templates: WIP: Composable roles NTP  https://review.openstack.org/31042109:22
derekhFYI, our repo pin updated last night also, so our base packages are now from yesterday09:23
ccamachoderekh, thanks for the info09:29
derekhccamacho: no prob09:29
openstackgerritMerged openstack/python-tripleoclient: Trove integration  https://review.openstack.org/23324109:30
*** saneax_AFK is now known as saneax09:30
gfidentemichchap_, so this https://review.openstack.org/#/c/309450/8 is now replicating the tht behaviour09:32
jistranyone to +A a backport here? https://review.openstack.org/#/c/306391/09:32
gfidentemichchap_, looks good to you?09:32
jistrit would allow me to land a different patch to master09:32
jistrb/c of depends-on not being branch aware09:32
gfidentejistr, in return https://review.openstack.org/#/c/309450/ ?09:33
jistrgfidente: you are a skilled trader09:33
* jistr looks09:33
gfidenteoh09:33
jistrah yea i saw that one09:33
jistri reviewed the t-h-t counterpart only though, will do this one too09:33
*** mbound has quit IRC09:35
gfidentejistr, it's just meant to replicate what the main manifests do09:35
gfidenteexcept it assumes sync_db true otherwise it never runs in non-ha09:35
gfidentebut that is what happens in current non-ha manifest as well, we just don't consider sync_db09:36
openstackgerritMerged openstack/python-tripleoclient: Unset UpdateIdentifier on deploy/scale, pass in StackAction  https://review.openstack.org/30639109:36
*** sshnaidm has quit IRC09:37
openstackgerritMerged openstack/tripleo-heat-templates: Use str_split to compute netmask in _v6 port templates  https://review.openstack.org/30783809:37
jistrgfidente: cool re https://review.openstack.org/#/c/309450 -- very nice that including of base from pacemaker classes is no longer done in a "if step > N" wrapping09:42
*** florianf has quit IRC09:43
gfidenteyeah I removed that09:43
gfidentecause it gets step anyway09:43
gfidenteprobably with some more work there we could diminish the pacemaker manifests to mostly pcmkr related things09:44
gfidentefor example09:46
gfidenteI think we could pass those enabled and manage_service from the pcmk specific yaml now09:46
gfidenteto the service class, and don't have any logic for that into the role .pp09:47
*** tosky has joined #tripleo09:50
openstackgerritMerged openstack/puppet-tripleo: Move databases creation and sync with the role  https://review.openstack.org/30945009:50
*** mbound has joined #tripleo09:51
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Noop start/stop/restart action for Glance and Keystone in the roles  https://review.openstack.org/31093509:53
*** athomas has quit IRC09:55
*** florianf has joined #tripleo09:56
*** fragatina has joined #tripleo09:59
*** dmk0202 has quit IRC09:59
*** cmyster has quit IRC10:01
*** jpena is now known as jpena|brb10:01
openstackgerritMerged openstack/tripleo-heat-templates: Fix distinguishing between stack-create and stack-update  https://review.openstack.org/30409410:01
*** jrist has quit IRC10:02
*** cmyster has joined #tripleo10:02
*** athomas has joined #tripleo10:02
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: Fix distinguishing between stack-create and stack-update  https://review.openstack.org/31242010:03
shardymarios: Hey, looking at https://review.openstack.org/#/c/311860/1 - you only +1'd it but it now has a +2 from jistr, are you ready for that to land?10:13
mariosshardy: looking10:15
mariosshardy: yeah that was mainly waiting for testing first10:16
mariosshardy: and jistr ran a sanity check with that which looks good10:17
shardymarios: Ok, thanks - it lgtm but I didn't want to approve if you were planning additional testing10:17
mariosshardy: thanks for checking10:19
*** sshnaidm has joined #tripleo10:24
openstackgerritMerged openstack/puppet-tripleo: Noop start/stop/restart action for Glance and Keystone in the roles  https://review.openstack.org/31093510:30
*** myoung has joined #tripleo10:31
openstackgerritCarlos Camacho proposed openstack-infra/tripleo-ci: Fix validation for stable branch  https://review.openstack.org/31235010:35
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add neutron profiles  https://review.openstack.org/29343610:35
openstackgerritMarios Andreou proposed openstack/instack-undercloud: Add the OS_MISTRAL_URL to the undercloud stackrc  https://review.openstack.org/31243710:35
openstackgerritCarlos Camacho proposed openstack/tripleo-heat-templates: Testing purposes do not merge  https://review.openstack.org/31244010:39
openstackgerritCarlos Camacho proposed openstack/tripleo-heat-templates: Testing purposes do not merge  https://review.openstack.org/31244110:39
openstackgerritCarlos Camacho proposed openstack/tripleo-heat-templates: Testing purposes do not merge  https://review.openstack.org/31244210:39
*** mburned_out is now known as mburned10:42
gfidentemichchap_, commented on https://review.openstack.org/#/c/293436/1810:43
gfidenteI just did that for glance/keystone and I am going to update it in cinder and trove10:44
gfidentenice split of the neutron manifest! :)10:44
*** links has quit IRC10:45
*** jrist has joined #tripleo10:45
*** zoli is now known as zoli|lunch10:51
trown|outtypewwwccamacho: is there another patch depending on https://review.openstack.org/#/c/31235010:51
trown|outtypewwwccamacho: we are not testing stable branches against changes to tripleoci, which is how we got in this situation to begin with10:52
*** stendulker has quit IRC10:52
trown|outtypewwwI would really like to merge https://review.openstack.org/#/c/312329/ until we get stable branch CI setup with https://review.openstack.org/#/c/297390/10:53
ccamachoyour are right, the thing is that in liberty should not be used openstack heat commands, is validated and testing it.10:53
ccamachofor me is OK the revert10:54
ccamachountil I get the output from stable/liberty in THT10:54
*** myoung|remote has joined #tripleo10:54
*** myoung has quit IRC10:54
ccamachoand the revert patch is correct10:55
*** lucasagomes is now known as lucas-hungry10:56
*** weshay has quit IRC10:56
*** david-lyle has joined #tripleo10:58
trown|outtypewwwderekh: could you look at https://review.openstack.org/#/c/312329/ all tripleo CI in RDO including tripleo-quickstart gate is broken10:58
derekhtrown|outtypewww: looking10:58
shardyccamacho: Hi!  Looking at https://bugs.launchpad.net/tripleo/+bug/1577903 and trying to figure out if a revert of the switch from heat to OSC CLI is the quickest fix11:03
openstackLaunchpad bug 1577903 in tripleo "[CI][Stable] Patch to fix deprecated heat commands broke stable branch CI" [Critical,Triaged]11:03
shardyinterested in your feedback as to if there's a quick fix that avoids that11:03
trown|outtypewwwshardy: ya https://review.openstack.org/#/c/312329/11:03
shardytrown|outtypewww: aha, it's not flipped the LP bug state11:04
shardythanks11:04
ccamachoderekh, the revert should work as the problem is the validation with the liberty branch, please go with the revert and I will fix the issue in the next minutes. THis in order to not having the gates broken11:04
derekhtrown|outtypewww: ccamacho its already gotten +A weird11:05
derekhtrown|outtypewww: ccamacho wronge patch sorry11:05
michchap_gfidente: Yep I think your comment is correct, though I'm not sure how non-ha works right now since sync_db defaults to false in neutron::server, so that's a mystery to me.11:05
openstackgerritMerged openstack-infra/tripleo-ci: Revert "Finished: Update deprecated commands in tripleo.sh (Old heat * calls)"  https://review.openstack.org/31232911:05
shardyccamacho: thanks, I just approved it11:05
trown|outtypewwwthanks folks!11:06
derekhshardy bet me to it11:06
shardyhehe :)11:06
trown|outtypewwwderekh: nice one on the tripleo promote last night, RDO is still really far behind11:06
gfidentemichchap_, yeah not sure why it defaults to false indeed11:06
gfidentebut it works because we still have11:06
gfidenteneutron::server::sync_db: true11:06
gfidentein controller.yaml11:06
gfidenteI need to make same split as your for cinder btw11:07
derekhtrown|outtypewww: thanks, the reason we hadn't promoted in the last few days is because we couldn't upload the instack image to the cahce11:07
michchap_gfidente: oh right. I guess it's more clear having true in the manifest11:07
derekhtrown|outtypewww: mainly because I had deleted it ....11:07
ccamachoshardy, derekh, jaosorior. thanks! and sorry for the broken CI I will test accordingly the patch with all versions (master stable/liberty and stable/mitaka)11:07
gfidentemichchap_, to be honest, I am not sure why it isn't true in the module itself as it happens for the others11:08
trown|outtypewwwccamacho: not your fault... we need stable branch CI on tripleoci repo11:08
derekhccamacho: np, my delay in approving the revert was because I was testing it to find the problem11:08
gfidenteccamacho and fwiw, breaking CI is a step everybody has to get into11:08
gfidenteimho :)11:08
derekhccamacho: looks like stack create isn't correct11:09
derekhccamacho: passing in this to the openstack function "openstack stack create asd asfd afa  sadfsd asdas asfas fa"11:09
gfidenteit's one of those things which reminds you how to learn11:09
gfidenteI broke it many times ;)11:09
derekhccamacho: translates to "heat stack-create -f sadfsd asdas asfas fa" , its ignored some of the params11:09
trown|outtypewwwhmm launchpad is not linking from patches to tripleo-ci either11:09
*** jaosorior has quit IRC11:10
trown|outtypewwwderekh: https://review.openstack.org/#/c/312212/ fixed stable/mitaka, but there was still an issue with stable/liberty11:10
derekhtrown|outtypewww: I've noticed that recently, don't know why11:10
ccamachoderekh, the issue is the backwards comp between openstack CLI and heat CLI from liberty to mitaka, so depending on the command its parsing and getting some options, I will double-tripleo-check it11:10
*** jaosorior has joined #tripleo11:10
derekhccamacho: ack11:11
trown|outtypewwwthanks ccamacho11:11
* trown|outtypewww is back to afk until after coffee11:11
michchap_gfidente: https://github.com/openstack/puppet-neutron/commit/0229d358ae45add73a0a8f4658632b54fd56be1211:12
*** chmouel has left #tripleo11:12
ccamachogfidente, my first broken CI, I will print a t-shirt with that phrase11:12
gfidenteyes then derek can do the laundry for all11:13
gfidentehe must hate us11:13
openstackgerritMerged openstack/tripleo-heat-templates: Disable VIPs before stopping cluster during version upgrade  https://review.openstack.org/31186011:14
*** ccamacho is now known as ccamacho|include11:15
*** ccamacho|include is now known as ccamacho|getting11:16
*** rodrigods has quit IRC11:16
*** rodrigods has joined #tripleo11:16
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add neutron profiles  https://review.openstack.org/29343611:16
*** ccamacho|getting is now known as ccamacho|coffee11:16
jaosoriorccamacho: No worries, there is a commit to address the issue (Lack of stable branch testing) https://review.openstack.org/#/c/297390/11:18
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: Disable VIPs before stopping cluster during version upgrade  https://review.openstack.org/31245411:21
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: Disable VIPs before stopping cluster during version upgrade  https://review.openstack.org/31245511:21
*** thrash|g0ne is now known as thrash11:22
sshnaidmderekh, trown|outtypewww check out my status page: https://goo.gl/bI9fCi11:23
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Default keystone manage_endpoint to false  https://review.openstack.org/31237311:24
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Create keystone endpoints in keystone profile  https://review.openstack.org/30412511:24
*** myoung|remote has quit IRC11:24
derekhsshnaidm: nice11:27
sshnaidmderekh, I thought maybe to enable elasticsearch on CI, do you have any objections to this? I think it was working before...11:28
*** MaxPC has joined #tripleo11:28
*** jpena|brb is now known as jpena11:29
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: WIP: integrate PCS provider in the merge.  https://review.openstack.org/31071311:29
jaosoriorAny +A for this? https://review.openstack.org/#/c/311833/11:30
derekhsshnaidm: as far as I remember the last time we tried it infra were against it because they didn't want the tripleo results on the elastic search status page, at least without having its own tab or something11:30
jistrjaosorior: done11:30
openstackgerritMerged openstack-infra/tripleo-ci: Add aodh endpoints to enable-tls  https://review.openstack.org/31183311:30
derekhsshnaidm: somebody was going to looking adding it but it never happened, one sec and I'll find the patch11:31
*** masco has quit IRC11:31
jaosoriorjistr: thanks dude11:31
derekhsshnaidm: https://review.openstack.org/#/c/98154/11:32
sshnaidmderekh, weird, why not tripleo. I see there are tracked files, most of them are not relevant, but I can fix it: https://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/files/logstash/jenkins-log-client.yaml11:32
derekhsshnaidm: actually that was elastic recheck11:33
sshnaidmderekh, yeah, exactly11:33
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Pass parameters to manage endpoints via puppet  https://review.openstack.org/24416211:33
derekhsshnaidm: so ya, we have the console logs in there at the moment, feel free to add to the list11:34
sshnaidmderekh, ok, I'll handle it..11:34
derekhsshnaidm: also looks like the tripleo stuff in that file is out of date, may need a refresh11:34
sshnaidmderekh, yes, most of files don't exist anymore, and there is pack of new files instead11:35
derekhsshnaidm: yup, would be great to have it updated11:35
jistrchem: hi, we talked about the PCS integration ^^ briefly with bandini, gfidente and EmilienM. I wonder do you think there would be a way how to ease the transition a bit? If the patch to puppet-pacemaker didn't remove the original pieces so that we could still keep using them during a transition period, or if we started a new pacemaker puppet module instead of redoing the current one, we might be able to make the transition in TripleO in11:36
jistrsteps. The way it's proposed currently to puppet-pacemaker means a huge leap and we were a bit worried some things might break in the transition, so in case that happens and we have to revert for some reason (if a fix can't be found quickly), we'd have to revert both huge patches in both tripleo and puppet-pacemaker. If we are able to keep the original things usable during the transition period, the puppet changes can land independently of11:36
jistrtripleo, and any potential reverts would be done only in tripleo, and they'd be probably much smaller.11:36
sshnaidmderekh, the problem that I can not set there overcloud logs, because it's archived as overcloud-controller-0.xz, or I don't know something..?11:36
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Pass parameters to manage endpoints via puppet  https://review.openstack.org/24416211:36
jistroh and marios too :) ^^11:37
derekhsshnaidm: yup unfortunately it is, if there are a few high value logs we could stash, maybe we could extract those single files at the end of the job11:37
jistrchem: actually i'll post that to the puppet-pacemaker patch review too11:39
chemjistr: the puppet-pacemaker integration with the fuel's one, is definitively not comptatible with the current module interface wise.11:40
jistrchem: yea :(11:40
jistrchem: so maybe starting a new module could be the easiest path11:40
chemjistr: some wrapper could be made to keep the tripleo code *looking* the same, but behind the scene everything would have changed11:40
*** rhallisey has joined #tripleo11:40
hewbroccajistr: glad you raised this... we really have to pay close attention to the migration path11:41
chemjistr: ^ I don't get "starting a new module" ? In what it is different from the current module integration ?11:41
jistrchem: unfortunately that wouldn't solve the problem. It's not just about being able to use the same interface, it's also about using the same internals if need be.11:41
hewbroccaor we will get AODH*10^x11:41
gfidentechem, so we actually talked about also another issue11:41
*** panda has quit IRC11:41
chemjistr: for "same internal", I'm currently porting all the xml provider to pcs to have the same triggered as the old module11:42
*** dixiaoli_ has quit IRC11:42
*** panda has joined #tripleo11:42
*** mbound has quit IRC11:42
gfidentewe need to make sure the resulting configuration is identical even in naming the resources otherwise they get duplicated11:42
jistrchem: it's very different because in case we find out the new module doesn't do something right, we'll be able to revert to the old behavior *only in tripleo*, without reverting the migration patch in puppet-pacemaker too11:42
gfidenteor at leat that was one thing bandini was concerned about11:43
*** pkovar has joined #tripleo11:43
gfidenteone more reason for a separate module is that we can do maintenance on the old one if needed11:43
chemgfidente: the commands are the same pcs command as the old module, and naming is taken care of by pcs, so that should be covered11:44
chemgfidente: anyway I'll be diffing the cib on install as a test11:45
gfidentejistr, chem , in fact I think if it was a different module and we were diffing the cib, it wouldn't really need to have the same interface11:46
chemjistr: so you want to create a new pacemaker module ? I really don't get this :)11:46
chemjistr: we could just import the currently reworked module in tripleo, maybe naming it transpacemake and so you would have the tripleo only patch, no ?11:48
jistrchem: yes. I realize that you're being super careful here, but let's say hypothetically that we discover that the new behavior is broken in some way. Maybe not resulting in wrong config, but perhaps breaks upgrades for some reason, or causes trouble in combination with composable roles for some reason, etc. What do we do then? Revert the 18 000 lines in puppet-pacemaker?11:49
jistrchem: the "yes" was for "you want to start a new pacemaker module"11:49
chemjistr: out of the 18000 most of them are for a library that does what pcs does written in ruby and tests (lot of them).  Most of the "real" change relative to the previous module are there https://review.openstack.org/#/c/310713/11:51
chemjistr: this is the "pcs" only code transition11:51
chemjistr: I can do quite easily a "pcs" only module that could be added to tripleo11:52
chemjistr: removing all the xml code11:53
chemjistr: that wouldn't be that much extra work.  Then the patch would be much more lightweigh and only in tripleo.  What do you think ?11:53
*** weshay has joined #tripleo11:54
*** yamahata has quit IRC11:54
*** lucas-hungry is now known as lucasagomes11:55
chemjistr: when this transition module get clearance for update, install, so on, then we can move to the "merged" pacemaker module as the interface and the code would be the same.11:55
jistrchem: hmm i'm not sure i understand how any of this guarantees our ability to keep the original behavior though (unmodified, 1:1)? You suggest to move the old puppet-pacemaker code to puppet-tripleo, migrate tripleo-heat-templates to it, then merge the puppet-pacemaker overhaul, then migrate tripleo-heat-templates to it? To me that sounds like much more work than just starting a new puppet module.11:55
chemjistr: hum, I've just made a new proposition ^11:56
jistrlol11:56
chemjistr: It's the same as doing a new module no ?11:58
*** links has joined #tripleo11:59
jistrchem: not *exactly* the same, there's still possibility for breakage, but it probably gets pretty close to doing a new module, yea11:59
chemjistr: but if you do a *new* how do you garanty unmodified 1:1  ?12:00
chem*new module*12:00
jistrchem: i meant new module for the new stuff, and keep puppet-pacemaker the way it always was. The only problem with this is that the new module would have to have a creative name, e.g. puppet-pcmk :)12:00
jistrbut that way it would be guaranteed unmodified :)12:01
chemjistr: hehe.  Okiii, I though you wanted to *write* a new module12:02
*** noslzzp has joined #tripleo12:02
*** noslzzp has quit IRC12:02
*** ccamacho|coffee is now known as ccamacho12:03
*** noslzzp has joined #tripleo12:03
jistrchem: so in this option we'd have Dmitry's and your puppet work unmodified, but we'd put it into a new module, not puppet-pacemaker12:03
chemjistr: oki, I've got it12:03
*** morazi has joined #tripleo12:05
adarazsdmsimard: hi, are you around?12:05
sshnaidmdoes somebody know where is admin user/pass to mariadb on controller in tripleo?12:05
chemjistr: so where do we maintain this, in github, and the puppet-pcmk seems oki to me as a name.12:06
*** zoli|lunch is now known as zoli12:06
chemin github?12:06
*** zoli is now known as zoliXXL12:06
jistrchem: i think it would be under openstack/12:06
jistropenstack/puppet-pacemaker == old stuff we migrated under openstack12:06
jistropenstack/puppet-pcmk == new stuff (Dmitry's + yours)12:06
openstackgerritCarlos Camacho proposed openstack/puppet-tripleo: WIP: Puppet manifest to configure NTP as a composable service  https://review.openstack.org/31072512:07
jistrchem: we'd probably just need an ack from EmilienM on that12:07
openstackgerritHonza Pokorny proposed openstack/tripleo-ui: Disable links when deploy is in progress  https://review.openstack.org/31247312:07
EmilienMhello12:07
jistro/12:07
gfidenteEmilienM, hey12:09
adarazssshnaidm: most of the passwords are in /etc/puppet/hieradata/controller.yaml.gz -- I don't see the admin password, but there are passwords for all the services.12:09
jistrEmilienM: we're discussing how to ease the transition to the new pacemaker Puppet code (you may recall we discussed it briefly at the summit) if you need full context just scroll back 30 minutes please :)12:09
*** abehl has quit IRC12:09
EmilienMok for new module as long: 1) mirantis agrees to collaborate on this new module 2) we document what happens somewhere (mailing-list, etc) 3) the new module lives under openstack/12:10
jistrEmilienM, chem: ++12:10
chemjistr: EmilienM ++12:11
sshnaidmadarazs, thanks, do you know the key name for hiera for db admin?12:11
EmilienMif you need help for 3), let me know. Infra + TC will wonder what happens here, so we can explain easily with 2)12:12
adarazssshnaidm: nope, I didn't see it there12:12
*** davidlenwell has quit IRC12:13
chemjistr: EmilienM I'll write a wrap up of the discussion (2), and I'll get in touch with Dmitry later today12:13
EmilienMexcellent12:13
jistrthanks a lot chem12:14
jistr++12:14
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add neutron profiles  https://review.openstack.org/29343612:15
*** davidlenwell has joined #tripleo12:15
*** ChanServ sets mode: +v davidlenwell12:15
*** jayg|g0n3 is now known as jayg12:15
*** mbound has joined #tripleo12:16
*** dprince has joined #tripleo12:17
*** olap has quit IRC12:17
openstackgerritHonza Pokorny proposed openstack/tripleo-ui: Add hints to run validations  https://review.openstack.org/31247612:23
bandinithanks chem! ;)12:29
*** fultonj has joined #tripleo12:29
*** pradk has quit IRC12:34
*** jpena is now known as jpena|lunch12:34
*** myoung has joined #tripleo12:36
*** myoung has quit IRC12:37
*** trozet has quit IRC12:46
*** trown|outtypewww is now known as trown12:46
openstackgerritCarlos Camacho proposed openstack-infra/tripleo-ci: Fix validation for stable branch  https://review.openstack.org/31235012:46
*** links has quit IRC12:46
*** julim has joined #tripleo12:47
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Deploy RabbitMQ as a composable role  https://review.openstack.org/31186312:54
*** akshai has joined #tripleo12:55
*** tzumainn has joined #tripleo12:57
*** Goneri has joined #tripleo13:00
*** trozet has joined #tripleo13:03
*** gfidente has quit IRC13:05
*** gfidente has joined #tripleo13:06
*** gfidente has quit IRC13:06
*** gfidente has joined #tripleo13:06
dprinceEmilienM: In general https://review.openstack.org/#/c/311863/6 looks good to me. I have one question about maintaining the RabbitCookie resource within the nested stack template however.13:06
dprinceshardy: could you look at my comment there? ^^^13:06
dprincegfidente: you too? ^^^13:07
*** krotscheck_dcm is now known as krotscheck13:10
*** rbrady has joined #tripleo13:13
openstackgerritFlorian Fuchs proposed openstack/tripleo-ui: [WIP]Adds thunk component, actions, records, reducer for a service status page.  https://review.openstack.org/31249413:14
akrivokadtantsur: is there any work in progress for fixing instackenv.json to more closely match Ironic?13:15
*** ifarkas has quit IRC13:15
dtantsurakrivoka, not that I'm aware of. I would love it to happen though13:15
akrivokadtantsur:  where would I file a RFE for that?13:17
*** zoliXXL is now known as zoli|afk13:17
dtantsurakrivoka, tripleo, dunno how they track it. maybe a bug is fine13:18
*** coolsvap has quit IRC13:18
akrivokadtantsur: ack, thanks13:19
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Create dbs in step 3 for the roles  https://review.openstack.org/31249613:20
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Create dbs in step 3 for the roles  https://review.openstack.org/31249613:20
shardyakrivoka, dtantsur: A bug in https://bugs.launchpad.net/tripleo will be fine, thanks13:21
akrivokashardy: https://bugs.launchpad.net/tripleo/+bug/157820313:21
openstackLaunchpad bug 1578203 in tripleo "Fix instackenv.json to more closely match Ironic" [Undecided,New]13:21
shardyakrivoka: thanks - there has been some discussion about renaming that file also, so we could possibly deprecate instackenv.json, and move to $something_better.yaml with content that better maps to Ironic13:22
gfidenteEmilienM, https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/manifests/overcloud_controller_pacemaker.pp#L35113:22
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Remove calls to ::mysql from the manifests  https://review.openstack.org/31087913:24
*** olap has joined #tripleo13:24
gfidenteEmilienM, and this https://review.openstack.org/#/c/31093613:25
dtantsurshardy, in this case we can probably move it to python-ironicclient as well13:26
shardydtantsur: that sounds good, although there has been some discussion around putting more data in there, which could be consumed a deploy time (such as networks/IP's assigned to a node)13:28
*** electrofelix has quit IRC13:28
shardywhat would be ideal is if there was some way to define e.g the "extra" data blob in ironic via that file13:28
shardyas well as the capabilities13:28
*** trozet_ has joined #tripleo13:28
shardyand have a way to inject that into each node (e.g via a config drive) on boot13:28
shardyI know there was a patch to nova which enabled that some time ago, and reviving such an approach could make certain per-node things much easier13:29
*** ifarkas has joined #tripleo13:29
*** trozet has quit IRC13:30
shardydprince: You're right, moving RabbitCookie will definitely change it's value13:30
*** trozet_ is now known as trozet13:30
shardyso we have to ensure this will work on upgrade - I assumed such random salts should not be changed on upgrade13:30
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Leave start/stop/restart for Keystone and Glance in charge to the role  https://review.openstack.org/31093613:30
dprinceshardy: yeah, unclear to me if this would functionally break a rabbit cluster or if swapping in the new value would just work13:31
*** electrofelix has joined #tripleo13:31
shardyI guess in this case we do have RabbitCookieSalt which seems to indicate you'd sometimes want to force it to change13:31
dprinceshardy: yep13:32
hewbroccaRabbitCookieSalt sounds like a tasty treat13:32
shardylol :)13:33
dtantsurshardy, nova hates being treated as a pass-thru mechanism for ironic :)13:33
dtantsurhewbrocca++13:33
dprinceits delicious13:33
*** zoli|afk is now known as zoli13:33
*** zoli is now known as zoliXXL13:33
shardydtantsur: Yeah, but we're having to do some pretty extreme stuff like implement our own scheduler filter, just so we can pin node placement such that data can be passed via heat13:33
dprincedtantsur: and we (or some of us) are starting to hate Nova for that ;)13:34
*** egafford has joined #tripleo13:34
dtantsurdprince, ;)13:34
shardythings like IP's, hostnames, networks, and mac address to os-net-config aliases are really attributes of the physical nodes, so it'd be far easier if we could group that data with the nodes themselves :)13:34
dprinceunless you are multi-tenant the use case for Nova as a front end to Ironic isn't great13:34
dprincestill, there are some things...13:34
shardyall we need is a way to make that data available to each node at runtime13:35
shardywe could even push all the data to swift as a mapping, but we still need a way to inject a unique key (e.g the ironic node ID) on deployment13:35
shardyunless we do something horrible like lookup node data via mac address or something13:36
dprinceshardy: if we use Ironic directly we could use config drive (which Nova prevents us from tapping into at this point)13:36
shardydprince: we actually don't need nova for this, we just need a way to tell Ironic to always attach a config drive with the data from the "extra" blob13:37
*** dkehn__ has joined #tripleo13:37
dprinceshardy: Heat resources for Ironic? :)13:37
*** lblanchard has joined #tripleo13:37
shardydprince: Heh, there has been further discussion on that topic, could be worth prototyping13:39
hewbroccaIt would sure as hell solve our scheduling problem13:39
hewbroccathe only issue is, we can't really use it for metal-to-tenant I don't guess13:39
*** dkehn_ has quit IRC13:39
dprinceshardy: I'm curious, do you have a link to your last attempt (that was blocked)13:40
* dprince wants to pick a fight13:40
dprinceshardy: seriously though. I don't see how anyone would block this feature in heat13:41
trownsounds like the vehement blocker the first go around has seen the light13:41
shardydprince: https://review.openstack.org/#/c/145302/ was the patch we previously used, related to https://review.openstack.org/#/c/173379/13:42
shardytrown: Yeah, apparently devananda doesn't hate the idea quite so much anymore13:42
shardyI read over the discussion we had on the reviews and it's pretty funny if you can get past the frustrating waste of effort13:43
hewbroccasunk cost13:43
dprinceshardy: who cares if he hates it? I mean how does a project get to control who uses their API anyway?13:43
hewbrocca:D13:43
dprinceIts an API. Use it however you'd like13:44
shardydprince: at the time he was the PTL, and, in general, it's just easier if the PTL of $project doesn't show up 30seconds after you post every patch with $project in the title saying how much he absolutely hates your idea ;)13:45
dprinceshardy: sounds like motivation to me :)13:45
shardyyou're right though, we can revisit it, but as hewbrocca mentions, we'll be papering over some gaps that will eventually bite metal-to-tenant users13:46
*** d0ugal has quit IRC13:46
*** d0ugal has joined #tripleo13:47
dprincegood quality paper I hope13:47
dprinceif we put an owl sticker on it it'll be fine13:47
trownwho needs patches when you have owl stickers13:48
*** jprovazn has quit IRC13:49
trownany of those getting sent to RDU btw?13:49
*** electrofelix has quit IRC13:52
*** electrofelix has joined #tripleo13:52
*** mbound has quit IRC13:54
dtantsurtrown, come to BRQ, we have a lot ;)13:55
*** jpena|lunch is now known as jpena13:55
*** mbound has joined #tripleo13:55
*** electrofelix has quit IRC13:57
trowndtantsur: I would love to come to BRQ... might be a year or more though13:57
*** electrofelix has joined #tripleo13:57
*** trozet has quit IRC13:57
*** trozet has joined #tripleo13:58
*** jtomasek has quit IRC14:02
*** jtomasek has joined #tripleo14:04
*** dtrainor has joined #tripleo14:05
*** mbound has quit IRC14:08
jristtrown: I'll probably do another order and ship it to RDU :)14:09
jristalong with some shirts14:09
jristright hewbrocca ^14:10
trownjrist: nice, I would like to give some to support folks in RDU too14:10
jristcool14:10
adarazsderekh: hi! I have a DLRN question... I'm trying to build THT, and I get this error:14:11
adarazsfatal: unable to access 'http://review.rdoproject.org/r/p/openstack/tripleo-heat-templates-distgit.git/': Encountered end of file14:11
adarazsI don't see that this is a valid repo... how does it work upstream? what did I forget to set?14:11
*** dmk0202 has joined #tripleo14:14
derekhadarazs: this is what we do in tripleo http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/scripts/tripleo.sh#n28614:14
*** nico_auv has joined #tripleo14:15
derekhadarazs: the process your using must have been effected by the rdo move to 'http://review.rdoproject.org'14:15
*** liverpooler has quit IRC14:15
adarazsderekh: yeah, I was actually trying to do the same things as the script you linked.14:15
*** absubram has joined #tripleo14:16
adarazsderekh: so why do I have this problem with the review.rdoproject.org? I clone the rdonifo from https://github.com/redhat-openstack/rdoinfo.git -- shouldn't that have some working URLs for the distgit dirs?14:18
*** mgould has quit IRC14:19
*** yamahata has joined #tripleo14:20
openstackgerritHonza Pokorny proposed openstack/tripleo-ui: Add hints to run validations  https://review.openstack.org/31247614:20
sshnaidmtrown, do you know maybe what this command does? "sudo crm_resource -r openstack-heat-api --wait " - after overcloud install14:21
*** tosky has quit IRC14:21
*** ebalduf has joined #tripleo14:21
*** bnemec has joined #tripleo14:21
*** mgould has joined #tripleo14:23
derekhadarazs: I'm able to git clone that exact url you posted14:23
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Pass parameters to manage endpoints via puppet  https://review.openstack.org/24416214:24
*** tosky has joined #tripleo14:24
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Create keystone endpoints in keystone profile  https://review.openstack.org/30412514:24
*** jaosorior has quit IRC14:25
*** ppradhan has joined #tripleo14:25
*** jaosorior has joined #tripleo14:25
bandinisshnaidm: it waits until openstack-heat-api resource has no transitions in the graph (aka it is stable - be it stopped or started)14:26
*** akshai has quit IRC14:26
sshnaidmbandini, thanks, so if it's not - what is consequences..?14:27
adarazsderekh: drat... this could be some transparent proxy issue again in the BRQ office o_o14:29
*** mbound has joined #tripleo14:29
bandinisshnaidm: well it means that the cluster is constantly transitioning. what's the issue?14:29
derekhadarazs: ack14:29
sshnaidmbandini, trying to understand what happened here exactly: http://logs.openstack.org/79/310879/2/check-tripleo/gate-tripleo-ci-f22-upgrades/69d84f5/console.html14:29
*** jaosorior has quit IRC14:30
*** akshai has joined #tripleo14:30
sshnaidmbandini, so cluster hasn't finished it work within 3 minutes after overcloud is deployed, are there other logs I can see for details maybe?14:31
*** pradk has joined #tripleo14:31
*** julim has quit IRC14:32
bandinisshnaidm: usually a crm_report will so (or simply collect an sosreport), but I assume we cannot connect to that vm, correct?14:33
*** dmk0202 has quit IRC14:33
sshnaidmbandini, yes, all log are there though: http://logs.openstack.org/79/310879/2/check-tripleo/gate-tripleo-ci-f22-upgrades/69d84f5/logs/  at lease everything that was in /var/log14:34
*** mbound has quit IRC14:34
*** julim has joined #tripleo14:34
*** jprovazn has joined #tripleo14:35
*** dprince has quit IRC14:39
*** ramishra has quit IRC14:47
*** ramishra has joined #tripleo14:48
*** sshnaidm has quit IRC14:50
*** yamahata has quit IRC14:52
*** yamahata has joined #tripleo14:52
*** akshai has quit IRC14:53
openstackgerritFlorian Fuchs proposed openstack/tripleo-ui: [WIP] Status Page UI - work in progress  https://review.openstack.org/31067014:53
openstackgerritFlorian Fuchs proposed openstack/tripleo-ui: [WIP]Adds thunk component, actions, records, reducer for a service status page.  https://review.openstack.org/31249414:53
*** mbound has joined #tripleo14:53
*** akshai has joined #tripleo14:54
*** dsariel has quit IRC14:54
*** tremble has quit IRC14:55
*** cwolferh has joined #tripleo14:57
*** bvandenh has joined #tripleo15:00
*** sshnaidm has joined #tripleo15:05
chemjistr: actually, there may be a way to keep the legacy code of the current puppet pacemaker module inside the merged one: same interface and same code.  I would like to give it a try tomorrow, so I will put the mail on hold till then.  Would this be ok for you?15:08
jistrchem: yea i think that would be fine /cc bandini gfidente15:10
*** zoliXXL is now known as zoli|relocating15:11
*** zoli|relocating is now known as zoli_gone-proxy15:12
*** r-mibu has quit IRC15:13
*** r-mibu has joined #tripleo15:13
trownsshnaidm: nice one on that status page... I really like the reason field15:15
trownsshnaidm: what is the Build column for? seems always empty15:16
*** absubram has quit IRC15:17
bandinisshnaidm: will look at the logs later tonight15:18
*** absubram has joined #tripleo15:18
*** absubram has quit IRC15:19
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Use megabytes for overcloud image size  https://review.openstack.org/31259415:22
*** aufi has quit IRC15:27
*** sshnaidm_ has joined #tripleo15:30
*** devvesa has quit IRC15:31
sshnaidm_trown, the build should have point to Jenkins build page, but it's blocked now15:32
*** sshnaidm has quit IRC15:32
trownoh right15:32
*** absubram has joined #tripleo15:34
*** jistr has quit IRC15:37
sshnaidm_trown, suggestions are welcome :)15:37
*** fragatina has quit IRC15:37
*** panda has quit IRC15:41
*** panda has joined #tripleo15:42
*** mikelk has quit IRC15:43
ppradhanMetadata quesiton, the overclod deply failed on both controller and computer nodes. on the nodes I see No Local Metadata found. How do I verify this?15:43
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Deploy RabbitMQ as a composable role  https://review.openstack.org/31186315:46
*** Guest23319 has quit IRC15:46
*** ccamacho is now known as ccamacho|away15:47
*** sshnaidm_ has quit IRC15:48
gfidentejdob, anything good?15:57
jdobgfidente: i haven't actually even started yet, finishing up a patch and then it's next on my list15:57
*** jistr has joined #tripleo15:58
gfidenteoh okay so I tried to put in LP what seemed useful to me15:58
gfidentebut couldn't in fact reproduce it locally15:58
gfidenteI mean, the same code change works for me locally15:58
gfidenteso it seems something to do with slowliness ... or extreme parallelization due to slowliness15:59
jdobya, this is gonna be annoying :)15:59
openstackgerritBen Nemec proposed openstack/instack-undercloud: Add dib-lint to pep8 tox job  https://review.openstack.org/30652116:00
*** lucasagomes is now known as lucas-afk16:01
*** saneax is now known as saneax_AFK16:01
gfidente:(16:02
*** bvandenh has quit IRC16:04
*** athomas has quit IRC16:06
openstackgerritBen Nemec proposed openstack/tripleo-heat-templates: Sample environment generator  https://review.openstack.org/25363816:07
*** veteran has joined #tripleo16:09
*** dmacpher has joined #tripleo16:09
*** Paul_ has joined #tripleo16:11
ansiwenpradk: btw, was it planned to have the integration meeting always that late?16:11
pradkansiwen, depends on what timezone you're in :)16:11
*** Paul_ has quit IRC16:11
*** zoli_gone-proxy is now known as zoliXXL16:11
*** zoliXXL is now known as zoli|mtg16:11
pradkansiwen, we can change it if its too late for you16:12
ansiwenpradk: today I realized it collides with my yoga classes, but the time window you proposed was 14-16UTC, so since it's summertime were in... :-)16:12
*** pkunicki has joined #tripleo16:12
*** fragatina has joined #tripleo16:13
pradkansiwen, ah ok, so you prefer earlier ? i can revisit the times and see if fits everyone16:13
*** athomas has joined #tripleo16:14
ansiwenpradk: well, I don't want to cause a trouble, I might switch to another class. but if it's ok for everybody I would prefer (during DST) something before 1500UTC16:14
pradkansiwen, understood, i'll look into it16:15
*** dprince has joined #tripleo16:16
ansiwenpradk: great, thanks16:16
*** dtantsur is now known as dtantsur|afk16:17
*** oshvartz has quit IRC16:19
*** mbound has quit IRC16:23
*** hewbrocca is now known as hewbrocca-afk16:28
jristderekh: had a spark of inspiration https://i.imgur.com/zobe3ko.png16:37
jristanyone else that does DLRN ^16:38
derekhjrist: nice16:38
pino|workuniversal pictures attorneys coming in 3.. 2.. 1..16:38
jristhaha16:39
jristnah16:39
jrist:)16:39
jristpino|work: http://everythingisaremix.info/16:39
dmsimardjrist: oh god, that's nice16:39
jristdmsimard: I like it :)16:39
jristwould make a good sticker...16:39
*** sshnaidm_ has joined #tripleo16:44
*** veteran has quit IRC16:44
ansiwenjrist: what did you use to illustrate it? inkscape?16:46
jristansiwen: admittedly illustrator, I'm still learning inkscape16:46
ansiwenwell, that was my first guess of course, but I thought it would be uncool to assume that :-)16:47
jristhaha16:47
jristI was an illustrator guy before inkscape existed16:47
ansiwenoh, you were?16:47
jristaye. I'm old!16:48
ansiwennot as old as me16:48
jristpossible :)16:48
jristI'm 3516:48
jristsoftware-old16:48
ansiwen4216:48
ansiwenjrist: you look a lot younger... like me and marios :-)16:49
dprincejrist: nice DLRN16:49
*** florianf has quit IRC16:49
ansiwenwe all have baby-face, as marios described it16:49
jristdprince: thanks :)16:49
dprinceslagle: hey, so what about calling it just 'Planet' instead of 'Blog Posts'16:49
ansiwenis dlrn the official name of delorean?16:50
trownansiwen: it is the rename of delorean, because delorean already exists on pypi16:52
ansiwentrown: I see, tha nks16:53
*** jistr has quit IRC16:54
*** trown is now known as trown|lunch16:57
*** derekh has quit IRC16:57
*** akshai has quit IRC17:00
*** akshai has joined #tripleo17:02
*** mcornea has quit IRC17:06
*** Marquis42 has quit IRC17:07
*** tosky_ has joined #tripleo17:08
*** tosky has quit IRC17:09
*** ibravo has quit IRC17:09
*** ohamada has quit IRC17:15
slagledprince: Planet TripleO?17:16
slaglePlanet Owl?17:16
dprinceslagle: either I guess :). I left a comment on the review17:17
openstackgerritBrent Eagles proposed openstack-infra/tripleo-ci: Fail pingtest more quickly  https://review.openstack.org/30848617:18
*** nico_auv has quit IRC17:19
*** paramite is now known as paramite|afk17:22
*** coolsvap has joined #tripleo17:25
*** jpena is now known as jpena|off17:28
*** shivrao has joined #tripleo17:30
*** shivrao_ has joined #tripleo17:30
*** akshai has quit IRC17:34
*** shivrao has quit IRC17:35
*** shivrao_ is now known as shivrao17:35
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: WIP: integrate PCS provider in the merge.  https://review.openstack.org/31071317:36
*** electrofelix has quit IRC17:36
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: WIP: integrate PCS provider in the merge.  https://review.openstack.org/31071317:42
*** akshai has joined #tripleo17:44
*** bvandenh has joined #tripleo17:47
*** shivrao has quit IRC17:48
*** shivrao has joined #tripleo17:50
*** pkovar has quit IRC17:52
*** rbrady has quit IRC17:53
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: WIP: integrate PCS provider in the merge.  https://review.openstack.org/31071317:56
*** tosky_ has quit IRC17:57
*** mgould has quit IRC18:02
*** zoli|mtg is now known as zoli|gne18:03
*** zoli|gne is now known as zoli|gone18:03
*** sambetts is now known as sambetts|afk18:03
*** shivrao has quit IRC18:03
*** zoli|gone is now known as zoli_gone-proxy18:03
*** ibravo has joined #tripleo18:03
*** shivrao has joined #tripleo18:06
*** shivrao_ has joined #tripleo18:10
*** shivrao has quit IRC18:13
*** shivrao_ is now known as shivrao18:13
*** paramite|afk is now known as paramite18:14
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Deploy RabbitMQ as a composable role  https://review.openstack.org/31186318:19
*** Idrot has joined #tripleo18:25
*** gfidente has quit IRC18:26
*** akshai has quit IRC18:28
*** trown|lunch is now known as trown18:31
*** akshai has joined #tripleo18:31
*** shardy has quit IRC18:34
*** liverpooler has joined #tripleo18:38
*** rook has joined #tripleo18:48
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Deploy RabbitMQ as a composable role  https://review.openstack.org/31186318:51
rookping TripleO peeps... if a update fails (due to timeout) is there a way to determine why? This is after scaling to 280 compute nodes.18:54
rookathomas ^ not sure if you might have some thoughts18:55
fultonjrook: did you try http://hardysteven.blogspot.com/2015/04/debugging-tripleo-heat-templates.html ?18:56
*** abehl has joined #tripleo18:57
* cmyster wonders if shardy gets payed every time his blog is being mentioned...19:02
* rook looks for ads19:04
rook:P19:04
rookfultonj yeah, i have this pretty well scripted.19:04
slaglewhat resource did it time out on? or was it just that there are too many resources and the stack timed out after 4 hours?19:04
rookslage one of the existing compute nodes that already deployed19:05
slaglewhat resource in the nested stack for that compute node?19:05
slagleor can you ssh to that node and check the os-collect-config log to see where it might be stuck19:06
rookhttps://gist.github.com/jtaleric/7ed6739362728d0a9af1fc2eef28e8b119:06
rookslagle my assumption is that we are bumping into the timeout, which I hard set to 400M19:06
rookminutes*19:06
slagleyea that should be a relatively quick deployment, all that does is write some hieradata yaml files19:08
rookslagle we are seeing once we get around 250 nodes, we are pushing against the default 240M19:09
rookonce we started the 280 node deployment, we hit it... I have bumped things to 400minutes but it is still failing19:10
rookslage However, It isn't clear to me what is failing19:10
slaglecan you heat event-list 1105d375-0b68-431e-afba-fd8f0ddb772e19:10
trownlarsks: have you seen https://bugs.launchpad.net/tripleo-quickstart/+bug/157780919:10
openstackLaunchpad bug 1577809 in tripleo-quickstart "virsh error: error: no valid connection error: Cannot create user runtime directory '/run/user/1000/libvirt': Permission denied" [High,Triaged]19:10
openstackgerritDan Prince proposed openstack/tripleo-common: Initial Integration of Mistral into TripleO  https://review.openstack.org/28236619:14
openstackgerritDan Prince proposed openstack/tripleo-common: Mistral action to drive Heat software deployments  https://review.openstack.org/31096619:14
dprincestevebaker: yo, that one works ;) ^^^19:14
rookslage interesting...19:14
rookhttps://gist.github.com/jtaleric/39324f579374f07563b11b2766f6a11d19:16
rookslagle: ^19:16
slaglerook: hmm, did you get OOM killed perhaps?19:16
rookhm, we have swap and tons of memory19:17
rooklooking19:17
rookwtf19:17
rookslage, naw, OOM isn't trolling us19:18
rookwonder if I need to bump the rpc timeouts?19:19
rookit might think heat engine went away because things timed out?19:19
openstackgerritJames Slagle proposed openstack/instack: Update $IMAGE_ELEMENT after excluding elements  https://review.openstack.org/31270919:20
slaglerook: possibly, or maybe those events are from a previous attempt19:21
rookpossibly19:21
slagleit's hard to tell from the timestamps in the gists19:21
openstackgerritJames Slagle proposed openstack/instack: Update $IMAGE_ELEMENT after excluding elements  https://review.openstack.org/31270919:23
rookslage they are all the same attempt i bet19:24
rookwith me setting the timeout to 400 minutes it is going to span a couple of hours.19:24
rookheat RPC is set to 600 (default it seems)19:26
slaglerook: it looks like 70810c8d-ac87-44d8-b6cc-998717c43bde is the resource that may have actually timed out19:27
slaglecan you resource-list 1105d375-0b68-431e-afba-fd8f0ddb772e and see which one 70810c8d-ac87-44d8-b6cc-998717c43bde is?19:28
slagleor is 70810c8d-ac87-44d8-b6cc-998717c43bde the event uuid19:28
slaglemaybe it is19:28
*** mbound has joined #tripleo19:29
slagleyea that's the event uuid, nm19:29
slaglerook: can you run the event-list again without the grep? i want to see the time between the creation and the time out19:30
larskstrown: looking...19:31
larskstrown: yeah, I was talking to weshay about that yesterday (or so?); I didn't have the chance to investigate yet.19:32
rookhttps://gist.github.com/jtaleric/165c20cca2dda76fe43da5381c7867c0 slage19:32
trownlarsks: k, it is an odd one19:33
larsksYeah, totally.19:33
larsksGoing to try to poke at that this afternoon.19:33
*** paramite has quit IRC19:35
trownI had some weird results, where the behavior seemed different depending on the order I shut things down in19:36
ayoungThe files ('overcloud-without-mergepy.yaml', 'overcloud.yaml') not found in the /usr/share/openstack-tripleo-heat-templates/ directory19:36
trownman... I thought we fixed that error19:36
ayoungtrown, that was with tip of tree, checked out last night or this morning19:36
trownayoung: the error is not very helpful, it wants to say "some file you referenced can't be found"19:37
slaglerook: ok, so it started at 2016-05-04T18:07:32 and timed out at 2016-05-05T00:46:2819:37
trowninstead of specifically pointing only to the top level template19:37
ayoungtrown, right.  I just assumed that something was either broken in tripleo-heat-templates or the quickstart was out of date19:37
slaglerook: that makes me think the whole stack got hung up on this one resource19:37
rookhm19:38
rooki can ssh to that guest just fine19:38
rookguest/instance19:38
slaglerook: can you ssh to this node and look at os-collect-config output?19:38
ayoungtrown, knwon issue?  Workaround?19:38
slaglerook: you're looking to see if os-collect-config triggered a run around 2016-05-04T18:07:3219:38
trownayoung: hmm, the error being not so helpful is a python-tripleoclient issue, that I know at least has a patch19:38
trownayoung: hitting that error with tripleo-quickstart is not a known issue... what is your command?19:39
trownayoung: or more importantly what config are you passing in19:39
ayoungnothing custom19:39
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-quickstart: lots and lots of comments (1/2)  https://review.openstack.org/31098019:40
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-quickstart: lots and lots of comments (2/2)  https://review.openstack.org/31098119:40
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-quickstart: use yaml blocks for really long strings  https://review.openstack.org/31097919:40
ayoungtrown, from my laptop: ./tripleo-quickstart/quickstart.sh ayoung-dell-t170019:40
ayoungthen19:40
ayoung ssh -F /home/ayoung/.quickstart/ssh.config.ansible undercloud19:40
ayoung ./undercloud-install.sh19:40
ayoungwait a while19:40
ayoung./undercloud-post-install.sh19:40
ayoungwait some more19:41
ayoung./overcloud-deploy.sh19:41
ayoungand then come here to bug you19:41
trownwtf19:41
ayoungtrown, I saw this last night, but it was an old install, so I wanted to start from scratch.19:42
*** panda has quit IRC19:42
trownessentially `quickstart.sh -t all $VIRTHOST` which is what runs in CI19:42
ayoungtrown, yeah, just step by step19:42
*** panda has joined #tripleo19:42
ayoungtrown, want me to debug?  I can go through the Hardy-Approved-Process19:43
rookslagle: nothing notable in os-collect-config19:43
ayoungtrown, but there is no stack19:43
ayoungheat stack-list --show-nested -f "status=FAILED"19:43
ayoungBupkis19:43
trownayoung: well, it is failing before ever starting... could you paste overcloud-deploy.sh somewhere?19:44
ayoungyep19:44
pandaayoung: it's missing network-environment.yaml19:44
ayounghttp://paste.openstack.org/show/496143/19:44
slaglerook: ok, if it didn't trigger a run of os-refresh-config around that time, then you may have tickled a heat bug where it never made the new metadata available, so the node never ran os-collect-config which would signal back to heat. so heat timed it out19:44
rookoh slagle it did trigger19:45
rookslage a couple of times19:45
rookslage i just see the host names over, and over again19:45
rookslage, sorry i didn't see any errors19:45
ayoungpanda, any idea why? Is this an RPM build issue, where a new file needs to be added?19:45
rookslage however, it seems to just be inserting the host names in a loop19:45
slaglerook: can you send me or paste the log?19:46
*** mkovacik has quit IRC19:46
pandaayoung: I did the same thing I think yesterday, quickstart is not driving everything, if you don't specify the right tags19:46
trownayoung: hmm, ya is /home/stack/network-environment.yaml on the underclou?19:46
rookslage err how about  a snip of it19:46
ayoungtrown, no it is not19:46
slaglerook: sure we can start there :)19:46
ayoungpanda, where do I get one?19:47
trownpanda: ayoung, ok that seems like a valid quickstart bug... running scripts one by one should work the same as using --tags all19:47
rookhttps://gist.github.com/jtaleric/6e007cafd25172bd0d65971c7a963591 slage, this is what i see over and over in the log around this time19:47
ayoungtrown, want me to go back and try running --tags all?19:47
pandatrown: oh ? with --tags all it works19:47
ayoungto confirm?19:47
rookslage ah, one sec19:47
pandathe scripts do not create network-envinroment, only the task in ansible do19:47
*** openstackgerrit has quit IRC19:48
rookslage gist updated19:48
*** openstackgerrit has joined #tripleo19:48
rookso this is the last bit before we don't see anything else from the metadata19:48
trownpanda: oh right... yuck19:48
pandatrown: ayoung I tought this was intentional, to limit the amount of things quickstgart does to your environment if you don't tell it to do it.19:48
ayoungpanda, shouldn't that be done by the ./undercloud-postinstall.sh script, though?19:49
trownpanda: well ideally it provides a "choose your own adventure" and all of the adventures end in some level of success19:49
ayoungrunning with -t all19:49
pandatrown: yes, and if you choose the default adventure, you have to provide network-environment on you own19:50
pandathat's you adventure :)19:50
trownya, but that page is ripped out of the book19:51
trown(go to page 42... there is not effing page 42, eff this book)19:51
pandalet the people without imagination use the best tools ...19:51
panda:)19:52
trownayoung: mind filing a bug for that here https://bugs.launchpad.net/tripleo-quickstart19:52
*** athomas has quit IRC19:52
*** akshai has quit IRC19:55
*** abehl has quit IRC19:56
pandatrown: it's not a bug, it's an adventure.19:56
trownI am going to ask people to file an adventure19:57
*** yamahata has quit IRC19:58
*** dprince has quit IRC19:58
ayoungpanda, why can't it be both?19:59
ayoungtrown, https://bugs.launchpad.net/tripleo-quickstart/+bug/157836820:01
openstackLaunchpad bug 1578368 in tripleo-quickstart "Performing step by step fails overcloud deploy due to missing network config" [Undecided,New]20:01
trownayoung: thanks for the adventure20:02
ayoungHappy Paul Desmond Day20:02
pandaayoung: a bug's life. Sounds nice.20:02
ayounghttps://www.youtube.com/watch?v=vmDDOFXSgAs20:03
ayoungHint (date, time signature)20:03
trownah... it took me actually listening20:04
trown5/4 is pretty neat20:04
pandauhm, is it because take 5 - DB 4tet ? or because of the rhythm in 5/4 ?20:07
trowntime signature is 5/420:09
*** akshai has joined #tripleo20:10
jristslagle: Blog Nest ?20:11
*** athomas has joined #tripleo20:11
jristyes I know I'm late20:11
*** athomas has quit IRC20:11
*** akshai has quit IRC20:12
slaglejrist: how about "The Owl's Wisdom"20:13
jristI'm diggin' it20:13
slagleperhaps that's a stretch20:13
*** bvandenh has quit IRC20:13
*** dmk0202 has joined #tripleo20:14
*** akshai has joined #tripleo20:14
openstackgerritJohn Trowbridge proposed openstack/tripleo-quickstart: use yaml blocks for really long strings  https://review.openstack.org/31097920:14
openstackgerritJohn Trowbridge proposed openstack/tripleo-quickstart: lots and lots of comments (1/2)  https://review.openstack.org/31098020:15
openstackgerritJohn Trowbridge proposed openstack/tripleo-quickstart: lots and lots of comments (2/2)  https://review.openstack.org/31098120:16
*** larstobi has quit IRC20:17
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272320:19
*** larstobi has joined #tripleo20:20
*** shivrao has quit IRC20:22
openstackgerritMerged openstack/instack-undercloud: Add aodh as part of endpoints generation  https://review.openstack.org/30903620:23
*** jayg is now known as jayg|g0n320:24
*** jprovazn has quit IRC20:25
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272520:29
*** shivrao has joined #tripleo20:35
openstackgerritJohn Trowbridge proposed openstack/tripleo-quickstart: Template network-environment.yml with undercloud-scripts tag  https://review.openstack.org/31273020:37
trownayoung: ^ that should fix your adventure20:38
*** yamahata has joined #tripleo20:38
ayounggit review -d 31273020:38
pabelangerquestion, how do you test patches to openstack-infra/tripleo-ci?  There is only a noop check currently20:39
*** morazi has quit IRC20:39
openstackgerritPradeep Kilambi proposed openstack/tripleo-heat-templates: Use new redis profiles  https://review.openstack.org/31273220:39
trownpabelanger: there is the check-tripleo-jobs template in zuul/layout.yaml20:41
EmilienMpabelanger: jobs take time to run20:41
EmilienMpabelanger: but functional jobs run20:41
EmilienMyou can see them in zuul20:41
pabelangertrown: EmilienM Ah, okay. I missed that20:42
*** Goneri has quit IRC20:42
pabelangerI only thought there was a noop job20:42
trownpabelanger: and we are adding stable jobs there if you want to review :) https://review.openstack.org/#/c/297390/520:43
pabelangertrown: okay, will look20:44
pabelangerguess I should also an an experimental gate-tripleo-ci-centos-7-nonha to openstack-infra/tripleo-ci too20:45
trownpabelanger: what is that?20:46
pabelangerworking on migrating tripleo from fedora-22 to centos-7 for jenkins slave20:47
pabelangerthen you'll be using the same jenkins slaves as the rest of openstack20:47
trownoh sweet20:47
EmilienMpabelanger: that's awesome yeah20:49
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272520:55
*** rhallisey has quit IRC20:59
*** julim has quit IRC20:59
*** julim has joined #tripleo21:01
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272321:03
*** lblanchard has quit IRC21:11
*** MaxPC has quit IRC21:13
*** mbound has quit IRC21:13
openstackgerritMerged openstack/instack-undercloud: Add ability to auto-generate self-signed certificates  https://review.openstack.org/27323321:15
*** alextricity25_ has joined #tripleo21:15
bnemec\o/21:16
*** trown is now known as trown|outtypewww21:17
*** weshay has quit IRC21:20
*** dbecker has quit IRC21:20
EmilienMbnemec: ++21:21
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Enable undercloud ssl on nonha job  https://review.openstack.org/27374321:21
*** fultonj has quit IRC21:24
alextricity25_hey guys. I'm trying to build a precise image using diskimage-builder. Has anybody every seen this:21:24
alextricity25_installing linux-image-generic from ubuntu21:24
alextricity25_E: Unable to locate package jq21:24
alextricity25_install failed with error Reading package lists...21:24
*** ibravo has quit IRC21:25
*** pkunicki has quit IRC21:31
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272321:34
*** shivrao has quit IRC21:44
*** chlong has quit IRC21:45
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272321:46
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272521:46
*** shivrao has joined #tripleo21:46
*** fragatin_ has joined #tripleo21:50
*** fragatina has quit IRC21:51
*** fragatin_ has quit IRC21:57
*** ppradhan has quit IRC21:57
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272321:57
*** fragatina has joined #tripleo21:57
*** dmk0202 has quit IRC22:01
*** chlong has joined #tripleo22:05
*** Goneri has joined #tripleo22:06
*** julim has quit IRC22:06
openstackgerritimain proposed openstack/tripleo-common: WIP: Replace docker compose hook with a straight docker one.  https://review.openstack.org/31272322:08
openstackgerritOpenStack Proposal Bot proposed openstack/python-tripleoclient: Updated from global requirements  https://review.openstack.org/26852822:13
*** ccamacho|away has quit IRC22:13
openstackgerritBen Nemec proposed openstack/instack-undercloud: Generate most of the pystache context automatically  https://review.openstack.org/29731922:15
*** absubram has quit IRC22:15
*** ccamacho has joined #tripleo22:15
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272522:16
alextricity25_Hey, does anybody know how to include the precise-backports mirror to the ubuntu disk element?22:23
EmilienMprecise?22:24
EmilienMouch22:24
alextricity25_EmilienM: Yes. I'm trying to build a precise image22:24
alextricity25_hahaha22:24
EmilienMI'm working on xenial, sorry22:24
alextricity25_I'm trying to add a mirror to the /etc/apt/sources.list for my image22:26
alextricity25_Does anybody know how I can do that?22:26
*** mbound has joined #tripleo22:46
*** pradk has quit IRC23:01
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272523:11
*** alextricity25_ has quit IRC23:14
*** akshai has quit IRC23:16
*** akshai has joined #tripleo23:20
*** akshai has quit IRC23:21
pabelangerEmilienM: FYI, ubuntu-precise DIB does work.... just bringing in online for openstack-infra23:25
EmilienMpabelanger: ok cool23:25
pabelangerthat was with the ubuntu-minimal element23:30
*** apetrich has quit IRC23:32
openstackgerritPaul Belanger proposed openstack-infra/tripleo-ci: [WIP] Initial support for centos-7 jenkins slave  https://review.openstack.org/31272523:35
*** fragatina has quit IRC23:39
*** fragatina has joined #tripleo23:40
*** panda has quit IRC23:42
*** panda has joined #tripleo23:42
*** mbound has quit IRC23:48
*** dkehn__ is now known as dkehn_23:49

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!