Tuesday, 2016-08-30

openstackgerritRatnakar Kolli proposed openstack/tripleo-heat-templates: Adding optional aci-opflex hiera data  https://review.openstack.org/35266100:11
*** onovy has quit IRC00:11
*** onovy has joined #tripleo00:13
*** bkero has joined #tripleo00:17
*** thrash is now known as thrash|g0ne00:26
*** fragatin_ has quit IRC00:27
*** fragatina has joined #tripleo00:28
bkeroHow does tripleo deploy to nodepool multinode during 'openstack overcloud deploy'?00:28
openstackgerritEmilien Macchi proposed openstack-infra/tripleo-ci: Implement scenari001, 002 and 003  https://review.openstack.org/36250400:44
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: DNM: test scenario001  https://review.openstack.org/36250500:46
*** limao has joined #tripleo00:48
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: DNM: test scenario002  https://review.openstack.org/36250600:48
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: DNM: test scenario002  https://review.openstack.org/36250600:49
EmilienMbkero: it uses 2-nodes deployment00:51
EmilienMit doesnt use ironic boot etc00:51
bkeroEmilienM: Yeah, I'm wondering what the heat templates and 'openstack overcloud deploy' look like for those, since it doesn't need to use the baremetal stuff since the host is already provisioned.00:52
EmilienMbkero: have you looked at openstack-infra/tripleo-ci?00:53
bkeroEmilienM: I have. In particular I've been reading scripts/tripleo.sh, especially around OVERCLOUD_DEPLOY_ARGS00:54
bkerobut AFAICT it's not using any ARGS, so I'm guessing this is handled through heat templates.00:55
openstackgerritRyan Brady proposed openstack/tripleo-common: WIP wire in jinja templating for custom roles  https://review.openstack.org/36246500:57
*** tzumainn has quit IRC00:58
bkeroThis might be what I'm looking for. http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/toci_gate_test.sh#n15801:05
*** xuao has joined #tripleo01:09
EmilienMbkero: indeed. Every job has its own variables defined here.01:13
*** bana_k has quit IRC01:22
*** xuao has quit IRC01:23
openstackgerritMerged openstack/puppet-tripleo: Add ML2/OVN support for neutron profile  https://review.openstack.org/31487501:23
*** fragatina has quit IRC01:30
*** saneax is now known as saneax-_-|AFK01:32
*** chem has quit IRC01:37
*** jcoufal has quit IRC02:04
*** bana_k has joined #tripleo02:07
openstackgerritEmilien Macchi proposed openstack-infra/tripleo-ci: Implement scenari001, 002 and 003  https://review.openstack.org/36250402:08
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: DNM: test scenario001  https://review.openstack.org/36250502:09
*** dtrainor has quit IRC02:18
*** bana_k has quit IRC02:22
*** kbyrne has quit IRC02:22
*** kbyrne has joined #tripleo02:24
*** Ryjedo_ has joined #tripleo02:31
*** Ryjedo has quit IRC02:32
*** Ryjedo_ is now known as Ryjedo02:32
*** jefrite has quit IRC02:39
*** jdob has quit IRC02:43
*** jefrite has joined #tripleo02:43
*** jcoufal has joined #tripleo02:45
*** yamahata has quit IRC03:13
openstackgerritKarthik S proposed openstack/tripleo-heat-templates: Customize the reservation of memory for host processes  https://review.openstack.org/32770503:13
openstackgerritKarthik S proposed openstack/tripleo-heat-templates: Added DPDK environment file and nic configs samples  https://review.openstack.org/36123803:20
*** jeckersb is now known as jeckersb_gone03:26
*** anbu has quit IRC03:32
openstackgerritOpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements  https://review.openstack.org/36150103:34
*** xinwu has joined #tripleo03:41
*** jcoufal has quit IRC03:41
colonwqEmilienM, I have verified RGW/Glance/IPv4/NoSSL. I hope to verify SSL tomorrow. http://paste.openstack.org/show/564728/03:41
*** limao has quit IRC03:50
*** links has joined #tripleo03:51
*** xinwu has quit IRC03:55
*** xinwu has joined #tripleo04:00
*** fragatina has joined #tripleo04:02
*** numans has joined #tripleo04:03
*** numans has quit IRC04:04
*** numans has joined #tripleo04:04
*** akrzos_ has quit IRC04:07
*** fragatina has quit IRC04:07
*** bana_k has joined #tripleo04:09
*** akrzos has joined #tripleo04:09
*** xinwu has quit IRC04:13
*** bana_k has quit IRC04:16
*** limao has joined #tripleo04:17
*** xinwu has joined #tripleo04:22
*** anbu has joined #tripleo04:28
*** xinwu has quit IRC04:31
*** jrist has quit IRC04:33
*** limao has quit IRC04:33
*** jrist has joined #tripleo04:34
*** ayoung has quit IRC04:58
*** limao has joined #tripleo05:03
*** jaosorior has joined #tripleo05:10
*** lifeless_ is now known as lifeless05:15
*** skramaja has joined #tripleo05:26
*** bkopilov has joined #tripleo05:27
*** xinwu has joined #tripleo05:31
*** bvandenh has quit IRC05:45
*** xinwu has quit IRC05:48
*** milan has joined #tripleo05:50
*** xinwu has joined #tripleo05:50
*** apetrich has quit IRC05:53
*** apetrich has joined #tripleo05:55
*** milan has quit IRC05:59
*** ebalduf has quit IRC06:00
openstackgerritSaravanan KR proposed openstack/python-tripleoclient: Tripleoclient leaks temporary files  https://review.openstack.org/33063806:01
*** oshvartz has quit IRC06:01
*** milan has joined #tripleo06:04
openstackgerritSaravanan KR proposed openstack/python-tripleoclient: Tripleoclient leaks temporary files  https://review.openstack.org/33063806:07
*** bfournie has quit IRC06:11
*** bfournie has joined #tripleo06:12
*** akrzos has quit IRC06:13
*** akrzos has joined #tripleo06:15
*** aqkhan_ has quit IRC06:18
*** aqkhan_ has joined #tripleo06:18
*** bana_k has joined #tripleo06:19
*** rcernin has joined #tripleo06:19
*** rhefner has quit IRC06:20
*** paramite has joined #tripleo06:21
*** xinwu has quit IRC06:22
*** pgadiya has joined #tripleo06:25
*** pcaruana has joined #tripleo06:26
*** leanderthal|afk is now known as leanderthal06:30
*** bvandenh has joined #tripleo06:32
*** florianf has joined #tripleo06:33
*** anbu has quit IRC06:37
*** ubijtsa has joined #tripleo06:39
*** ubijtsa is now known as Guest5456806:39
*** bana_k has quit IRC06:39
*** Guest54568 has quit IRC06:42
*** Guest54568 has joined #tripleo06:42
*** Guest54568 is now known as assassin06:43
*** HoloIRCUser6 has joined #tripleo06:44
*** masco has joined #tripleo06:46
*** HoloIRCUser6 is now known as paramite|cell06:46
*** xinwu has joined #tripleo06:46
*** jbadiapa has joined #tripleo06:47
*** saneax-_-|AFK is now known as saneax06:47
*** zaneb has quit IRC06:48
*** zaneb has joined #tripleo06:48
*** rasca has joined #tripleo06:50
*** liverpooler has joined #tripleo06:51
*** aufi has joined #tripleo06:52
*** oshvartz has joined #tripleo06:53
*** jprovazn has joined #tripleo06:55
*** mhenkel has joined #tripleo06:58
*** mcornea has joined #tripleo07:02
*** tesseract- has joined #tripleo07:07
*** zoli_gone-proxy is now known as zoliXXL07:09
*** dsariel has joined #tripleo07:09
*** jpena|off is now known as jpena07:12
*** dtantsur|afk is now known as dtantsur07:13
openstackgerritMerged openstack/tripleo-ui: Parameters listing and updating  https://review.openstack.org/35155407:19
*** tremble has joined #tripleo07:20
*** ifarkas_afk is now known as ifarkas07:22
*** milan has quit IRC07:22
*** milan has joined #tripleo07:24
*** xinwu has quit IRC07:28
*** hewbrocca-afk is now known as hewbrocca07:29
hewbroccaMorning folks07:30
*** tvignaud has quit IRC07:31
*** akuznetsov has joined #tripleo07:37
*** bvandenh has quit IRC07:38
*** bvandenh has joined #tripleo07:38
*** jpich has joined #tripleo07:40
*** abregman has joined #tripleo07:40
*** fzdarsky|afk has joined #tripleo07:41
*** tremble has quit IRC07:42
*** fzdarsky|afk is now known as fzdarsky07:44
*** tremble has joined #tripleo07:45
openstackgerritmathieu bultel proposed openstack/tripleo-common: Use Openstack cli instead in order to get the instance status  https://review.openstack.org/35399607:45
*** ohamada has joined #tripleo07:46
*** anbu has joined #tripleo07:46
shadowerd0ugal: last three to go: https://review.openstack.org/#/c/353899/ :-) (mistral validations)07:48
honzahello folks, the puppet module for apache is deleting the tripleo-ui config file when it runs --- is there a common fix for this kind of issue?07:49
honzaline 5007:49
*** yolanda has joined #tripleo07:51
d0ugalshadower: I'll trade you for reviews of this series ;) This is the first: https://review.openstack.org/#/c/359784/07:53
shadowerd0ugal: fair deal07:53
jaosoriorhonza: tripleo-ui also needs to be in the puppet catalog for it not to get deleted07:54
jaosoriorhonza: I stumbled upon this same issue yesterday07:54
d0ugalshadower: Going out for a run now, I'll look at the remaining validations patches when I get back.07:54
d0ugalshadower: I actually started reviewing the first earlier.07:54
* d0ugal will bbiab07:54
shadowerd0ugal: enjoy your run!07:54
honzajaosorior: i see!  what repo is this in?07:54
jaosoriorhonza: what?07:55
jaosoriorhonza: the deletion of stuff you mean? That should be in puppetlabs-apache07:55
jaosoriorhonza: but yeah, basically any apache-related configuration for tripleo-ui needs to be done via puppet. out-of-band stuff will get deleted.07:55
honzajaosorior: sounds like i need to add tripleo-ui to the catalogue as you said, so i'm wondering where this place is07:56
honzainteresting, right now it's done via rpm07:56
jaosoriorhonza: and that's fine... but you need to add your configurations somewhere. probably here: https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/ui.pp07:57
jaosoriorhonza: yesterday I did the mistake of just running the setup of an apache module locally. And it whiped my keystone configuration from the undercloud X_x07:57
jaosoriorso yeah, it doesn't work like one would think it does. But I guess there's not much to do about it :/07:58
honzajaosorior: i don't know much about puppet so i'm trying to orient myself07:58
honzajaosorior: but this sounds like a good starting point07:58
openstackgerritMatthias Runge proposed openstack/tripleo-heat-templates: Clarify horizon allowed hosts setting  https://review.openstack.org/36268307:59
jaosoriorhonza: If you can point to someone what exact configurations you need to have for tripleo-ui. I'm sure someone here can help out getting those to puppet.07:59
honzaall i need is to tell puppet to not delete /etc/httpd/conf.d/openstack-tripleo-ui.conf :)08:02
*** r-mibu has quit IRC08:02
*** r-mibu has joined #tripleo08:03
dtantsurvicious puppet...08:03
*** shardy has joined #tripleo08:03
*** derekh has joined #tripleo08:07
openstackgerritJiri Stransky proposed openstack/puppet-tripleo: Defaults to ceph_mon_node_ips and ceph_mon_node_names  https://review.openstack.org/36110808:07
*** bvandenh has quit IRC08:12
*** HoloIRCUser6 has joined #tripleo08:15
*** paramite|cell has quit IRC08:15
shardytherve: Hey, thanks for you yaql help on the ML :)08:18
shardytherve: I send another update showing what we need, any help appreciated, I'll be sure to credit you as a co-author when we work it out ;)08:18
*** HoloIRCUser6 has quit IRC08:18
*** HoloIRCUser7 has joined #tripleo08:18
*** HoloIRCUser7 is now known as paramite|cell08:21
jaosoriorshardy, jistr: Can you guys check this CR out? https://review.openstack.org/#/c/359645/ I've been testing it and mcornea just tried it out too manually08:23
*** gfidente has joined #tripleo08:25
jistrjaosorior: looks good, +A08:26
jaosoriorjistr: thanks dude08:26
jistrgfidente: morning, i published a minor edit on https://review.openstack.org/361108 and also rechecked the t-h-t patch that depends on it08:26
jistrgfidente: just a bracket that we missed08:27
*** vijaykc4 has joined #tripleo08:28
gfidentehey morning08:28
*** milan has quit IRC08:29
gfidentethanks :)08:29
*** yolanda has quit IRC08:33
Jokke_morning folks08:34
*** abregman has quit IRC08:35
*** electrofelix has joined #tripleo08:35
honzashardy: do you happen to know how to make puppet not delete my apache config file?08:36
therveshardy, No problem, glad I could help!08:37
shardytherve: yup, your example is really really close, I'm just trying to figure out how to flatten the list-of-lists vs selecting the first item08:38
therveshardy, There is nifty CL tool in the yaql repo, I don't know if you saw it08:38
shardytherve: No actually I didn't - I've been testing via heat with a minimal template08:38
openstackgerritMartin AndrĂ© proposed openstack/instack-undercloud: Unit tests for _create_default_plan()  https://review.openstack.org/36270108:39
therveIf you type "yaql" in the heat venv you should see it08:39
shardyaha, yeah, that may speed things up a bit08:39
*** chem has joined #tripleo08:40
shardytbh the main problem is lack of docs, I end up reading the yaql code/tests when figuring out what syntax is needed08:40
shardyI know there's an effort underway to improve that tho08:40
*** vijaykc4 has quit IRC08:46
*** abregman has joined #tripleo08:47
therveYeah I'm getting better at reading the source as well, but it's not an easy process08:48
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Move service/endpoint configuration to keystone profile  https://review.openstack.org/35968108:52
*** bvandenh has joined #tripleo08:55
*** links has quit IRC09:00
therveshardy, OK I may have one solution, but it's not pretty09:02
thervedict($.selectMany($.items()).groupBy($[0], $[1]).select($[0]).zip($.selectMany($.items()).groupBy($[0], $[1]).select($[1].flatten())))09:02
therveIt may look like I'm trying to crash your IRC client, but I'm not09:02
shardytherve: hehe, thanks - I'll give it a try09:03
shardyI got pretty close with a different approach:09:03
shardy$.data.l.selectMany($.items()).groupBy($[0], $[1].reduce($1 + $2))09:03
openstackgerritFlorian Fuchs proposed openstack/tripleo-ui: Use Mistral/Zaqar for folder-based plan-creation  https://review.openstack.org/36055009:04
shardythe problem is the reduce reduces the nested lists, not just the wrapper from the select09:04
shardytherve: if I just wrote a map_deep_merge function for heat right now, is there any chance we can land it for newton?09:05
shardythis yaql stuff is kinda fun, but it's pretty scary from a maintenance perspective ;)09:06
therveshardy, Possibly. I don't feel super stressed about the n3 deadline TBH09:06
shardytherve: Ok, I'll play with the yaql a little bit more, make some coffee and ponder the next move09:06
shardytherve: context is, this is needed for our custom-roles feature, which is going to be a FFE09:07
jaosoriormcornea: did you get a chance to test this one out? https://review.openstack.org/#/c/35536609:07
therveshardy, Yeah, I'm not super confident we should ship template with those undebuggable calls09:07
mcorneajaosorior: currently deploying, will update once it's done09:07
jaosoriormcornea: awesome :D09:08
shardytherve: Yeah, if we do then I'd say it has to be a stopgap until heat provides a cleaner way to do the same transformation09:08
shardythere are a few other yaql-isms landing where I'd like to refactor into heat native calls for ocata anyway09:08
*** jaosorior is now known as jaosorior_lunch09:08
therveshardy, dict($.selectMany($.items()).groupBy($[0], $[1], [$[0], $[1].flatten()])) looks more reasonable09:11
therveSeems to work with the basic example I have, and I think I can mostly comprehend it :)09:12
shardytherve: woohoo! It works! :)09:13
shardythanks :)09:13
therveNo problem :)09:13
shardyI'll do some more testing, but that seems like it may do what we need, and I'll add the map_deep_merge thing on my todo list for ocata09:13
*** vijaykc4 has joined #tripleo09:14
*** links has joined #tripleo09:15
*** limao has quit IRC09:17
vijaykc4shardy, we are waiting for the OVB ha to clear our patches (SRIOV/DPDK)  and it has been consistently failing... i was wondering if this continues what we need to do considering today being the last day.09:19
*** limao has joined #tripleo09:22
openstackgerritMerged openstack/tripleo-heat-templates: Fix check of rpm-python.  https://review.openstack.org/35702209:23
*** chem has quit IRC09:34
*** chem has joined #tripleo09:35
openstackgerritMerged openstack/puppet-tripleo: Handle galera_node_names being an array  https://review.openstack.org/36204609:37
shardyvijaykc4: do you have a link to the failing patches please?09:40
shardyvijaykc4: if patches don't land due to CI issues they will most likely be considered valid feature freeze exceptions09:40
jaosorior_lunchseems to me like some merrors are having issues for some reason09:41
vijaykc4shardy, https://review.openstack.org/#/c/327705/ https://review.openstack.org/#/c/361430/09:41
openstackgerritMerged openstack-infra/tripleo-ci: Inject CA certificate via service profile  https://review.openstack.org/35779109:41
jaosorior_lunchshardy: what are the requirements for applying for an FFE? I want to submit the TLS work as an FFE, but I haven't finished all the patches (some depend on the composable roles work... somewhat)09:42
*** jaosorior_lunch is now known as jaosorior09:42
openstackgerritMerged openstack/tripleo-heat-templates: Conditionally expose haproxy endpoints  https://review.openstack.org/35964509:42
shardyjaosorior: I replied to your ML question, did you see my response?09:42
jaosoriorshardy: ah, let me check that out09:42
shardyjaosorior: basically we've got to balance the risk vs benefit09:42
shardyif a feature is nearly complete and/or low risk I think a FFE is fine09:43
*** tosky has joined #tripleo09:43
shardyif it's large, incomplete and/or risky we have to ensure there's a very strong justification of why we don't defer to ocata09:43
jaosoriorshardy: in this case I would say medium. most of the blockers have been worked on. But some composable roles work is still needed before the haproxy parts09:43
saneaxshardy, it seems like the failures are on timeout of dismantling the stack? https://paste.fedoraproject.org/417163/50205147/09:44
jistryea the nova memory reservation patch went all the way to successful pingtest and then it timed out http://logs.openstack.org/30/361430/3/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/22f09dc/console.html#_2016-08-30_08_44_09_13930109:44
jistrjust +2'd09:44
saneaxshardy, http://logs.openstack.org/05/327705/12/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/6acd1ca/console.html#_2016-08-30_08_52_33_84615209:45
shardyjaosorior: ack - I think your tls stuff should be a valid FFE, but we really need a mail which clearly enumerates the remaining patches, so we can focus review attention, and help with any remaining WIP pieces09:45
saneaxjistr, seems like same issue09:45
jaosoriorshardy: alright, will do09:46
shardysaneax: Yeah in that case I think the job hit the infra timeout09:47
shardylately we've been seeing that a lot, we have to figure out ways to optimize such that the job walltime decreases09:48
mandred0ugal: remember the mistral workflow to copy SSH keys I told you about yesterday?09:48
shardyfor now a recheck is valid, I think provided all jobs pass on some combination of rechecks we can merge the patches09:48
mandred0ugal: well... it turns out it needs to run at the end of overcloud deploy, because it copies the keys to the overcloud nodes09:49
saneaxshardy, doing a once more recheck :) after previous two failure at the same point09:50
vijaykc4shardy, thanks for the info..one is a samples  patch and two T-H-T's (all low risks and good to have patches) i believe if CI continues to fail we should be eligible for FFE. Yes for now we will continue to recheck.09:50
shardyvijaykc4: ack, thanks - I'll definitely support FFEs for anything blocked by CI09:51
mandred0ugal: I suppose my code woud go there? https://github.com/openstack/python-tripleoclient/blob/master/tripleoclient/v1/overcloud_deploy.py#L94909:51
shardythanks for your patience, hopefully we can work on a proper fix for these timeouts very soon09:51
jaosoriorjistr: hey dude, with the FQDN stuff that landed, I made a profile to add those VIP hostnames to /etc/hosts in the nodes. Can you check it out? https://review.openstack.org/#/c/357765/09:51
vijaykc4shardy, JFYI these are the three patches, https://review.openstack.org/#/c/327705/ , https://review.openstack.org/#/c/361430/ and https://review.openstack.org/#/c/361238/ Thanks for your inputs and hopefully we can close them today.09:55
*** vijaykc4 has quit IRC09:57
d0ugalmandre: I would rather it was part of the overcloud deploy workflow, we don't want more logic in the CLI.09:57
d0ugalmandre: Is there a way we can automatically detect if validations have been enabled?09:57
mandred0ugal: there is an enable_validations hiera key09:59
mandrecan we easily retrieve the value from the workflow?09:59
*** nyechiel has joined #tripleo10:01
*** vijaykc4 has joined #tripleo10:02
d0ugalmandre: good question, I guess so, but I am not sure.10:03
mandred0ugal: it might require a custom mistral action10:03
*** r2bit__ is now known as r2bit10:03
*** r2bit has quit IRC10:04
*** r2bit has joined #tripleo10:04
d0ugalmandre: Yeah, it will, I think10:04
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Add HAProxy TLS handled by certmonger as composable service  https://review.openstack.org/35643010:07
*** akrivoka has joined #tripleo10:07
openstackgerritMerged openstack/tripleo-validations: Added check for keystone token_flush cronjob.  https://review.openstack.org/35555710:07
dtantsurfolks, https://review.openstack.org/#/c/361165/ passed both gates in several attempts, could you please merge it?10:08
mandred0ugal: we're talking about the deploy_plan workflow, right?10:08
dtantsurotherwise future changes in puppet might break tripleo10:09
shardydtantsur: +2, I agree I think we can probably merge it10:11
*** ebarrera has joined #tripleo10:13
d0ugalmandre: Yeah, but we have not really added anything optional to it yet so it would be good to have a discussion with a few folks about it10:13
d0ugalmandre: rbrady is looking at adding something else optional that is based on shardy's work: https://review.openstack.org/#/c/362465/10:14
dtantsuractually, we're delaying https://review.openstack.org/#/c/359138/ because otherwise it'll break tripleo10:14
*** vijaykc4 has quit IRC10:16
*** paramite|cell has quit IRC10:17
d0ugalmandre: (that linked patch is integrating in the process templates action which is called at deploy time btw)10:18
d0ugalmandre: I'd expect for your integration, you would want a new action that is triggered in the deploy workflow.10:18
*** HoloIRCUser4 has joined #tripleo10:18
*** adarazs is now known as adarazs_lunch10:19
*** vijaykc4 has joined #tripleo10:20
*** HoloIRCUser6 has joined #tripleo10:21
*** vijaykc4 has quit IRC10:21
*** HoloIRCUser7 has joined #tripleo10:22
*** HoloIRCUser6 has quit IRC10:22
*** HoloIRCUser6 has joined #tripleo10:23
*** HoloIRCUser7 has quit IRC10:23
*** ebarrera has quit IRC10:23
*** HoloIRCUser6 has quit IRC10:24
*** limao has quit IRC10:24
*** HoloIRCUser8 has joined #tripleo10:24
*** HoloIRCUser4 has quit IRC10:25
*** HoloIRCUser6 has joined #tripleo10:25
*** HoloIRCUser8 has quit IRC10:26
*** HoloIRCUser4 has joined #tripleo10:26
*** HoloIRCUser6 has quit IRC10:26
*** apetrich has quit IRC10:28
*** HoloIRCUser4 has quit IRC10:29
*** HoloIRCUser4 has joined #tripleo10:30
d0ugalshardy: Hey10:31
*** paramite has quit IRC10:31
jaosoriord0ugal: hey dude, trying to review https://review.openstack.org/#/c/360305/19 but am I bit confused about the plan deploy10:32
jaosoriorwhat's up with that?10:32
d0ugaljaosorior: So, with the new Mistral workflows we have plans that are created and stored in swift. Without that command there wont actually be a way to deploy plans from the CLI10:33
d0ugaljaosorior: The current command is being updated to create plans in the fly, but to be backwards compatible it doesn't really follow the workflow.10:33
*** HoloIRCUser4 has quit IRC10:33
*** HoloIRCUser6 has joined #tripleo10:33
jaosoriord0ugal: wasn't plan the old term used for tuskar?10:34
d0ugaljaosorior: Yeah, it's back.10:34
d0ugaljtomasek: You use "plan" in the GUI, right?10:34
dtantsurjaosorior, hey! when you have a second, mind reviewing https://review.openstack.org/#/c/361165/ please? it's small and actually passed both gates (in 2 attempts)10:35
jaosoriord0ugal: I think that since that term was being used before, it's quite confusing... so, either another term should be used. Or more documentation should be added10:35
jaosoriordtantsur: sure thing. Just one question there. Weren't we using the pxe element still? seems like this stuff should go alongside the deletion of that element, no10:35
d0ugaljaosorior: Okay, I guess we will need to go with more documentation. Changing it now is probably a non-starter. I have started some related documentation here but I probably need to tell people what a plan is. https://review.openstack.org/35868510:36
jtomasekd0ugal: yes10:36
dtantsurjaosorior, no, these are not related. this is only for ironic.conf, and pxe element is for building the actual environment10:36
jaosoriordtantsur: got it10:36
*** HoloIRCUser8 has joined #tripleo10:36
d0ugaljtomasek: phew, just a sanity check, thanks :)10:36
*** apetrich has joined #tripleo10:37
jaosoriord0ugal: I'm fine if it's just more documentation10:37
jtomasekd0ugal: we call it deployment plan10:37
d0ugaljtomasek: That is a long name for the CLI :) but maybe we should have used it...10:38
jaosoriorso... these new plans in CLI are the same as deployment plans in UI?10:38
jtomasekjaosorior: yes10:39
jaosoriorsorry for the basic questions, things got confusing terms-wise10:39
jtomasekjaosorior: the general reason is that gui can't operate with files in the filesystem, so a plan resembles to a set of templates10:39
d0ugaljaosorior: Thanks!10:40
*** HoloIRCUser9 has joined #tripleo10:40
*** HoloIRCUser6 has quit IRC10:40
jaosoriord0ugal: I would really suggest you write a blog post or something of the sort with the ideas behind the mistral work10:41
jaosoriord0ugal: that way it'll be easier for folks to review10:41
d0ugaljaosorior: Sure, I guess it is best wrapped up in the spec at the moment. Adding something that is updated and relevant now to the tripleo-docs would be a good idea.10:42
jaosoriorthat would be good too10:42
d0ugaljaosorior: Once people stop me landing code for N I'll move my focus to documentation :)10:43
jaosoriorgfidente: ping10:43
*** HoloIRCUser8 has quit IRC10:43
*** ebarrera has joined #tripleo10:45
mcorneajaosorior: https://review.openstack.org/#/c/355366 looked good in my testing10:46
jaosoriormcornea: thanks dude!10:46
gfidentejaosorior, pong10:47
gfidentejistr, think I found the issue in https://review.openstack.org/#/c/355067/10:48
jaosoriorgfidente: Hey dude, can you check out again the commit adding the VIP hosts to /etc/hosts? https://review.openstack.org/#/c/357765/10:50
jaosoriorgfidente: I'm still unsure if we should have that enabled by default or just enable it via an environment file10:50
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Move ceph cluster config out of allNodesConfig  https://review.openstack.org/35506710:52
gfidentejaosorior, I think if we can replace the hosts element with that10:53
gfidentewe should and we should enable it by default10:53
gfidentebut if we can't, or if we can't by the code freeze, then probably not10:53
gfidenteand maybe enable it from one of the tls endpoints10:53
*** ebarrera has quit IRC10:53
gfidenteto start with, until we can enable it by default when deprecating the hosts element10:53
jaosoriorgfidente: I actually don't remember which element was that :/10:56
*** sshnaidm|afk is now known as sshnaidm10:56
*** pkovar has joined #tripleo10:57
*** vijaykc4 has joined #tripleo10:59
*** rhallisey has joined #tripleo10:59
gfidentejaosorior, I think it's named just hosts11:02
jaosoriorgfidente: so, right now the only hosts we set are the ones for the specific nodes. We never set the overcloud VIP hosts11:02
jaosoriorgfidente: that's why I tried to introduce that profile11:03
*** fzdarsky is now known as fzdarsky|lunch11:03
jaosoriorgfidente: I guess it wouldn't be too hard to get the other hosts on another profile too11:03
gfidentebut it feels like we don't want competing stuff editing the hosts file11:03
jaosoriorgfidente: well, it seems to me that they get ran in different times. Do they're not really competing in that sense11:07
*** zoliXXL is now known as zoli|lunch11:08
jistrjaosorior, gfidente: i wonder how confined the methods are, that is if they just manage individual entries vs. the whole file. I guess puppet would be entry based but not sure about the element11:09
jaosoriorjistr: the element gets ran before puppet AFAIK11:09
jaosoriorjistr: and puppet is quite entry-based for that resource11:10
dtantsurjistr, hey! that update problem, was it fixed after all?11:10
jistrjaosorior: yea but on stack-update it could destroy the VIP records for a period of time before puppet runs again. Potentially, that is. I'm still not sure if the element actually rewrites the file.11:11
* jistr searches11:11
jistrdtantsur: yup, should be all right, but i'm not sure if it's just in trunk or also in current-tripleo pin11:11
*** apetrich has quit IRC11:11
* jaosorior is running an update with the VIP hosts profile right now11:11
gfidentejistr, jaosorior the script is this one https://github.com/openstack/tripleo-image-elements/blob/master/elements/hosts/os-refresh-config/configure.d/51-hosts11:12
jaosoriordaaaamn that's some funky bash right there11:13
gfidentejaosorior, yeah and it does get re-run at every os-refresh-config11:13
jistrjaosorior: yea trying to decrypt that :D11:14
sshnaidmdoes anybody know about keystone db-sync problem?11:14
jaosoriorgfidente: it seems to have different paths whether it finds the HEAT_HOSTS_START tag or not11:14
jaosoriorsshnaidm: which problem?11:14
sshnaidmAttributeError: 'module' object has no attribute 'RegionInvalidationStrategy'11:14
gfidenteI like especially the f=1 part11:14
sshnaidmin undercloud installation11:14
jaosoriorsshnaidm: no idea dude. I had other issues with the undercloud related to augeas11:15
sshnaidmjaosorior, and what are they?11:15
jaosoriorthe augeas call on lvm.conf fails11:15
jaosoriorsshnaidm: no logs, it just says it fails11:15
jaosoriorsshnaidm: I tried for a LONG time yesterday to get something relevant about that failure and couldn't11:16
jaosoriorjust makes the install fails and doesn't report anything else :(11:16
jaosoriorhappens on update for the undercloud11:16
sshnaidmjaosorior, which jobs does it affect?11:16
jaosoriorsshnaidm: locally11:16
sshnaidmundercloud update?11:16
jaosoriorsshnaidm: no jobs that I know of11:16
jaosoriorsshnaidm: yeah11:16
sshnaidmoh, I see11:16
jaosoriorsshnaidm: when trying to update to the latest undercloud in current11:17
sshnaidmmy errors are in periodic jobs..11:17
jaosoriorsshnaidm: well, it does sound like an actual error in keystone11:17
sshnaidmjaosorior, for example: http://logs.openstack.org/periodic/periodic-tripleo-ci-centos-7-ovb-nonha/37fe6d7/logs/undercloud/var/log/undercloud_install.txt.gz#_2016-08-30_06_29_34_00011:18
*** vijaykc4 has quit IRC11:18
*** milan has joined #tripleo11:18
jaosoriorjistr, gfidente: running an update seems to leave things in a pretty good state. I can see the stuff generated by the element like it was before. And the stuff generated by the profile is there too.11:18
jaosoriorsshnaidm: yep, looks like an actual keystone issue11:19
*** oshvartz has quit IRC11:19
*** ramishra has quit IRC11:21
*** oshvartz has joined #tripleo11:21
jaosoriorgfidente, jistr: Ah... well, at least judging from the code. It seems to me like the element overwrites the whole file every time it gets ran. the reason why I see that the puppet stuff is still there is cause on update it also gets ran; so the file ends up being updated first by the element and afterwards by puppet11:22
jaosoriorgfidente: jistr https://github.com/openstack/tripleo-image-elements/blob/master/elements/hosts/os-refresh-config/configure.d/51-hosts#L2411:23
*** ramishra has joined #tripleo11:23
jistrjaosorior: can you try to write some custom gibberish outside of the HEAT_HOSTS_START/END and see if it gets preserved?11:23
jistrjaosorior: i have a suspicion that yes, it does rewrite the whole file, but it has logic built in that awk madness that preserves everything outside those markers11:24
jaosoriorjistr: sure, I can do that. it'll just take a bit cause I'm updating an HA deployment.11:24
hewbroccaawk madness11:25
* hewbrocca despairs11:25
*** apetrich has joined #tripleo11:25
* hewbrocca rechecks that it is in fact 201611:26
* hewbrocca looks for a better profession11:26
*** vijaykc4 has joined #tripleo11:27
jaosoriormcornea: still have that deployment up?11:27
mcorneajaosorior: no but I'm currently redeploying11:28
jaosoriormcornea: if you do an update can you add a host in /etc/hosts to any controller?11:29
jaosorioroutside of the HEAT_HOSTS_START/END tags?11:29
mcorneajaosorior: I'll try. fwiw I know this worked in the past11:30
jaosoriormcornea: so the stuff that gets added outside of those tags gets preserved?11:31
mcorneajaosorior: yes, I tried that on mitaka - https://bugzilla.redhat.com/show_bug.cgi?id=1360398#c2211:32
openstackbugzilla.redhat.com bug 1360398 in rhel-osp-director "In IPv6 deployment cluster fails when 1/2 DNS servers are unreachable" [Urgent,New] - Assigned to athomas11:32
jaosoriorjistr, gfidente ^^11:32
openstackgerritMerged openstack/tripleo-common: Use Openstack cli instead in order to get the instance status  https://review.openstack.org/35399611:32
openstackgerritMerged openstack/tripleo-common: Workflows to run validations  https://review.openstack.org/35389911:33
openstackgerritMerged openstack/python-tripleoclient: Fix doc page for overcloud deploy  https://review.openstack.org/35857711:33
openstackgerritMerged openstack/instack-undercloud: Handling defaults for pxe_bootfile_name and pxe_config_template moved to puppet  https://review.openstack.org/36116511:33
jistrjaosorior, mcornea, gfidente: well if that survives stack-update then i think we're good11:34
jaosoriorgfidente, jistr: so now the question remains. Do you think that profile I added should be default in the overcloud deploy?11:35
*** akrivoka has quit IRC11:37
*** thrash|g0ne is now known as thrash11:38
*** oshvartz has quit IRC11:39
*** vijaykc4 has quit IRC11:39
*** maeca1 has joined #tripleo11:40
*** vijaykc4 has joined #tripleo11:40
rbradyshardy: ping re: j211:46
*** morazi has joined #tripleo11:47
karthiksshardy,  jaosorior  jistr  These reviews has +2 and all green in CI.  Need your attention.  https://review.openstack.org/#/c/327705/  https://review.openstack.org/#/c/361238/11:47
*** lucasagomes is now known as lucas-hungry11:47
*** akrivoka has joined #tripleo11:49
*** adarazs_lunch is now known as adarazs11:49
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Default haproxy listeners activation on hiera service _enabled  https://review.openstack.org/36279811:52
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Generate composable service node_names lists  https://review.openstack.org/35506811:53
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Upgrade ceph-osd  https://review.openstack.org/35731211:55
*** jayg|g0n3 is now known as jayg11:56
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Cleanup haproxy service from explicit services enablement  https://review.openstack.org/36280011:56
shardyrbrady: Hi!11:56
shardykarthiks: thanks, looking11:56
karthiksThanks shardy gfidente11:57
shardykarthiks: when those two merge, can we declare tripleo-ovs-dpdk complete?11:58
rbradyshardy: I just had a couple of questions...does the result of the jinja template need to be used as the value of the 'template' key in the dict we send to heat for deployment?11:59
karthiksyes shardy ... 1 pending for SRIOV which is still failing on CI11:59
jistrhmm i'm leaning a bit more towards yes... The default deployment sorta assumes that full DNS is not in place anyway, as it writes manual /etc/hosts entries for all the nodes. Seems expectable to me that we'd do it for VIPs too. I wonder if we should fix the default value for CloudName in that case though.11:59
shardyrbrady: basically yes, I wrote it to overcloud.yaml, which would then be read and passed via heatclient by tripleoclient12:00
shardyrbrady: I would like it if we persisted the data as overcloud.yaml in the plan, even if we then pass it directly to heat as well12:00
jistrjaosorior: i think CloudName isn't used by default anywhere, right? We use IPs by default?12:00
shardyas it will help a lot with debugging the j2 template12:00
*** jpena is now known as jpena|lunch12:00
*** pkovar has quit IRC12:00
jistrjaosorior: just trying to think how safe it would be to change the default there12:01
shardyrbrady: my debug cycle with the tripleoclient patch was to start a deployment, then look at the /tmp/.../overcloud.yaml12:01
shardyso we'll need a similar approach with the swift stored plan I think12:01
jaosoriorgfidente: don't we have to provide a default here? https://review.openstack.org/#/c/362798/1/manifests/haproxy.pp12:01
jaosoriorjistr: well... I actually don't know. That's why I didn't change it12:02
*** trown|outtypewww is now known as trown12:02
*** pradk has joined #tripleo12:02
jaosoriorjistr: but then again, it shouldn't be too bad. Don't see it being used in that many places12:02
jaosoriorI can try12:02
shardyrbrady: I was thinking, particularly if it simplifies things, that we could just write out overcloud.yaml, then use the existing deployment workflow completely unmodified12:02
rbradyshardy: I think we'll need to.  when a plan is created, the capabilities map file is parsed and overcloud.yaml is identified as the root tempate.  that template is used for parameters management prior to deployment12:02
gfidentejaosorior, yep it's setting the defaults from l30612:02
shardyrbrady: ack, can we just do the j2 part prior to that step in the workflow?12:02
*** pkovar has joined #tripleo12:03
gfidenteI want to add the ceph_rgw listener, that's why I am changing that12:03
rbradyshardy: that's the path I started on after your email, but I'm just looking where it needs to happen at this point12:03
shardyI can imagine this being a pretty early step, e.g the first thing a custom-roles aware UI would do is ask for the list of role names I think12:03
jaosoriorgfidente: right now you can do it in puppet/services/haproxy.yaml that already takes those hiera parameters12:03
shardyjtomasek: ^^ FYI12:03
jaosoriorgfidente: but shouldn't it be hiera('keystone_enabled', false)  >?12:03
rbradyshardy: is this a one time operation or should there be a way to trigger this N times against a plan?12:03
*** jaosorior has quit IRC12:04
shardyrbrady: I think we'll need to allow re-running it, as folks might want to add a new role to an existing deployment12:04
shardyrbrady: the way I wired it into tripleoclient, we'd just run it every time "openstack overcloud deploy" was run12:05
*** jaosorior has joined #tripleo12:05
shardyso we probably need similar behavior via mistral I think12:05
gfidentejaosorior, oh a default for the hiera calls, better to have it yes, I'll do it12:05
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Change CloudName default value to include domain  https://review.openstack.org/36280612:05
shardyif the roles_data is unchanged, we'll just write out an identical overcloudl.yaml12:05
jaosoriorjistr: ^^12:05
*** fultonj has joined #tripleo12:06
*** mat128 has joined #tripleo12:07
shardyrbrady: so, I think that means store-plan -> optionally update roles_data -> j2 templating -> <normal deploy workflow>12:07
mat128good morning TripleO12:07
*** ebarrera has joined #tripleo12:07
jaosoriorjistr, gfidente: by the way, can you guys check that out? That sets the ports from haproxy to be configured by the endpoint map12:07
shardygood morning mat12812:07
*** rlandy has joined #tripleo12:08
mat128I've seen a TripleO failure that didnt "fail" my build, but I still wanted to act on it.12:08
gfidentejaosorior, which one?12:08
jaosoriorgfidente, jistr: https://review.openstack.org/#/c/355366/12:08
rbradyshardy: yep.  thanks.  I've separated the code into its own action again and now looking where to trigger.  we don't provide a means to update a plan, that is done directly through swift, yet I find myself now wanting some sort of post update trigger when plan files are changed12:09
* jtomasek reads back12:09
shardyrbrady: Yeah, I guess for now we can just say "optionally update roles_data" is a no-op, and later we can wire in pushing a modified yaml file into the plan12:10
shardyshould be easy enough to do directly via swift in the meantime?12:10
*** akuznetsov has quit IRC12:11
*** jaosorior has quit IRC12:11
*** mhenkel has quit IRC12:11
*** akuznetsov has joined #tripleo12:11
rbradyshardy: I'm thinking we have a AfterPlanUpdateAction that will need to be called manually at the moment via run-action or creating an execution12:12
*** jaosorior has joined #tripleo12:12
shardyrbrady: Ok, sounds good, either wfm12:12
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Default haproxy listeners activation on hiera service _enabled  https://review.openstack.org/36279812:13
jtomasekrbrady: so that action is going to be called as part of the plan creation, right?12:13
jistrjaosorior: is there a matching puppet-tripleo change or not yet?12:13
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: DONT MERGE: test periodic job  https://review.openstack.org/35585912:13
jtomasekflorianf: create plan is called as an action now in GUI?12:13
jaosoriorjistr: for what?12:13
rbradyshardy: unless there is some piece of swift middleware that sends zaqar notifications in the event of container or object updates :)12:13
jistrjaosorior: for the haproxy endpoint settings. Or it doesn't need one?12:13
jaosoriorjistr: it doesn't need one12:14
jaosoriorjistr: the interface has been there for a while. It was just never used12:14
karthiksEmilienM,  gfidente jaosorior jistr bnemec  Could you please have a look at the only outstanding patch for DPDK :) . Please do the honors  !!!!   https://review.openstack.org/#/c/361238/12:14
rbradyjtomasek: yeah, I'll have it run in the CreatePlanAction.  If you update files, you will need to call AfterUpdatePlanAction directly12:15
jtomasekrbrady: ack12:15
jtomasekrbrady: sounds good12:15
*** masco has quit IRC12:15
*** akuznetsov has quit IRC12:16
jtomasekshardy, rbrady: so the custom roles work is almost done?12:16
karthiksThanks EmilienM jaosorior !!!12:17
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: Replace ssh public key for sshnaidm  https://review.openstack.org/36281412:17
sshnaidmderekh, hi12:17
shardyjtomasek: it's getting closer, and wiring this in via mistral is a key part of completing it12:17
shardykarthiks: great, can you please update the implementation status of the blueprint when the patches merge?12:18
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Add Ceph RGW listener to HAProxy  https://review.openstack.org/36281612:18
jtomasekrbrady: roles listing patch will then get updated to accommodate that, right?12:18
karthiksSure shardy .12:18
shardyjtomasek: yes12:19
jtomasekshardy: does Heat send any zaqar messages when deployment is in progress?12:19
shardythe current roles listing patch will actually still work provided it runs after the j2 templating is completed12:19
jtomasekshardy: good point12:19
rbradyjtomasek, shardy: it should work without changes for now12:20
shardyjtomasek: You can define an event sink in the environment file, then all events can be sent to a zaqar queue12:20
shardyI don't think we currently do that though12:20
jtomasekshardy: ok, seems like we'll use polling for that for now12:20
shardyjtomasek: Yeah, but we should look at moving away from polling when possible12:21
florianfjtomasek: no, create plan is called as a workflow. list_plans is called as an action12:22
jtomasekshardy: yes, definitely something to discuss on the summit (or earlier) afaik, the deployment workflow is not able to run the whole time the deployment is in progress12:23
shardyjtomasek: Yeah, we risk hitting a token timeout because mistral can't yet reauthenticate via trusts12:24
shardyjtomasek: can't we just pass event_sinks to heat and have a queue defined to recieve the events from heat?12:24
florianfjtomasek: I wasn't aware, it was available as an action. Should change that in the UI.12:24
shardyseems like a fairly simple alternative to polling, although I've not tested it12:24
*** dprince has joined #tripleo12:26
jtomasekflorianf: hold it untill we know that rbrady calls that new jinja processing from the action or as a separate action in plan creation12:26
EmilienMshardy: have you looked yet why ci is not promoted since 1 week?12:26
florianfjtomasek: ok12:26
EmilienMshardy: if not i'll look today12:26
*** pkovar has quit IRC12:26
shardyEmilienM: Not yet, sorry - although ramishra did mention a regression related to deleting stacks in heat which I need to check on12:27
openstackgerritMerged openstack/tripleo-heat-templates: Upgrade ceph-mon  https://review.openstack.org/35713012:27
jtomasekshardy: is that a simple Heat api call to pass the event_sinks?12:27
shardyjtomasek: No it's an environment file you can pass, via the same method we pass all the environments defined e.g in the capabilities map12:28
shardyso we could have environments/ui-events.yaml or something12:28
shardywith a pre-defined queue name12:28
*** jdob has joined #tripleo12:29
shardyThe only gotcha is I suspect that interface won't yet support nested events12:29
shardyso we probably can't yet switch the CLI over to use it12:29
EmilienMpradk: i'm currently testing gnocchi in tripleo ci12:29
jtomasekshardy: yep, that could work, I am not 100% sure we're able to get it in for newton, but I totally see it as a viable solution for next12:29
EmilienMpradk: https://review.openstack.org/#/c/362505/ and https://review.openstack.org/#/c/362504/12:29
shardyjtomasek: yup, agreed, probably not high priority until we've got basic functionality proven12:30
EmilienMpradk: it fails to create OS::Aodh::GnocchiResourcesAlarm12:30
openstackgerritSagi Shnaidman proposed openstack/tripleo-incubator: Replace sshnaidm SSH public key  https://review.openstack.org/36282412:31
EmilienMpradk: I see this error http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/logs/subnode-2/var/log/gnocchi/app.txt.gz12:31
EmilienMpradk: have you seen that before?12:31
pradkEmilienM, sounds like its not able to talk to redis12:31
EmilienMpradk: right12:32
thrashd0ugal: was testing out 'overcloud plan create', and it just kinda hung there.12:32
EmilienMpradk: have you tried gnocchi using tripleo before?12:32
*** zoli|lunch is now known as zoli|wfh12:32
EmilienMpradk: it sounds like it does not work12:32
thrashd0ugal: I looked at the mistral executions, and the plan create was stuck in error, and it looked like a zaqar error.12:32
pradkEmilienM, yep and it works12:32
thrashd0ugal: I was trying to use a custom template dir.12:32
EmilienMpradk: ok good I hope I did a mistake in my templates12:32
pradkEmilienM, we even QA'ed and shipped in mitaka.. dont say it doesnt work :)12:32
EmilienMpradk: can you review the patch when you have time?12:33
jistrjaosorior, gfidente: +2'd the CloudName default change, it's only used from the tls-dns env file, and i'd assume if someone uses that, they're also feeding their own CloudName based on their DNS settings12:33
pradkEmilienM, hehe sure i'll take a look.. perhaps we have some regression from composable roles work12:33
thrashd0ugal: point being, I'm not sure it ever noticed the error.12:33
EmilienMpradk: https://review.openstack.org/#/c/362504/12:33
EmilienMpradk: specially https://review.openstack.org/#/c/362504/2/test-environments/scenario001-multinode.yaml12:33
shardymat128: Thanks for the info, I'll raise a bug as it seems like an issue with the testenv nic numbering, similar to a problem we fixed recently for the multinode job12:34
mat128cool :), didnt know what to do with it12:34
mat128thanks for your help12:34
shardymat128: I raised https://bugs.launchpad.net/tripleo/+bug/161844112:36
openstackLaunchpad bug 1618441 in tripleo "ovb job fails to find eth1 nic" [Undecided,New]12:36
jaosoriorjistr: :D12:36
d0ugalthrash: hrm, intersting. I am not sure how we handle that, if we never recieve messages on the zaqar queue, what do we do?12:37
thrashd0ugal: timeout perhaps?12:37
d0ugalthrash: I guess a sensible timeout per command could work. Plan create is quick, some things could be slow.12:37
d0ugalthrash: Yeah, okay, I guess this is an issue with all the merged commands too. Probably need something generic - I'll take a look.12:38
thrashd0ugal: I'll add a comment on the first patch for tracking.12:38
d0ugalthrash: Thanks12:38
*** jeckersb_gone is now known as jeckersb12:39
pradkEmilienM, looking at these logs http://logs.openstack.org/04/362504/2/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha-mitaka/4686f25/logs/overcloud-controller-0/var/log/messages12:40
shardyd0ugal: IME if we don't get the messages for any reason, we hang forever, so a timeout is definitely something we need to investigate12:40
pradkEmilienM, i dont see any aodh/gnocchi errors .. i do see some newutron/haproxy errors12:40
EmilienMpradk: wyh12:40
pradkEmilienM, if you see aodh/gnocchi logs the services come up fine and talk to redis12:40
EmilienMpradk: no...12:40
EmilienMwrong logs12:41
d0ugalshardy: +1, I'll look into it12:41
EmilienMpradk: you need to look at https://review.openstack.org/#/c/362505/12:41
pradkEmilienM, that the ha mitaka job failing12:41
EmilienMpradk: in scenario001 logs12:41
*** lucas-hungry is now known as lucasagomes12:41
pradk<EmilienM> pradk: specially https://review.openstack.org/#/c/362504/2/test-environments/scenario001-multinode.yaml12:41
pradkso looks at logs for that patch12:41
d0ugalshardy: I'm not sure it should block things landing, as it is an issue with the merged CLI workflow calls12:41
EmilienMyes because the scenario001 is only run when we modify telemetry files12:42
shardyd0ugal: Yeah, but we should probably raise a bug to track fixing it (for all commands) before the final release12:42
d0ugalshardy: Yeah, good point. I'll open a bug now.12:42
EmilienMpradk: one sec... i'll explain you how it works12:42
EmilienMpradk: logs are here anysay http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/12:42
shardyd0ugal: please target it at newton-rc112:42
d0ugalshardy: will do12:43
EmilienMpradk: https://review.openstack.org/#/c/361433/ please look what im doing in CI12:43
EmilienMpradk: specially https://review.openstack.org/#/c/361433/1/zuul/layout.yaml12:43
EmilienMpradk: now look https://review.openstack.org/#/c/362505/ and why I did this12:44
EmilienMpradk: you got it?12:44
pradkEmilienM, so looking at http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/logs/postci.txt.gz12:46
pradkEmilienM, where are the overcloud logs? the stack created successfully.. but dint gather any logs12:46
EmilienMpradk: pingtest failed12:47
d0ugalshardy, thrash: FYI https://bugs.launchpad.net/tripleo/+bug/161844512:47
openstackLaunchpad bug 1618445 in tripleo "Add a timeout in tripleoclient when waiting for Zaqar messages" [High,Confirmed] - Assigned to Dougal Matthews (d0ugal)12:47
jaosoriormcornea, gfidente: is the keystone admin port supposed to be exposed in the external network?12:48
*** fzdarsky|lunch is now known as fzdarsky|afk12:48
pradkEmilienM, right, just trying to understand why you think gnocchi is not working without any log suggesting so12:48
EmilienMpradk: http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/console.html#_2016-08-30_03_33_00_84084012:49
EmilienMpradk: the gnocchi resource alarm is not created12:49
gfidentejaosorior, external12:49
jaosoriorgfidente: is it?12:49
pradkEmilienM, k any way to gather logs.. hard to say whats happening without seeing the error12:49
gfidenteAFAIK v2 client defaults to the admin endpoint12:49
jaosoriorgfidente: or is it supposed to be in the internal?12:49
dtantsurfolks, how do you test stuff locally? how do you fight with constant inconsistency between THT (checkout) and puppet on the image?12:50
EmilienMpradk: well the redis error in gnocchi logs looks quite problematic12:50
gfidentedtantsur, if you need to pull updates in THT, the only way is to rebuild the overcloud-image too12:50
EmilienMpradk: did you have it in our local tests?12:50
dprinced0ugal: going to try it again today, rebuilding everything instead of just tripleo-common and python-tripleoclient12:50
gfidentedtantsur, or at least that is what I do12:50
jaosoriordtantsur: I recently started using this https://hardysteven.blogspot.fi/2016/08/tripleo-deploy-artifacts-and-puppet.html12:51
gfidentedtantsur, if you need a particular puppet change, then DeployArtifacts12:51
d0ugaldprince: Hey, no need to do that yet - I can confirm your problem :)12:51
dprinced0ugal: oh, you found it then?12:51
thrashd0ugal: I think it's just zaqar being "broken" at that moment.12:51
d0ugalthrash: agreed12:51
*** fzdarsky|afk is now known as fzdarsky12:52
d0ugaldprince: I don't know what the problem is yet - but I have these two test cases, interestingly one works and one doesn't: https://github.com/d0ugal/tripleo-util/tree/master/test-scenarios/network_isolation12:52
*** vijaykc4 has quit IRC12:52
pradkEmilienM, where do you see redis error in this review https://review.openstack.org/#/c/362505/12:52
thrashd0ugal: but definitely still a problem. :)12:52
dtantsurmeh, that's complex.... anyway, thanks jaosorior12:52
d0ugaldprince: (well, the one that works complained about a missing param, but I think that is because I don't have your custom.yaml file)12:52
d0ugalthrash: +112:52
jaosoriordtantsur: well, it's a long read but I found it helps out a lot in making developing faster12:52
shardydtantsur: it's really not complex, I just included a lot of context12:52
EmilienMpradk: http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/logs/subnode-2/var/log/gnocchi/app.txt.gz12:53
shardydtantsur: tl;dr tripleo-common/scripts/upload-puppet-modules -d <direcory of modules>12:53
d0ugaldprince: given it changed depending how the paths are passed in, it must be something to do with the merging code - I'll need to dive in and see.12:53
d0ugaldprince: trying to do that now12:53
EmilienMpradk: just read scenario001 logs, like I said.12:53
EmilienMpradk: scenario001 == telemetry scenario12:54
dprinced0ugal: ack, carry on then. Let me know if you need help12:54
dprinced0ugal: my custom.yaml file is just some 'parameter_defaults:'12:54
d0ugaldprince: Can you send me them? Just so I can confirm that it fully works eventually12:55
pradkEmilienM, so its not just gnocchi.. see aodh and ceilometer they cant talk to redis either12:55
*** dtrainor has joined #tripleo12:55
pradkEmilienM, so issue perhaps is that redis is not up12:55
pradkthe config for redis looks fine to me12:56
pradkin gnocchi.conf12:56
dtantsurshardy, okie, lemme try12:56
pradkEmilienM, see http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/logs/subnode-2/var/log/aodh/evaluator.txt.gz12:56
dprinced0ugal: sent you via email12:56
d0ugaldprince: Thanks12:56
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable TLS for keystone admin endpoint.  https://review.openstack.org/36285312:57
pradkEmilienM, similarly ceilo.. http://logs.openstack.org/05/362505/2/experimental/gate-tripleo-ci-centos-7-scenario001-multinode-nv/ad71d22/logs/subnode-2/var/log/ceilometer/central.txt.gz12:57
openstackgerritEmilien Macchi proposed openstack-infra/tripleo-ci: Implement scenari001, 002 and 003  https://review.openstack.org/36250412:57
pradkEmilienM, so its not just gnocchi.. redis is not up and hence all services are failing12:57
shadowermandre: I replied to: https://review.openstack.org/#/c/354775/  would appreciate your thoughs when you have a few12:57
jaosoriorshardy: by the way, great blog post. That's been really helpful12:57
shadowermandre: it's the openstack-config question with johnbro's patches12:58
EmilienMpradk: ok I'll try to investigate more today though i'm not sure i can12:58
pradkEmilienM, i guess check why redis is not up first12:58
pradkcool, i'll take a look this afternoon as well12:58
shardyjaosorior: thanks :)12:58
*** tzumainn has joined #tripleo12:58
*** vijaykc4 has joined #tripleo12:59
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Mv Nova, Neutron, Horizon out of controller.yaml  https://review.openstack.org/36136212:59
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Mv pacemaker and firewall out of controller  https://review.openstack.org/36139812:59
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Move snmp settings into composable services  https://review.openstack.org/36006212:59
*** ebarrera has quit IRC12:59
*** jpena|lunch is now known as jpena12:59
EmilienMpradk: yes lets sync a bit later on this13:00
gfidentejaosorior, I could very well add ceph/rgw specific data in the endpoint map though instead of reusing swift proxy data13:00
gfidenteif we don't want to share tcp ports13:00
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: configure heat with new authtoken parameters  https://review.openstack.org/36140913:03
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: configure Heat with new authtoken parameters  https://review.openstack.org/36141213:03
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: Add NetApp Manila driver integration and tidy up generic  https://review.openstack.org/35401913:03
jaosoriorgfidente: well, however you see fit. However, if ceph rgw is a replacement for swift. I think making them mutually exclusive would help avoiding issues.13:06
EmilienMshardy: https://review.openstack.org/36285813:07
mandreshadower: looking at your config parser module13:07
openstackgerritAthlan-Guyot sofer proposed openstack/tripleo-heat-templates: Refactor upgrade checks.  https://review.openstack.org/35775013:07
shadowermandre: thanks. I've tested it briefly but it's mostly an illustration of what we could do there13:07
thrashd0ugal: you have any docs/pointers/is it actually ready for 'overcloud plan deploy'?13:08
dtantsurshardy, the same failure... should I specify something explicitly for the uploaded stuff to be used?13:08
thrashd0ugal: got this: http://paste.openstack.org/show/564827/13:08
mandreshadower: looking good to me13:09
*** rodrigods has quit IRC13:09
shardydtantsur: No, but you have to be careful of naming, e.g puppet_modules/ironic not puppet_modules/puppet-ironic when cloning13:09
*** rodrigods has joined #tripleo13:09
d0ugalthrash: is what ready?13:09
shardydtantsur: could that be the problem?13:09
thrashd0ugal: 'overcloud plan deploy'13:09
dtantsurshardy, aha, that's what I missed13:09
*** david-lyle has quit IRC13:10
d0ugalthrash: well, you could deploy normally, then re-deploy it as a plan (to update or delete the stack first for a create)13:10
mandreshadower: I don't think it needs to be any more complex13:10
shardyEmilienM: thanks!13:10
d0ugalthrash: Parameters need to be set manually if you just create an empty plan - there is a patch to improve some of that. https://review.openstack.org/#/c/361440/13:10
*** pgadiya has quit IRC13:10
thrashd0ugal: that's what I'm missing I think. :)13:11
derekhlucasagomes: the problem with the pyghmi timeouts is something to do with its session handling (and possibly not getting rid of old sessions of of its cache)13:11
thrashd0ugal: looks like it's just those three, so I'll just make an env file with them.13:11
d0ugalthrash: Yeah, based on your error :) but you might be missing passwords etc. too?13:11
d0ugalthrash: cool13:12
thrashd0ugal: already fixed that with a yaml. :)13:12
thrashthe passwords.13:12
derekhlucasagomes: it looks like it happens any time the source UDP port that was used is one that had been used before (to talk to the same bmc)13:12
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Add NeutronGlobalPhysnetMtu to neutron-base.yaml  https://review.openstack.org/36153913:12
lucasagomesderekh, yeah, I didn't dig much into it to see. But have you seem it reproduce with ovb ?13:12
d0ugalthrash: FWIW, I done a regular deploy and saved the params so I could quickly do a deploy: https://github.com/d0ugal/tripleo-util/blob/master/mistral-params.json13:12
lucasagomesderekh, a-ha nice finding13:12
derekhlucasagomes: yup, reproducing it here, looking through the code now13:13
thrashd0ugal: ack.13:13
thrashd0ugal: *appears* to be working! :)13:13
thrashjust got all the CREATE_IN_PROGRESS messages! \o/13:14
*** vijaykc4 has quit IRC13:14
rbradyshardy: the jinja template needs to be processed only as jinja or processed first as jinja and then as yaml?13:14
*** egafford has joined #tripleo13:15
thrashd0ugal: stack create failed, but at least the mistral stuff worked. :)13:16
d0ugalthrash: \o/13:16
openstackgerritmathieu bultel proposed openstack-infra/tripleo-ci: WIP - Implement overcloud upgrade job - Mitaka -> Newton  https://review.openstack.org/32375013:16
mandreEmilienM: earlier with honza we noticed that the apache puppet module was purging the /etc/httpd/conf.d directory13:17
EmilienMmandre: yes13:17
mandreEmilienM: which is annoying since that where the tripleo-ui package puts its configuration :/13:17
EmilienMi said it yesterday to jaosorior13:18
EmilienMmandre: and like I said its normal13:18
jaosoriorEmilienM: I actually said that to honza too13:18
EmilienMpuppet manages config that we told it to manage13:18
EmilienMyou need to manage ui vhost with puppet13:18
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902713:18
EmilienMlike we do with horizon etc13:18
*** david-lyle has joined #tripleo13:18
EmilienMits puppet 10113:18
gfidentejaosorior, can you check https://review.openstack.org/289027 especially the part with the endpoints and see if I missed anything?13:18
EmilienMpuppet ensure manual actions are purged13:19
dprincemandre: is tripleo-ui getting installed with puppet?13:19
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902713:19
honzaEmilienM: that's what we figured but wanted to double check13:19
mandreEmilienM: ok, I see... can we ship that configuration file with puppet-tripleo or should it be it's own puppet module?13:20
EmilienMit could be in puppet tripleo13:20
EmilienMtripleo::ui for ex13:20
mandredprince: https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/ui.pp#L2013:20
EmilienMin theory it should not be in a profile13:20
EmilienMlogic should not be in the profile13:21
EmilienMbut in a regular class13:21
EmilienMand the class should be called in a profile13:21
EmilienMlook at how we do other profiles13:21
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902713:21
jaosoriorgfidente: will we need a keystone endpoint for ceph rgw?13:21
*** cdearborn has joined #tripleo13:22
gfidentejaosorior, it does need one as it's public facing service replacing swiftproxy13:22
jaosoriorgfidente: where is that created?13:22
gfidentejaosorior, but it should be of object-store type13:22
mandreEmilienM: perfect, will give it a try, thanks13:22
openstackgerritMerged openstack/tripleo-specs: Specification for tripleo-lldp-validation blueprint  https://review.openstack.org/32920313:23
Jokke_shardy: https://blueprints.launchpad.net/tripleo/+spec/manila-cephfs-integration13:23
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: Add NetApp Manila driver integration and tidy up generic  https://review.openstack.org/35401913:23
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: IGNORE - testing manila-generic deploy after tidyup  https://review.openstack.org/36286613:23
gfidentejaosorior, either we use swift::keystone::auth13:24
*** mhenkel has joined #tripleo13:25
gfidenteor I'll have to copy that into puppet-ceph13:25
*** lblanchard has joined #tripleo13:25
gfidentejaosorior, and I think I'll have to copy it to accomodate a slightly different endpoint url13:26
gfidente$hit :)13:26
EmilienMrbrady, dprince: do you guys know when mistral will switch to modern keystone authtoken parameters? https://github.com/openstack/mistral/blob/c2424d6ccedc6d8a4448d78612a82521472443e1/mistral/utils/openstack/keystone.py13:26
dprinceEmilienM: when you push a patch for it :)13:27
rbradyEmillienM: I don't know of any planned efforts to change that currently in mistral.13:27
EmilienMrbrady, dprince: in mistral.conf it's still old keystone params13:27
dprinceEmilienM: not sure what the plans are for that...13:27
EmilienMdprince: ok asking because it's one of the last projects in OpenStack that does it13:28
rbradyEmillienM: maybe we can get that into the planning for O-113:28
EmilienMdprince: context https://bugs.launchpad.net/puppet-aodh/+bug/1604463/13:28
openstackLaunchpad bug 1604463 in puppet-swift "Modules should use keystone::resource::authtoken to configure keystone_authtoken settings" [Medium,Fix committed] - Assigned to Iury Gregory Melo Ferreira (iurygregory)13:28
dprincerbrady: no need to plan it I think, just do it13:28
d0ugalEmilienM: I think there is a bug for it somewhere, but I don't know of anyone actually working on it.13:28
EmilienMd0ugal: ack13:28
jaosoriorgfidente: well, you need to set up the keystone::auth stuff in that case13:31
*** pkovar has joined #tripleo13:31
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: POC: WIP: oooq undercloud install  https://review.openstack.org/35891913:32
gfidentejaosorior, yeah13:32
lucasagomesderekh, yeah if I send a lot of ipmi commands quickly I can reproduce the problem http://paste.openstack.org/show/564834/13:32
gfidenteworking on it right now13:32
lucasagomesderekh, (that's running with virtualbmc)13:32
derekhlucasagomes: try this first to limit the source port allocation13:33
derekhlucasagomes: echo "32768   32777" | sudo tee /proc/sys/net/ipv4/ip_local_port_range13:33
derekhlucasagomes: you'll reproduce it a lot faster13:33
lucasagomesderekh, ty, will do13:33
*** jeckersb is now known as jeckersb_gone13:34
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902713:35
*** jeckersb_gone is now known as jeckersb13:37
rookbeagles curious, why do we ship with two "default"13:38
beaglesrook: in overcloud?13:38
beaglesrook, or undercloud? but to answer your question - I don't know off hand but now I'm curious too :)13:39
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: POC: WIP: oooq undercloud install  https://review.openstack.org/35891913:39
rookyup overcloud13:39
rbradydprince: guess it depends if you're trying to work in an organized sprint13:40
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Migrate to using osc-lib  https://review.openstack.org/33546013:40
rookdprince if you have a newton cloud, do you have two Default sec-groups?13:40
dprincerbrady: I always "sprint" man :)13:41
dprincerbrady: spinning up one now, I will look in a bit13:41
openstackgerritAttila Darazs proposed openstack/tripleo-quickstart: Use relative paths in ci-scripts for quickstart.sh  https://review.openstack.org/36064913:41
d0ugaldprince: daily sprints! That sounds super agile!13:42
openstackgerritAttila Darazs proposed openstack/tripleo-quickstart: Use relative paths in ci-scripts for quickstart.sh  https://review.openstack.org/36064913:42
openstackgerritAttila Darazs proposed openstack/tripleo-quickstart: Document Third Party CI and developer usage  https://review.openstack.org/36000713:44
*** fultonj has quit IRC13:44
*** fultonj has joined #tripleo13:46
*** mcornea has quit IRC13:47
openstackgerritJohn Browning proposed openstack/tripleo-validations: Added check for ceilometerdb size  https://review.openstack.org/35477513:47
openstackgerritJohn Browning proposed openstack/tripleo-validations: Added check for ceilometerdb size  https://review.openstack.org/35477513:48
*** ayoung has joined #tripleo13:52
*** liverpooler has quit IRC13:53
openstackgerritRyan Brady proposed openstack/tripleo-common: [WIP] Wire in jinja templating for custom roles  https://review.openstack.org/36246513:54
*** ebarrera has joined #tripleo13:55
d0ugalyes, We should all go there :)13:55
shardyGood reminder dtantsur ;)13:55
shardymeeting is in 5mins13:55
*** yamahata has joined #tripleo13:57
*** myoung|afk is now known as myoung14:00
*** jprovazn has quit IRC14:01
*** pkovar has quit IRC14:01
openstackgerritJohn Browning proposed openstack/tripleo-validations: Added check for ceilometerdb size  https://review.openstack.org/35477514:02
*** mcornea has joined #tripleo14:02
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902714:04
*** jcoufal has joined #tripleo14:06
openstackgerritJohn Browning proposed openstack/tripleo-validations: Added check for ceilometerdb size  https://review.openstack.org/35477514:07
openstackgerritBabu Shanmugam proposed openstack/tripleo-heat-templates: OVN heat templates  https://review.openstack.org/30773414:08
*** jprovazn has joined #tripleo14:09
jaosoriorgfidente: why not just reuse swift::keystone::auth?14:10
gfidentejaosorior, I think we could yes14:11
gfidenteI am not sure what level of separation we should have amongst the two14:11
gfidenteand where to draw the line14:11
larsksEmilienM: you mentioned that puppet/{compute,controller}.yaml are going away.  With what are they being replaced?14:11
EmilienMlarsks: tripleo meeting right now14:12
larsksEmilienM: okay.  Whenever you have time...14:12
*** chem has quit IRC14:13
*** chem has joined #tripleo14:13
*** ebalduf has joined #tripleo14:14
openstackgerritJeff Peeler proposed openstack/tripleo-common: Add List Roles Action  https://review.openstack.org/33028314:14
gfidentejaosorior, but swift::keystone::auth is pulling in some swift::deps14:14
jaosoriorgfidente: alright, then it's fine to have the specific one14:15
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: [NO MERGE] Test Ceph RadosGW as replacement for Swift  https://review.openstack.org/35718214:16
gfidentethough it currently depends on two different puppet-tripleo changes I should stack up14:18
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: Add NetApp Manila driver integration and tidy up generic  https://review.openstack.org/35401914:18
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Enable TLS in the internal network for keystone  https://review.openstack.org/32702914:19
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Add Ceph RGW listener to HAProxy  https://review.openstack.org/36281614:19
openstackgerritGiulio Fidente proposed openstack/puppet-tripleo: Add Ceph RGW profile  https://review.openstack.org/33408114:21
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902714:22
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: Split additional features across the periodic jobs  https://review.openstack.org/36290414:23
*** vijaykc4 has joined #tripleo14:24
*** tosky has quit IRC14:25
beaglesdprince, I owe you a beer for this MTU thing - I also need to adapt my MO to vary my what network environment's I deploy with14:25
larsksdprince: in https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/controller.yaml#L394, shouldn't that be bootstrap_nodeid, rather than bootstack_nodeid?14:26
*** tosky has joined #tripleo14:27
*** jprovazn has quit IRC14:27
dprincebeagles: thanks for having a look and digging into it14:28
*** jprovazn has joined #tripleo14:28
*** vijaykc4 has quit IRC14:28
*** dtantsur is now known as dtantsur|mtg14:28
dprincelarsks: bootstack is the right one14:28
larsksdprince: huh, okay.  because there also exists bootstrap_ in other places, so I was confused.14:28
dprincelarsks: a legacy thing... we might be weening off of it. I haven't checked. but bootstack was a thing at one point14:28
larsksdprince: as long as you're here, can you point me at any info (reviews?) about what will be happening to puppet/{controller,compute}.yaml? I hear they are going away...14:29
*** ppradhan has joined #tripleo14:33
dprincelarsks: rbrady and shardy are working on it. Look for there incoming patches regarding jinja templating within the Mistral workflows14:33
openstackgerritDan Prince proposed openstack/tripleo-common: Fix upload-swift-artifacts work new OSclient  https://review.openstack.org/36291514:35
jaosoriorEmilienM: where is the keystone URL set for the provider that does the endpoints and all that stuff?14:35
EmilienMits tripleo meeting now FYI14:36
rbradylarsks: writing a test for this now to remove the WIP flag https://review.openstack.org/#/c/362465/14:36
adarazspabelanger: I don't see that we break anything there by the 'rdo-ci-check' behavior. -- we're not triggering by default so we don't need recheck support either for now IMO.14:36
adarazspabelanger: of course if we would start to trigger on all patchsets uploaded we'd add recheck and remove this extra codeword.14:37
sshnaidmEmilienM, can you please to update the bug with patch number, in your time? I'll close it then: https://bugs.launchpad.net/tripleo/+bug/161841814:38
openstackLaunchpad bug 1618418 in tripleo "CI: periodic jobs fail on undercloud install because "AttributeError: 'module' object has no attribute 'RegionInvalidationStrategy'"" [High,Triaged]14:38
pabelangeradarazs: The reason is, we don't want to run per 3rd party CI on patches in review.openstack.org.  Other wise, people can just abuse the command and keep rechecking a broken 3rdparty CI system until it is green14:39
*** jprovazn has quit IRC14:39
pabelangeras such, we only support recheck systax14:40
*** cdearborn has quit IRC14:40
pabelangerand this is outlined in the requirements: http://docs.openstack.org/infra/system-config/third_party.html#requirements14:40
openstackgerritMerged openstack-infra/tripleo-ci: Replace ssh public key for sshnaidm  https://review.openstack.org/36281414:41
*** cdearborn has joined #tripleo14:41
adarazspabelanger: I understand the reasoning, but we're not even voting on anything for now. you can think of it as a demo mode. would love to have some jobs that will actually vote and then we can conform to the standard.14:41
pabelangeradarazs: unfortunately this is how the process works today. And how openstack-infra requires 3rd party CI system to work.14:42
pabelangerIf you think the policy should change, I encourage to raise a topic in #openstack-infra IRC or mailing list14:43
*** david-lyle has quit IRC14:43
adarazspabelanger: I don't see how this break anything by having an undocumented magic word that some developer can use to test their change against our latest RDO images/testing images. it doesn't trigger normally, it doesn't vote, it's only enabled for a few selected repos.14:44
adarazspabelanger: I can't enable it on any repo normally because it will crush our CI system if it runs on all changes for now.14:44
*** jaosorior is now known as jaosorior_away14:45
*** david-lyle has joined #tripleo14:48
adarazspabelanger: I showed that we have this kind of functionality on the meeting just so that people can actually use the method, with maybe more resources than we have on centos ci currently :)14:48
pabelangeradarazs: see https://review.openstack.org/#/c/329049/14:48
*** aufi has quit IRC14:49
pabelangerFirst of all third party requirements at the items _every_ third party system has to have in place in order to keep their gerrit account. Failure to implement _all_ requirements risks having your account shut off. Requirements are stable at this point and are not changed lightly. Since I'm the one that does most of the decision making regardling toggling of system accounts and following up with account14:49
pabelangerholders, I consider it a courtesy to be consulted should something actually need to be changed in third party requirements.14:49
pabelangerbasically, you run the risk of having your account shut off. I don't know your CI specific issues, and not able to help fix that. However, I encourage you to use recheck systax14:50
pabelangerand, I encourage you to raise your objections in #openstack-infra. Specifically, anteaya and fungi would have more to say on the subject14:52
openstackgerritOpenStack Proposal Bot proposed openstack/os-collect-config: Updated from global requirements  https://review.openstack.org/35090514:52
openstackgerritMerged openstack/tripleo-validations: Added check for ceilometerdb size  https://review.openstack.org/35477514:52
adarazspabelanger: I really don't see your point. what are you trying to prevent here? yes. it doesn't work like that. but this functionality is in demo mode. we have proper gate jobs on this account for tripleo-quickstart that adheres to the rules completely. otherwise we have a few repos that this magic command works and it does a comment about the results. if I make the jobs silent would that help?14:52
adarazsnobody minds that on those repos that we have this.14:53
fungiyeah, we're generally flexible on ci systems dedicated to specific projects where the team maintaining the ci is the team maintaining the project being tested/commented on14:53
openstackgerritOpenStack Proposal Bot proposed openstack/python-tripleoclient: Updated from global requirements  https://review.openstack.org/36187514:53
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add support for deploying Ceph RGW role  https://review.openstack.org/28902714:54
adarazsthat's pretty much what's happening now /o\14:54
openstackgerritOpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements  https://review.openstack.org/36150114:54
fungiwhat it comes down to is that if people in charge of the project whose changes are receiving these ci comments object to them, then we convince the ci operator to stop doing that (or we disable their account if we can't track them down quickly enough)14:55
*** bana_k has joined #tripleo14:55
fungithe only other typical case is that if the ci system's comment volume is significant enough to impact performance of our gerrit server (mostly only happens when a system begins looping and responding to its own comments), then we tend to shoot first and ask questions afterward14:56
adarazsyeah, that won't be a problem with the tripleo repos we're testing this CI, to eventually have proper upgrade jobs that the regular infra can't do.14:56
saneaxhttps://review.openstack.org/#/c/361238/ failed on merge  (After workflow) - is this something which we need to address?14:56
pradkEmilienM, i think its an issue with redis.. seems like the haproxy config is getting the ping but not ponging back for redis and hence redis is not reachable14:57
adarazsfungi: this seems all reasonable. :)14:57
pabelangerfungi: okay, that is a little different from what anteaya had explained before. There was objections to having pre CI syntax regardless of project14:57
EmilienMpradk: tripleo meeting now14:57
fungieffectively, while the current third-party ci documentation doesn't really get explicit about it, there are two fairly distinct classes of third-party ci... those which are focused on testing a specific project in ways our upstream ci cannot but is run by that project, and those focused on testing integration with some other project or projects who are not in control of the ci14:58
*** rajinir has joined #tripleo14:58
fungifor ci systems operated by a project and only testing repos for that same project, the infra team tends not to get involved in its operation at all14:59
mariosshardy: fyi filed https://bugs.launchpad.net/tripleo/+bug/1618479 for manila and added the closes-bug to the tht https://review.openstack.org/#/c/35401914:59
openstackLaunchpad bug 1618479 in tripleo "Add integration for manila-netapp backend" [Undecided,In progress] - Assigned to Marios Andreou (marios-b)14:59
shardymarios: ack, thanks!14:59
fungihowever, those are a very small fraction of the overall number of third-party ci systems reporting into our gerrit, so the documentation doesn't really cover them directly14:59
saneaxshardy, https://review.openstack.org/#/c/361238/ failed on merge - I havent seen this, any idea why?15:00
pandaEmilienM: so you're only creating the jobs definitions for the periodic ha-ipv6 jobs ?15:00
pabelangerfungi: okay, great. So I stand corrected. adarazs you shouldn't have an issue then15:00
*** vijaykc4 has joined #tripleo15:00
openstackgerritJuan Antonio Osorio Robles proposed openstack/puppet-tripleo: Enable TLS in the internal network for keystone  https://review.openstack.org/32702915:00
fungiwe're mostly focused on the third-party ci systems operated by hardware vendors doing things like test neutron/cinder/nova/ironic drivers and report on changes to those projects15:00
adarazsfungi, pabelanger: so yeah, these jobs are by tripleo people, for tripleo projects and for now very special cases (when we're trying to fix some bug in the rdo promotion and we want to make sure the specific patch helps), so I think we're okay.15:00
shardysaneax: it depends on https://review.openstack.org/#/c/327705/ (in a series), which failed the gate CI15:01
saneaxoh ok shardy, thanks15:01
shardyIf you think that failure was spurious, you can use "reverify" to re-run the gate verify jobs15:01
adarazsyeah, we're not trying to do that. :)15:01
EmilienMpanda: yes in project config15:01
EmilienMpradk: can you file a bug please? i think its critical15:02
pabelangerfungi: we should likely revisit https://review.openstack.org/#/c/329049/ with updates about per-project CI systems. That is where my confusion came from15:02
*** david-lyle has quit IRC15:02
fungiadarazs: _if_ you later decide that the scenarios being tested should also validate and comment on, say, proposed ironic changes, then we'll need to revisit15:02
*** mcornea has quit IRC15:02
adarazsfungi: ack, understood. thanks for the clarification :)15:03
pandaEmilienM: you talking to me regarding the bug ?15:03
fungipabelanger: it's more that almost none of the documentation in that file relates to those rare third-party ci systems which are operated by the projects tey're commenting on15:03
EmilienMpanda: no pradk about redis bug15:04
sshnaidmtripleo cores, please consider to merge this tempest patches: https://review.openstack.org/#/c/355641/ https://review.openstack.org/#/c/361936/15:04
*** rhallisey has quit IRC15:05
*** tremble has quit IRC15:05
*** HoloIRCUser9 has quit IRC15:05
*** HoloIRCUser has joined #tripleo15:05
*** HoloIRCUser2 has joined #tripleo15:06
*** HoloIRCUser5 has joined #tripleo15:07
*** HoloIRCUser2 has quit IRC15:07
*** coolsvap is now known as coolsvap_15:07
*** chem has quit IRC15:08
*** panda is now known as panda|bbl15:08
openstackgerritMerged openstack/puppet-pacemaker: Fix CI puppet-lint and type issues  https://review.openstack.org/35520215:08
*** chem has joined #tripleo15:08
*** HoloIRCUser5 has quit IRC15:09
*** HoloIRCUser has quit IRC15:10
*** numans has quit IRC15:12
pradkEmilienM, https://bugs.launchpad.net/tripleo/+bug/161851015:12
openstackLaunchpad bug 1618510 in tripleo "unable to reach redis service" [Undecided,New]15:12
*** bana_k has quit IRC15:13
openstackgerritMerged openstack/puppet-pacemaker: Move other-requirements.txt to bindep.txt  https://review.openstack.org/35500115:13
openstackgerritBrent Eagles proposed openstack/tripleo-heat-templates: Add support for configuring the OVS firewall driver  https://review.openstack.org/35755615:14
*** saneax is now known as saneax-_-|AFK15:16
*** abregman is now known as abregman|mtg15:17
*** zoli|wfh is now known as zoli|afk15:19
*** ohamada has quit IRC15:20
openstackgerritBrent Eagles proposed openstack/tripleo-heat-templates: Add base neutron service configuration  https://review.openstack.org/36136715:20
*** abregman|mtg has quit IRC15:22
*** david-lyle has joined #tripleo15:22
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Update configuration docs  https://review.openstack.org/35670015:22
*** dtantsur|mtg is now known as dtantsur15:25
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add `overcloud parameters set` to set Heat params in a plan  https://review.openstack.org/36039115:25
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add 'openstack overcloud plan create' command  https://review.openstack.org/35978415:25
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Print the execution ID when starting Mistral workflows  https://review.openstack.org/34071115:25
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add `openstack overcloud plan deploy`  https://review.openstack.org/36030515:25
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Update the 'overcloud deploy --templates' command to use Mistral  https://review.openstack.org/34226115:25
dtantsurEmilienM, re https://review.openstack.org/#/c/362504/ is it possible to add ironic to one of these?15:26
d0ugaldprince: ^ I think that should solve your issue, if you could give it another test that would be great.15:26
dtantsureven without fancy checks, just make sure it still can be installed15:26
*** leanderthal is now known as leanderthal|afk15:26
EmilienMdtantsur: thats the goal Sir15:26
EmilienMdtantsur: but let me finish this baby step15:26
EmilienMdtantsur: and I15:27
dtantsuryeah, cool, thanks! just making sure I understand the plan15:27
EmilienMerr this new keyboard15:27
EmilienMand I'll add it to scenario002 like in puppet ci15:27
dtantsurdprince, I've finally managed to build a local environment to test your ironic patch..15:27
*** links has quit IRC15:28
EmilienMbnemec: what is periodic-ovb-ha2 ?15:30
EmilienMbnemec: random question feel free to say idk15:30
bnemecEmilienM: I believe that's the version of the ha job that we're running right now.15:31
bnemecIt was added when derekh started working on ovb support in tripleo-ci.15:31
EmilienMbnemec: ok merci15:31
*** bana_k has joined #tripleo15:32
*** xinwu has joined #tripleo15:34
openstackgerritMerged openstack/puppet-tripleo: Defaults to ceph_mon_node_ips and ceph_mon_node_names  https://review.openstack.org/36110815:34
*** jcoufal has quit IRC15:34
derekhbnemec: if tou see random ipmitool calls to the ovb BMC timeout in ironic, I belive this fixes them https://review.openstack.org/#/c/362944/115:34
derekhbnemec: I spent some time tracking it down today and it turns out if the same source udp port is used twice things didn't work15:35
*** xinwu has quit IRC15:36
openstackgerritMerged openstack/tripleo-quickstart: Revert "Move unique socket generation into quickstart.sh"  https://review.openstack.org/36229215:36
*** jcoufal has joined #tripleo15:36
EmilienMbnemec, dprince, panda|bbl, shardy: can you guys review please https://review.openstack.org/36296615:37
derekhEmilienM: I added ha2 as the ha job on ovb was different to the origional ha job , we shouldn't need it any longer, is it still being used?15:39
EmilienMderekh: i have not idea15:39
EmilienMderekh: i havent seen this job running lately15:39
bnemecYes, ha2 is still the ha job we're running.15:39
derekhEmilienM: bnemec ok we should probably make the ha and ha2 cases in the switch statment the same15:40
*** david-lyle has quit IRC15:40
derekhEmilienM: bnemec then change the JOB_TYPE in project-config to ha and finally delete the ha2 entry from the toci_gate_test switch statment15:41
EmilienMwho's on this?15:42
*** dtantsur is now known as dtantsur|afk15:42
derekhEmilienM: I'll push up the first patch now and will come back to it later15:42
openstackgerritAttila Darazs proposed openstack/tripleo-quickstart: Use relative paths in ci-scripts for quickstart.sh  https://review.openstack.org/36064915:42
openstackgerritAttila Darazs proposed openstack/tripleo-quickstart: Document Third Party CI and developer usage  https://review.openstack.org/36000715:44
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Update section header indicators  https://review.openstack.org/36052215:44
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Update configuration docs  https://review.openstack.org/35670015:44
*** ebarrera has quit IRC15:46
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Sync the config of the ha JOBTYPE with ha2  https://review.openstack.org/36297415:46
derekhEmilienM: bnemec ^ patch one, gotta run will do the project-config patch this evening15:46
*** derekh has quit IRC15:46
bnemecgfidente: Now that we're testing ceph in the nonha job again, do we still need it in the ha job too?15:48
bnemecIt would be nice to drop it if we can.  We'd need one less node and the ha job would probably be a little faster.15:49
*** david-lyle has joined #tripleo15:50
gfidentebnemec, we can drop it yes15:50
gfidentethough I wanted to turn it into osd on compute scenario15:50
gfidentewhich would still permit us to remove the additional node15:50
gfidenteuntil then, if we need resources, I think it's fine testing it in a single job15:51
bnemecgfidente: Yeah, we can certainly do that too.15:51
ayoungquickstart seems to be setting an obsolete value15:51
ayoung$  openstack baremetal import --json instackenv.json15:51
ayoungInvalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]15:51
gfidentebnemec, problem with that was steps synchronization outside roles which we don't have yet15:51
bnemecIt would probably be good to test the non-ceph path anyway.  Right now we aren't.15:51
ayoungassuming there is a patch to work around that?15:51
gfidentebnemec, also yes15:52
*** trown is now known as trown|lunch15:52
*** pkovar has joined #tripleo15:52
ayoungrunning using devmode...is that the problem?  Running like this: ~/.quickstart/tripleo-quickstart/quickstart.sh  --release master --extra-vars @config/general_config/devmode.yml  ayoung-dell-t1700.test15:52
bnemecgfidente: Okay, I'll write a patch.15:52
openstackgerritimain proposed openstack/tripleo-common: Create new docker command hook.  https://review.openstack.org/31272315:53
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: Enable modern authtoken for Neutron and Glance  https://review.openstack.org/36298015:53
openstackgerritJeff Peeler proposed openstack/tripleo-common: Include environments in capabilities output  https://review.openstack.org/35559815:53
*** pkovar has quit IRC15:54
jpeelerjtomasek: i imagine you don't have much day left, but if you do ^15:55
jtomasekjpeeler: cool, will test it tomorrow morning!15:55
jtomasekjpeeler: or you need it reviewed immediately?15:55
jpeelerjtomasek: if you have time to do it now, that would be great since i have ~6 hours left15:56
jtomasekjpeeler: ok, I'll check it in horizon of 2 hours15:57
jpeelerof 2 hours?15:57
jtomasekjpeeler: I am at the meeting now and then need to go afk for a bit, so I can do it right when I get back15:58
ayounghrybacki, what is the blessed configuration to run quickstart today?15:58
jtomasekjpeeler: is that fine?15:58
jpeelerjtomasek: yes, thanks!15:58
* jpeeler was thinking horizon UI...15:59
*** bana_k has quit IRC15:59
*** dprince has quit IRC15:59
*** dprince has joined #tripleo16:00
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Remove Ceph from the ha job  https://review.openstack.org/36299016:00
bnemecgfidente: ^16:01
*** ifarkas is now known as ifarkas_afk16:01
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add `overcloud parameters set` to set Heat params in a plan  https://review.openstack.org/36039116:02
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add 'openstack overcloud plan create' command  https://review.openstack.org/35978416:02
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Print the execution ID when starting Mistral workflows  https://review.openstack.org/34071116:02
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Add `openstack overcloud plan deploy`  https://review.openstack.org/36030516:02
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Update the 'overcloud deploy --templates' command to use Mistral  https://review.openstack.org/34226116:02
*** pkovar has joined #tripleo16:04
ayoungUm...guys?  Does anyone here use quickstart? Has anyone run it in, say, 2-3 days?  Cuz the last known good incantation I had no longer works, and I am trying to figure out if it is me or if it is Memorex.16:06
*** zoli|afk is now known as zoli|wfh16:09
*** pkovar has quit IRC16:09
*** pkovar has joined #tripleo16:10
*** bana_k has joined #tripleo16:14
openstackgerritJohn Browning proposed openstack/tripleo-validations: Check to ensure debug is disabled on undercloud services.  https://review.openstack.org/35555916:15
chemjistr: marios hi guys, could you give the final workflow to https://review.openstack.org/#/c/354713/, we still can see if gate-tripleo-ci-centos-7-ovb-hais  fixed during POST_FAILURE gating ?16:15
jristmandre, shadower ^16:15
shadowerjrist: what are you pointing at?16:16
shadowerjrist: as in, what should I look at?16:17
jristshadower: what hewbrocca said :)16:17
ayoungSo are we giving up on Tripleo Quickstart?16:17
shadoweroh, johnbro's patch?16:18
shadoweryea I'm on it16:18
shadowermerged two of his today16:18
shardyayoung: You may want to chat with trown|lunch and larsks re any quickstart issues16:18
shardyfolks are just super overloaded atm with newton3 stuff16:18
*** akrivoka has quit IRC16:18
shadowerhewbrocca, jrist: need to relocate but it's on my radar, don't worry :-)16:18
ayoungshardy, they the only ones that use it?16:18
shardyI can't help personally because I don't use it myself atm16:18
shardyayoung: No, but a lot of folks don't use it upstream because it's not currently what we run in CI16:19
shardyRDO CI does use it, as do quite a few developers16:19
ayoungshardy, yeah, I am aware, which is why I don't bother you.  I'll bug people in #rdo then16:19
shardyso, unfortuntely that dilutes the pool of folks in this channel equipped to help16:19
shardyayoung: I identifed trown|lunch and larsks because they were the original authors of it16:20
jristshardy, ayoung we just had a retrospective and we decided that we're going to centralize on one installer16:20
jristit will probably be tripleo quickstart16:20
ayoungshardy, yes, and I appreciate that.  It is the continuing difficulty of getting a working Tripleo development environment that leads me back here.16:21
*** david-lyle has quit IRC16:21
shardyayoung: Yeah, it's far from ideal16:21
ayoungjrist, it makes sense to me.  Ansible based provisioning is the ost important part.16:21
jristthat was one of the things we just discussed as well in our retrospective16:21
ayoungjrist, are you running it?16:21
shardytbh I'm perfectly happy with tripleo.sh myself16:21
jristayoung: I am not, but shadower is and he has some instructions16:21
ayoungshardy, I'm neutral.   I just want to use the tool that the team is using.  But it needs to work for the single host, resource constrained developer use case...cuz that is me.16:22
ansiwenchem: around?16:23
*** david-lyle has joined #tripleo16:23
chemansiwen: what's up ?16:23
ansiwenchem: hi. could you point me to the keystone tests that also used a provider to create credentials for tests? I think you mentioned something like that, no?16:24
marioschem: done16:25
shardyayoung: Yeah I am too, but it works for me, and gets me close to what we actually run in CI16:25
shardyif we switch to quickstart in CI, I will start using that instead16:25
chemmarios: thanks16:25
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Upgrade ceph-osd  https://review.openstack.org/35731216:26
chemansiwen: hum, it's for the ec2 tempest credential api creator... I don't recall something similar in keystone test.  Everything is hardcoded, not dinamically generated.  Your problem is not like the openrc stuff.16:27
openstackgerritMathieu Mitchell proposed openstack/diskimage-builder: (WIP) Zuul-cloner first attempt  https://review.openstack.org/36301316:27
*** dmacpher has quit IRC16:28
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add Ceph cluster health validation on upgrade  https://review.openstack.org/35941016:28
pradkshardy, Hey did we recently make any changes to how redis is configured in OSP10? we're seeing some issues16:29
ansiwenchem: but sketched an idea about a provider that created a file with the credentials, I thought you mentioned something similar would be in keystone...16:29
chemansiwen: this is the openrc stuff16:29
ansiwenchem: so, whats the openrc stuff?16:32
shardypradk: yes, possibly, we reworked how the redis_node_ips hieradata is created16:32
chemansiwen: hold no16:32
shardypradk: we now auto-generate a list in hiera for all services, when before it was a yaml string16:32
shardygfidente and jistr have been helping with some adjustments in puppet-tripleo to align with that16:32
shardypradk: got a link to the failures you're seeing?16:32
pradkshardy, hmm i see ..16:33
shardythis passed CI so I guess we missed something16:33
pradksure i opened a bug: https://bugs.launchpad.net/tripleo/+bug/161851016:33
openstackLaunchpad bug 1618510 in tripleo "unable to reach redis service" [Undecided,New]16:33
pradkshardy, so you can see the issue in ceilo/aodh/gnocchi logs where it fails to connect to redis coordination backend16:33
jaosorior_awayEmilienM: hey dude, seems that the keystone provider only works with ssl if the ssl parameters were set in the keystone manifest. I'm setting them in the keystone::wsgi::apache. How do we get the provider to take those into account too?16:34
*** jpich has quit IRC16:35
shardypradk: hmm, *coordination_url references "%{hiera('redis_vip')}", so not *_node_ips16:35
EmilienMjaosorior_away: ::keystone::enable_ssl: True16:35
shardypradk: can you reproduce this locally?16:35
EmilienMalso set public_bind_host and admin_bind_host16:35
pradkshardy, i'm in progress of deploying locally16:35
EmilienMjaosorior_away: /me afk lunch a bit let me know please16:36
*** david-lyle_ has joined #tripleo16:36
shardypradk: ack, it'd be good to see the output of hiera aodh::evaluator::coordination_url on the CLI16:36
pradkshardy, but shouldnt the node ips map to vip in haproxy config?16:36
chemansiwen: https://github.com/openstack/puppet-openstacklib/blob/master/lib/puppet/provider/openstack/auth.rb16:36
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: [NO MERGE] Test Ceph RadosGW as replacement for Swift  https://review.openstack.org/35718216:36
shardyand probably grep redis /etc/puppet/hieradata/*16:36
shardypradk: Yeah, there's two data items, the ip of each node, and the vip associated with the service16:37
shardyonly the former (so far) has been reworked16:37
shardyso the change I referenced may not be related16:37
* shardy has a vip rework patch posted, but not yet landed16:37
chemansiwen: in the end whatever code you're doing should ends up in puppet tempest.  It should be some kind of provider that takes a project as input and add the ec2 credential into the tempest conf.16:37
pradkshardy, if we comment out the tcp-check in haproxy it goes through and we can connect to redi via vip16:38
chemansiwen: that basically the only proper way to do it in puppet16:38
pradkshardy, but i dont think we should have to do that16:39
*** electrofelix has quit IRC16:39
chemansiwen: the file you're creating could be used for a simple idempotency test, and should be hashed by project16:39
*** yamahata has quit IRC16:39
*** david-lyle has quit IRC16:39
shardyjaosorior_away had some patches related to haproxy which just landed, so we might want to check those aren't related16:39
pradkok, i'll look at the recently merged16:40
chemansiwen: but the best would be to ask the api for the value each time, but not sure that is possible, so saving parameter in a file might be the only solution.16:40
shardypradk: there was that one, so perhaps ensure hiera redis_enabled returns true16:41
ansiwenchem: which api? keystone api?16:41
shardyand that the interpolation e.g hiera tripleo::haproxy::redis is true16:41
chemansiwen: the ec2 api16:41
openstackgerritMerged openstack/tripleo-heat-templates: Customize the reservation of memory for host processes  https://review.openstack.org/32770516:41
chemansiwen: that's the one that give you the cred back, isn't it ?16:42
ansiwenchem: no, I think you need the credentials to access the api...16:42
chemansiwen: the keystone cred are already taken care by puppet-openstacklib, you don't need to worry about those, or I'm missing something here.16:43
ansiwenchem: you create the credentials with "openstack ec2 credentials create" or so... and then they are stored somewhere (I thought it's keystone)16:43
gfidentepradk, that is sort of "superseded" by https://review.openstack.org/#/c/362800/16:43
gfidentepradk, and it's depends-on https://review.openstack.org/#/c/362798/16:43
gfidentepradk, if there is something we need to amend there, please comment on those16:43
ansiwenchem: well, I'm completely confused by all this, so I'm probably absolutely wrong :-)16:43
ansiwenchem: the problem is I have no idea about the ec2api, but I have to integrate it into tripleo... that's not exactly helpful :-)16:44
*** hewbrocca is now known as hewbrocca-afk16:44
hrybackiayoung one second.  omw back from lunch16:45
chemansiwen: inside puppet provider you can use "request("ec2 credential", "create" ...) and the keystone authentication will be done for your, no problem there.16:45
ansiwenchem: aha! so I need an equivalent of this bash code for puppet-tempest: https://github.com/openstack/ec2-api/blob/master/devstack/create_config#L51-L5316:46
pradkgfidente, hmm so these default to false? https://review.openstack.org/#/c/362798/2/manifests/haproxy.pp16:46
pradkgfidente, where are these overridden to true?16:46
gfidentepradk, heat will dump _enabled: true16:47
gfidentefor those roles which are deployed16:47
openstackgerritimain proposed openstack/tripleo-heat-templates: WIP: Containerized Services for Composable Roles  https://review.openstack.org/33065916:47
openstackgerritimain proposed openstack/tripleo-heat-templates: Bind mount files to run DiD in latest atomic host  https://review.openstack.org/34721816:47
hrybackiayoung: what do you mean by which config? What are you trying to deploy with ooq?16:47
Slowerflaper87: there you go.. hopefully it's all there :)16:48
*** thrash is now known as thrash|biab16:49
pradkgfidente, i see, ok.. so redis_enabled should be turned to true16:49
*** bana_k has quit IRC16:49
*** liverpooler has joined #tripleo16:50
ayounghrybacki, simple developers setup.  HA, one controller16:50
*** fragatina has joined #tripleo16:50
gfidentepradk, redis_enabled should be true yes, let me check CI configs to confirm16:50
pradkgfidente, weird donno why else the vip wouldnt work16:50
*** fragatin_ has joined #tripleo16:51
*** jaosorior_away has quit IRC16:51
gfidentepradk, just found it redis_enabled: true16:51
hrybackiayoung: I don't often deploy HA locally but for vanilla deploy I use the tripleo-quickstart/ci-scripts/fully-deploy.sh script (to closely mimic CI) with this local wrapper https://paste.fedoraproject.org/417438/75851147/16:51
*** trown|lunch is now known as trown16:52
hrybackiayoung: but swap out 'minimal' for 'ha' on line 22 and give that a go16:52
ayounghrybacki, ah, no don't want Mitaka, trying to get up to Newton dev16:52
hrybackiflip 'mitaka' for master16:52
chemansiwen: oki, so you create ec2 cred, then use them in "aws configure ..."16:53
hrybackiayoung: yeah, so swap 'mitaka' with 'master' and 'minimal' with 'ha'16:53
hrybackinote that I'm pulling a local clone of tripleo-quickstart (you might not be doing that?)16:53
*** pradk has quit IRC16:53
*** pradk- has joined #tripleo16:53
ayounghrybacki, so, I was running something pretty minimal, and was getting :16:53
ayoung $  openstack baremetal import --json instackenv.json16:53
ayoung<ayoung> Invalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]16:53
chemansiwen: what is that "aws" command ?16:54
sshnaidmayoung, do you use devmode?16:54
ayoungsshnaidm, I was16:54
gfidentetty tomorrow!16:54
hrybackiayoung: I haven't seen that before (I don't use devmode though)16:54
ayoungsshnaidm, I am running a deploy now without it, but the image fetch takes close to an hour16:54
sshnaidmayoung, I saw such errors in devmode jobs16:54
*** gfidente has quit IRC16:54
*** fragatina has quit IRC16:55
ayoungsshnaidm, I suspected it was something along those lines.  I see a patch to remove devmode16:55
beaglesayoung, did you recently yum update in an existing undercloud?16:55
ayoungWTF does doing anything with images take so long:  either downloading or creating them from scratch is by far the longest part of the deploy.16:55
ansiwenchem: I think the aws configure part is not necessary, is it?16:55
openstackgerritDmitry Ilyin proposed openstack/puppet-pacemaker: Fixes for race condition when adding primitives without OCF present  https://review.openstack.org/35608616:55
beaglesayoung, if so, you probably need to update the mistral workbooks16:55
ayoungbeagles, blew away the undercloud.  THe failure is coming from quickstart files16:55
sshnaidmayoung, afaik there is a problem with performance on rdo artifacts site16:56
hrybackiayoung: yes the ^^16:56
beaglesayoung, oh.. interesting16:56
openstackgerritDmitry Ilyin proposed openstack/puppet-pacemaker: Fixes for race condition when adding primitives without OCF present  https://review.openstack.org/35608616:56
openstackgerritIvan Berezovskiy proposed openstack/puppet-pacemaker: Execute reset_cib in wait_for_online  https://review.openstack.org/36303516:56
ayoungis there any way to speed up the image issue?16:56
beaglesayoung, same answer probably :) I just don't know where you'd need to do it :)\16:56
hrybackiayoung: note that if you have cached images you can circumvent the potential of downloading by handing this extra param (/me goes to get)16:56
chemansiwen: got no idea :)16:56
hrybackiayoung: -e _force_cached_image=true16:57
ayounghrybacki, the problem is that I seem to lose the cached files.  Don't they get put into the /home/stack directory?16:57
hrybackithis will reuse the previously used undercloud image16:57
*** rhallisey has joined #tripleo16:57
hrybackiayoung: no, /var/cache/<something>16:57
pradk-same problem here.. i'm just waiting on fetch_image step for quite some time now16:57
sshnaidmayoung, you can try devmode though, it should still work, but with force_cached_images=false and using the patch: https://review.openstack.org/#/c/335953/16:57
ayoungOh wow look at all them16:58
*** lucasagomes is now known as lucas-dinner16:58
hrybackiayoung: ya, and they eat up a lot of space at like 2.8GB/image16:58
ayoungsshnaidm, dangit, I *KNEW* about that patch16:58
ayounghrybacki, its OK, space I got16:58
*** jpena is now known as jpena|off16:58
chemansiwen: could you just show me a paste of how you would like it to be working as user .  A sample puppet code that would do what you want as if it existed.16:58
ayoungsshnaidm, what would it take to get that patch merged?16:59
hrybackiayoung: did I give you enough to get rolling?16:59
chemansiwen: it would be much easier to work on16:59
sshnaidmayoung, there is a work in progress about new "devmode", so..16:59
sshnaidmweshay, adarazs ^^17:00
sshnaidmweshay, adarazs maybe let's merge this anyway?17:00
*** rasca has quit IRC17:00
ayounglarsks, trown and Gabriele Cerami17:01
ayoungtrown, TYVM17:01
sshnaidmtrown, just was writing to you :))17:01
larsksayoung: what what?17:01
ayounglarsks, you too slow!  trown got it17:01
trownthough... do we even have overcloud images for "devmode" now?17:01
larsksThat means I win!17:02
ayounglarsks, yes.  yes you do.17:02
sshnaidmtrown, yes17:02
trownI looked yesterday and there was no new overcloud images from tripleo-ci17:02
larskstrown: my understanding was that devmode is basically useles now.17:02
larskstrown: I think there is even a "get rid of devmode" review open.17:02
chemansiwen: in the meantime that how you would get in in puppet-tempest https://github.com/openstack/puppet-tempest/blob/master/lib/puppet/provider/tempest_glance_id_setter/openstack.rb#L3117:02
sshnaidmtrown, actually now they are in rh1 cloud, different ip, but should be still available in rh2 too17:02
ansiwenchem: so, I thought the credentials just have to end up in the tempest.conf17:02
trownlarsks: ya it seems that way... though one use case is not suffering the horrible download speeds from centos artifacts server :P17:02
trownsshnaidm: do you have the IP? I think it is not updated in quickstart17:03
ansiwenchem: what I wanted to do with the puppet-tempest module17:03
larskstrown: so actually, solving that is I hope one of our priorities, because I have over the past two days had several people say basically "I can't use quickstart because the downloads take forever"17:03
*** pradk- is now known as pradk17:03
ayoungSo no devmode, right?  Or devmode good?17:03
sshnaidmtrown, yes, but there is no "instack.qcow2", only overcloud:
larsksayoung: devmode not really useful.  I've been using config/release/master.yml for recent versions of things.17:03
larskstrown: this is no instack.qcow2, ONLY ZUUUUUUL.17:04
larsksAh, darn. s/this/there/17:04
ansiwenchem: but maybe you are right, and the aws command is also necessary?17:04
sshnaidmayoung, devmode should work17:04
*** abregman has joined #tripleo17:04
ayoung~/.quickstart/tripleo-quickstart/quickstart.sh  --release master   --extra-vars force_cached_images=true  ayoung-dell-t1700.test17:04
larskssshnaidm: reaallly?  I thought it was trying to fetch images that are no longer updated.17:04
trownlarsks: ya, there is a centos bug opened to fix the md5sum issue with the CDN, https://bugs.centos.org/view.php?id=11367 but ya until we have some alternative place to pull images quickstart is pretty much useless17:04
sshnaidmlarsks, there are updated anyway in devmode17:04
ayoungor do I not want  --extra-vars force_cached_images=true17:05
ansiwenchem: all I know is that I have to put ec2 credentials into the tempest.conf17:05
larskssshnaidm: okay. the last discussion I had with weshay was "devmode is totally broked", but I haven't followed closely since then.17:05
hrybackitrown: do you know if they have made any movement on that bug? adarazs pinged them earlier but no comments :(17:05
trownhrybacki: I dont see anything, but I can bug kb in #centos-devel17:05
chemansiwen: ok, so this is a good description, input: "a project", output: "ec2 cred in the tempest.conf"17:05
sshnaidmlarsks, it should work, this merged patch should have solved all problems17:06
weshayya.. we're updating devmode to use the ansible-role-tripleo-gate and the master image built from current-tripleo17:06
openstackgerritMerged openstack/tripleo-quickstart: Devmode: rename delorean current so any existing repo is overwritten  https://review.openstack.org/33595317:06
sshnaidmlarsks, ^17:06
ayoungweshay, so devmode good, or wait a while?17:06
* ayoung hoping for a set of command line parameters for newton based setup reasonably likely to succeed...17:07
weshaythe incarnation that sshnaidm is talking about will go away.. ayoung if it's working w/ that patch great.  Expect an updated workflow and doc for devmode in the next week or so17:07
ayoungOK, so not today17:07
ayoungany reason to do anything more than17:07
ayoung~/.quickstart/tripleo-quickstart/quickstart.sh  --release master   ayoung-dell-t1700.test17:08
weshayayoung, if you just want to deploy newton that's all you need17:08
trownayoung: I would actually use devmode... since the image will get pulled from tripleo server instead of super slow centos server17:09
ayoungweshay, I want a 3 VM setup:  one undercloud, one controller, one compute.  HA proxy.17:09
ayoungOK  so17:09
ayoung~/.quickstart/tripleo-quickstart/quickstart.sh  --release master --extra-vars @config/general_config/devmode.yml  ayoung-dell-t1700.test17:09
trownayoung: devmode does a complete update of all of the packages on undercloud and overcloud images, so it should not matter that the image itself is out of date17:09
sshnaidmayoung, master-tripleo17:10
trownayoung: s/--release master/--release master-tripleo/17:10
ayoungtrown, now, the last time I ran that, I was getting the error I posted above ..I'll repost17:10
trownthat is the part that points to tripleo server17:10
sshnaidmayoung, devmode uses its own config and "master-tripleo" config17:10
ayoungso don't need devmode to get that?17:10
ayoung~/.quickstart/tripleo-quickstart/quickstart.sh  --release master-tripleo --extra-vars @config/general_config/devmode.yml  ayoung-dell-t1700.test17:11
sshnaidmayoung, yeah17:11
ayoungmy last couple of runs were failing on the post-undercloud, due to a malformed json file:17:11
ayoung$  openstack baremetal import --json instackenv.json17:11
ayoung<ayoung> Invalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]17:11
trownya IIUC the patch I just merged should fix that17:11
*** rcernin has quit IRC17:12
ayoungOK,  running now17:12
*** bana_k has joined #tripleo17:12
openstackgerritMerged openstack/tripleo-validations: Check to ensure debug is disabled on undercloud services.  https://review.openstack.org/35555917:13
*** thrash|biab is now known as thrash17:13
*** tesseract- has quit IRC17:14
*** milan has quit IRC17:15
*** mhenkel has quit IRC17:17
chemansiwen: so are we oki about the provider jobs ?17:17
ayounghrybacki, that option to force cache...what does it do in devmode?17:20
hrybackiayoung: I can't speak to that. I haven't tested devmode personally17:20
ayoungso OOOqs failed again but this time much faster, so ++ there.17:21
hrybackimy work has me mocking CI as closely as possible17:21
ayoungfatal: [undercloud]: FAILED! => {"changed": true, "cmd": "/home/stack/undercloud-install.sh > /home/stack/undercloud_install.log 2>&1", "delta": "0:00:10.040743", "end": "2016-08-30 17:20:42.711475", "failed": true, "rc": 1, "start": "2016-08-30 17:20:32.670732", "stderr": "", "stdout": "", "stdout_lines": [], "warnings": []}17:21
*** yolanda has joined #tripleo17:21
*** pkovar has quit IRC17:21
ayoungand that is the same error I saw before.17:21
sshnaidmayoung, about instackevnv.json?17:21
ayoungthis looked more like the yum error,...17:21
hrybackican you paste undercloud_install.log17:22
ayounghrybacki, yeah, looks like the yum error...debugging17:22
sshnaidmayoung, did you update the quickstart to latest?17:22
sshnaidmayoung, because patch was merged a 15 minutes ago17:23
ayoungsshnaidm, hrybacki http://paste.openstack.org/show/564865/17:23
ayoungsshnaidm, I thought I had...but maybe I fooled myself and jumped the gunnnn17:23
ayoungsshnaidm, yep...when I rebased, the patch, while merged on gerrit, had not made its way to origin/master17:24
ayoungeasy to fix that one...17:24
*** tesseract- has joined #tripleo17:24
ayoungI don't mind failures so long as they fail fast....17:25
*** pcaruana has quit IRC17:25
hrybackiayoung: can you also paste the call to ansible-playbook for me?17:25
ayounghrybacki, for the origianl error?17:26
ayoungthe JSON one...sure?17:26
hrybackiwell for whichever is the immediate problem17:26
ayounghrybacki, I'm good for the moment, as I had missed a known-necessary patch17:27
* hrybacki failed to read up17:27
ayoungit failed in a different way this time than it did before, just looked the same at first glance17:27
*** tesseract- has quit IRC17:28
ansiwenchem: sorry, was afk17:29
ansiwenchem: so, isn't puppet-tempest already responsible for adding stuff to tempest.conf17:30
*** tosky has quit IRC17:30
chemansiwen: hum ... yes, what's the question ?17:31
chemansiwen: anyway, the way to go is take inspiration from https://github.com/openstack/puppet-tempest/blob/master/lib/puppet/provider/tempest_neutron_net_id_setter/openstack.rb#L3117:34
openstackgerritMerged openstack/tripleo-heat-templates: Mv Nova, Neutron, Horizon out of controller.yaml  https://review.openstack.org/36136217:35
chemansiwen: you see the keystone auth is done there https://github.com/openstack/puppet-tempest/blob/master/lib/puppet/provider/tempest_neutron_net_id_setter/openstack.rb#L817:35
chemansiwen: then you make the request as in line 3217:35
chemansiwen: you get back the values and add them to tempest.conf17:36
*** florianf has quit IRC17:36
chemansiwen: it's all there really17:36
ansiwenchem: unfortunately I have to go now, but thanks a lot for now! I will look at it tomorrow, and come back to you with my many questions ;-)17:36
ansiwenchem: again, thanks a lot!17:36
chemansiwen: np :)17:36
EmilienMdprince, bnemec: any chance to get https://review.openstack.org/361409 ? I have more coming17:38
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: Enable modern authtoken for Neutron and Glance  https://review.openstack.org/36298017:43
colonwqEmilienM, Can you look at 339106 and 247956? I have hopefully knocked all of the rough edges off.17:43
EmilienMcolonwq: looking17:44
EmilienMcolonwq: done17:49
EmilienMgood work here17:49
EmilienMread my comment17:49
*** vijaykc4 has quit IRC17:51
*** egafford has quit IRC17:56
*** egafford has joined #tripleo17:56
colonwqEmilienM, Thanks. I'll fix17:56
EmilienMshardy: mitaka releases17:56
EmilienMerr, released17:56
EmilienMi hate my new US keyboard17:56
*** mcornea has joined #tripleo17:56
*** zoli|wfh is now known as zoli|gone17:58
*** akrivoka has joined #tripleo17:59
* bnemec wonders if they swapped the s and d on EmilienM's new keyboard ;-)17:59
*** zoli|gone is now known as zoli_gone-proxy17:59
EmilienMbnemec: dude I have a new laptop18:00
EmilienMand i lost french keyboard18:00
EmilienMi'll need 3 months to learn that18:00
bkerooh no D:18:00
openstackgerritAde Lee proposed openstack/tripleo-heat-templates: Add Barbican to the overcloud - WIP  https://review.openstack.org/35245818:00
colonwqEmilienM, Do you need some key cap stickers?18:00
bkeroNo French keyboard in Canada? This sounds like it violates some law.18:00
bnemecThis is why I have a very nice mechanical keyboard.  I don't care what keyboard my laptop has.18:00
*** jdob has quit IRC18:01
EmilienMyeah i need to buy one18:01
*** jdob has joined #tripleo18:01
jtomasekjpeeler: reviewed, https://review.openstack.org/#/c/355598/218:02
jtomasekjpeeler: almost there18:02
jpeelerjtomasek: perfect, i'll get it cleaned up for you to look at tomorrow18:02
jtomasekjpeeler: cool, once you have it, I could take a look immediately18:03
jtomasekjpeeler: thanks!18:04
*** tesseract- has joined #tripleo18:05
beaglesDas Keyboard ftw!18:05
*** tesseract- has quit IRC18:05
beaglesmy keyboard has no labels18:06
*** tesseract- has joined #tripleo18:06
beaglessecurity by obscurity from the non-touch-typist18:06
bkeroI hope you remember which of 3 the insert and scroll-lock keys are :)18:07
*** tesseract- has quit IRC18:08
EmilienMbeagles: 260 CAD !!! what???18:08
beaglesEmilienM, what? holy crap18:08
beaglesEmilienM, it was not that expensive when I bought it18:08
beaglesmm.. the new ones have more features18:09
beaglesEmilienM, mine was like 130 -140 .. of course the CDN dollar wasn't so crappy then either18:09
*** yamahata has joined #tripleo18:10
*** tesseract- has joined #tripleo18:10
openstackgerritBrad P. Crochet proposed openstack/tripleo-common: Allow for building specific images  https://review.openstack.org/36309518:11
openstackgerritMerged openstack/tripleo-heat-templates: Add NeutronGlobalPhysnetMtu to neutron-base.yaml  https://review.openstack.org/36153918:11
ayounghrybacki, specatacular failage18:12
ayoungthere is so much log spew I can;t scroll to the top of the failing section18:12
EmilienMI found what I need18:12
bkeroEmilienM: can you fix it with setxkbmap?18:14
hrybackiayoung: lol. I know that pain &> <some_file>.log is a need18:14
ayoungEmilienM, I once had to try and type on A French Keyboard.  T'was not easy18:14
EmilienMbkero: yeah18:14
ayounghrybacki, looks like something in the image update.  Lots of RPM/Yum type commands in there18:14
bkeroPut them in French DVORAK layout :D18:14
trownayoung: I am seeing the same thing... looking into it18:15
openstackgerritSagi Shnaidman proposed openstack-infra/tripleo-ci: Split additional features across the periodic jobs  https://review.openstack.org/36290418:18
openstackgerritDimitri Savineau proposed openstack/tripleo-heat-templates: apache: Remove product informations  https://review.openstack.org/36310518:19
sshnaidmshardy, can you please merge it? https://review.openstack.org/#/c/355641/18:20
EmilienMsshnaidm: done18:22
sshnaidmEmilienM, thanks!18:22
EmilienMsshnaidm: I want to work with you on something when you can18:23
sshnaidmEmilienM, what is it?18:23
EmilienMsshnaidm: now we have the scenarios almost in place I would like to investigate if we could run tempest sometimes18:23
EmilienMsshnaidm: have you followed my stuff on scenarios?18:23
sshnaidmEmilienM, partially, where should I look firstly?18:24
openstackgerritBrad P. Crochet proposed openstack/tripleo-docs: Switch docs to use tripleo-build-images  https://review.openstack.org/33766918:24
EmilienMsshnaidm: https://review.openstack.org/#/c/362504/18:24
sshnaidmEmilienM, in which utc are you?18:25
EmilienMsshnaidm: I'm in polar circle18:25
EmilienMnow I'm joking18:25
EmilienMsshnaidm: nothing urgent, we can talk about it later this week18:25
sshnaidmEmilienM, ok, then I'll look at it today and let's sync tomorrow when you're here18:26
sshnaidmEmilienM, ok18:26
EmilienMbnemec, dprince: I have another patch for composable CI jobs, all green https://review.openstack.org/#/c/362504/18:27
*** nyechiel has quit IRC18:28
*** tesseract- has quit IRC18:30
*** tesseract- has joined #tripleo18:30
*** jdob has quit IRC18:30
*** tesseract- has quit IRC18:30
*** jdob has joined #tripleo18:31
*** tesseract- has joined #tripleo18:31
openstackgerritBrad P. Crochet proposed openstack/tripleo-docs: Switch docs to use tripleo-build-images  https://review.openstack.org/33766918:32
*** yamahata has quit IRC18:37
*** tesseract- has quit IRC18:37
openstackgerritMerged openstack/tripleo-incubator: Replace sshnaidm SSH public key  https://review.openstack.org/36282418:38
*** fragatin_ has quit IRC18:38
*** rbrady has quit IRC18:38
*** egafford1 has joined #tripleo18:41
sshnaidmpabelanger, hi18:43
*** myoung is now known as myoung|bbl18:43
sshnaidmpabelanger, I use on rh1 cloud the template-centos-7-1472086539 image as undercloud, but I don't see there /etc/nodepool/provider. /etc/nodepool dir is empty. Do you know what could be a problem?18:44
*** egafford has quit IRC18:44
pabelangersshnaidm: how did you launch it?18:45
sshnaidmpabelanger, with ovb18:45
pabelangersshnaidm: okay, manually? That info is populated by nodepool18:45
sshnaidmpabelanger, create-env etc18:45
sshnaidmpabelanger, yeah, manually. On which step is it populated?18:45
sshnaidmpabelanger, I thought is comes with this already18:46
pabelangerno, there is a process in nodepool.o.o, when we launch a node we first try to SSH into it, if we can, then we populate the provider / node information18:46
pabelangeryour create-env process will need to do that18:47
pabelangeror we see how to get nodepool.o.o to manage OVB18:47
sshnaidmnot good.. but how can it get this info and from where18:47
*** dsariel has quit IRC18:47
*** yamahata has joined #tripleo18:48
sshnaidmat least those AFS mirrors and env vars that are required for toci_gate_test.sh18:48
pabelangersince you launch ovb from your test jobs, you'll have to sync of the data from /etc/nodepool from the zuul worker18:49
sshnaidmpabelanger, ok, how do I do it?18:50
pabelangersshnaidm: http://git.openstack.org/cgit/openstack-infra/nodepool/tree/nodepool/nodepool.py#n630 is the process18:50
pabelangerthat nodepool uses today18:50
*** ramishra has quit IRC18:50
*** etp has quit IRC18:50
pabelangersshnaidm: how do you copy data to ovb nodes today?18:50
*** saneax-_-|AFK is now known as saneax18:51
*** abregman has quit IRC18:51
*** jtomasek has quit IRC18:51
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Add env. setup for virt deployments doc  https://review.openstack.org/36312018:51
sshnaidmpabelanger, what do you mean? undercloud image I have it there already as "template_centos-<timestamp>"18:51
*** ianw has quit IRC18:51
*** apetrich has quit IRC18:52
*** jrist has quit IRC18:52
*** NobodyCam has quit IRC18:52
*** cmyster has quit IRC18:52
*** therve has quit IRC18:52
*** ramishra has joined #tripleo18:52
pabelangersshnaidm: right, so after the ovb node is online, scp /etc/nodepool/providers ovh_node:/etc/nodepool/providers, then run your test18:53
*** cmyster has joined #tripleo18:53
sshnaidmpabelanger, but I don't have /etc/nodepool/providers anywhere, that's the problem18:53
*** yolanda has quit IRC18:53
sshnaidmpabelanger, you said it's populated when nodepool takes this image for job, right?18:54
*** rbrady has joined #tripleo18:54
pabelangersshnaidm: are you saying it is failing in the gate or for manual testing?18:54
*** NobodyCam has joined #tripleo18:54
sshnaidmpabelanger, manual testing on the cloud, it should be the same as in jobs, but without nodepool18:54
sshnaidmpabelanger, https://etherpad.openstack.org/p/tripleo-ci-devenvs18:55
*** etp has joined #tripleo18:56
*** therve has joined #tripleo18:56
pabelangersshnaidm: logic needs to be added to toc_gate_test.sh to support this then.  Because, this works today in the gate right? Otherwise ovb nodes would not be able to access the AFS mirror18:58
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Update section header indicators  https://review.openstack.org/36052218:58
pabelangerbut, from a tripleo-centos-7 node, you'd scp the nodepool info to the ovb node18:58
sshnaidmpabelanger, I think nodepool/zuul inject there these dirs when starting the job, and I don't have it because don't use them18:58
pabelangersince nodepool doesn't manage OVH18:58
pabelangersshnaidm: yes, nodepool will add them to tripleo-centos-7 nodes18:59
pabelangernot ovh18:59
sshnaidmpabelanger, yeah, I'm looking how to do it, so the question - can I get the data from anywhere?18:59
pabelanger/etc/nodepool/provider on tripleo-centos-7 nodes18:59
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Update section header indicators  https://review.openstack.org/36052219:00
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Add env. setup for virt deployments doc  https://review.openstack.org/36312019:00
bnemecExcellent.  I think we may indeed have broken ipv6.19:00
bnemecError: Error from DataBinding 'hiera' while looking up 'tripleo::trusted_cas::ca_map': (<unknown>): found unexpected ':' while scanning a plain scalar at line 102 column 32 on node overcloud-novacompute-0.localdomain19:00
sshnaidmpabelanger, sorry, I meant nodepool injects it to tripleo-centos-7 nodes - so I don't have it there right now19:00
bnemecSeems to be unhappy about this line: nova::compute::vncproxy_host: [2001:db8:fd00:1000::14]19:01
pabelangersshnaidm: which server are you looking at?19:01
sshnaidmpabelanger, I'll reformulate - can I get this info without any involvment of nodepool?19:01
sshnaidmpabelanger, for example template-centos-7-1472086539 on rh1 cloud19:01
pabelangersshnaidm: you can, once you have /etc/nodepool/provider the contents are static. you'll be able to reuse it mulitple times in tripleo-test-cloud-rh1, however I would not manually manage the file19:03
pabelangerin case we make changes in openstack-infra19:03
*** jtomasek has joined #tripleo19:03
pabelangersshnaidm: http://paste.openstack.org/show/564885/ is all the file contains19:04
*** apetrich has joined #tripleo19:04
sshnaidmpabelanger, in best case I wanted to do it transparent for CI, without additional patches.. seems like it's not the case19:04
*** abregman has joined #tripleo19:05
*** david-lyle_ is now known as david-lyle19:05
pabelangersshnaidm: Can you provide more context? I am having a hard time following what the issue is19:05
sshnaidmpabelanger, sorry need to leave now, I'll talk to you tomorrow then, when you'll have a time, is it ok? maybe in bluejeans it'll be faster to explain19:07
*** jrist has joined #tripleo19:07
*** jrist has joined #tripleo19:07
sshnaidmfor now will try you file, thanks19:07
*** sshnaidm is now known as sshnaidm|afk19:08
*** coolsvap_ has quit IRC19:12
hrybackitrown: do you have time to BJ about docs?19:14
*** mcornea has quit IRC19:14
trownhrybacki: sure19:15
pabelangersshnaidm|afk: if you are trying to get reproducibility CI build downstream from zuul / nodepool. I wouldn't spend too much time write code. We are already working on solving that issue with zuulv3. We already have the ansible playbooks created today, for somebody to re-run a job. EG: http://logs.openstack.org/10/346510/3/check/gate-openstack-ansible-docs-ubuntu-xenial/9b5a401/_zuul_ansible/playbook19:15
pabelangerand we can solve for the issue of launching a cloud resource today with ansible19:16
*** vijaykc4 has joined #tripleo19:16
hrybackitrown: https://bluejeans.com/855901362319:16
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: Enable modern authtoken for Neutron and Glance  https://review.openstack.org/36298019:19
*** mlupton has joined #tripleo19:21
*** ianw_ has joined #tripleo19:28
saneaxFor bp/sriov to close these needs a workflow, jenkins passed - https://review.openstack.org/#/c/361430/19:30
*** dprince has quit IRC19:35
ayoungtrown, any clue?  Need any help debugging it?19:35
trownayoung: something is failing in the update image part of devmode and I was hoping it was something with a cached image so I tried nuking all my old images, but hit the same thing19:37
openstackgerritMathieu Mitchell proposed openstack/diskimage-builder: (WIP) Zuul-cloner second attempt  https://review.openstack.org/36301319:38
*** jcoufal has quit IRC19:38
ayoungtrown, I'm running again now redirecting to a logfile trying to get a better error message19:38
ayoungit was not clear what actually failed19:38
*** rcernin has joined #tripleo19:39
openstackgerritMathieu Mitchell proposed openstack/diskimage-builder: (WIP) Zuul-cloner third attempt  https://review.openstack.org/36301319:39
*** jcoufal has joined #tripleo19:40
openstackgerritMerged openstack-infra/tripleo-ci: Bump tempest version to latest  https://review.openstack.org/35564119:42
*** fragatina has joined #tripleo19:43
openstackgerritMartin AndrĂ© proposed openstack/puppet-tripleo: Manage tripleo-ui configuration files with puppet  https://review.openstack.org/36316719:43
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Add fluentd client service  https://review.openstack.org/35350619:45
EmilienMbnemec: another one ready https://review.openstack.org/#/c/361412/19:48
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Availability monitoring agents support  https://review.openstack.org/25478819:50
*** dprince has joined #tripleo19:50
*** fzdarsky is now known as fzdarsky|afk19:52
flaper87Slower: thanks19:53
*** akshai has joined #tripleo19:53
*** jcoufal has quit IRC19:54
openstackgerritEmilien Macchi proposed openstack/puppet-tripleo: midonet: update neutron parameter  https://review.openstack.org/36317819:59
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Add fluentd client service  https://review.openstack.org/35350620:00
*** shardy has quit IRC20:02
ayoungtrown, so...the image errors went away, and I am back to20:04
ayoung+ openstack baremetal import --json instackenv.json20:04
ayoungInvalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]20:04
*** shardy has joined #tripleo20:05
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Add fluentd client service  https://review.openstack.org/35350620:05
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Last round of modern authtoken update  https://review.openstack.org/36318220:05
trownayoung: hmm... I am still stuck on the image update errors20:06
trowntrying to boot up the undercloud.qcow2  to manually hack on it since the virt-customize output isnt helping20:07
*** jcoufal has joined #tripleo20:08
shardypradk: Hey re your Redis issue, try openstack stack environment show overcloud | grep RedisPassword20:09
shardythat should tell you if the problem is with the value passed by the client, or something wrong in the templates/puppet20:10
EmilienMshardy: fyi we found this issue with my scenario testing work, scenario001 fails to run Gnocchi API calls20:10
ayoungtrown, what is "capabilities": "profile:{{ node.flavor }},boot_option:local"20:11
ayoungin instackenv.json.j220:11
shardyEmilienM: Yeah I think that's what pradk is investigating, and it's been traced to a misconfiguration related to Redis20:11
*** pradk has quit IRC20:11
trownayoung: that is the capabilities that get applied to the node in ironic, it is used to match nodes to the correct roles20:12
shardy(unless your failure is some other issue)20:12
EmilienMshardy: no it's the same20:12
shardyEmilienM: pradk was saying the hiera for the redis password was unset, so the first step is the command above to check the parameter is set20:13
EmilienMshardy: do we need this parameter manually set or is ther a default?20:13
shardythen work backwards from the hieradata files on the node20:13
shardyEmilienM: it's set inside python-tripleoclient20:13
ayoung$ openstack baremetal import --json instackenv.json20:13
ayoungInvalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]20:13
EmilienMok so the template in CI should be good20:13
shardywe never set defaults for passwords inside the templates20:13
ayoungseems to happen whether I have that boot_option:local value or not20:14
EmilienMshardy: have you gone through all bugs for triage? I finally have time now20:14
shardyEmilienM: Not yet, been busy with other things this afternoon20:16
*** pradk has joined #tripleo20:16
*** akrivoka has quit IRC20:16
shardyany help appreciated, but I know you're busy so don't spend too much time on it20:16
*** dsariel has joined #tripleo20:18
trownayoung: I am pretty confused... booted the undercloud that was failing in that update script and ran yum update manually successfully... maybe devmode is indeed borked atm... I will continue to dig on it tomorrow, but in the meantime you could use the CDN location to pull a master image for RDO20:20
trownit is only a couple days old20:20
ayoungtrown, but that does not seem to be what is killing my installs20:20
EmilienMshardy: there are 61 bugs not triaged20:21
EmilienMshardy: in "new" status20:21
ayoungthis is with the JSON used to enroll nodes, and I've had a consistent error with it, module some other things messing up earlier in the process20:21
EmilienMshardy: I'll try to get them triaged at least20:21
openstackgerritBrent Eagles proposed openstack/tripleo-heat-templates: Add missing reference to ExternalPort for DVR  https://review.openstack.org/36318820:22
*** jayg is now known as jayg|g0n320:22
trownayoung: ya, nothing has changed with that instackenv.json in months20:22
ayoungtrown, then it is something in Ironic?20:23
ayoungNo one else is seeing this, are they:20:23
trownayoung: so it has to be some issue with the set of packages we are getting via devmode20:23
ayounghmmm, ok20:23
ayoungtrown, so  ~/.quickstart/tripleo-quickstart/quickstart.sh  --release master  ayoung-dell-t1700.test  \20:24
trownthe error is a cryptic mistral error... so I am not sure exactly what is going on there, and I can not even get my environment to that point20:24
openstackgerritMathieu Mitchell proposed openstack/diskimage-builder: Introduce 'zuul' as a new source-repositories type  https://review.openstack.org/36301320:24
trownayoung: "--extra-vars undercloud_image_url=http://buildlogs.centos.org/centos/7/cloud/x86_64/tripleo_images/master/delorean/undercloud.qcow2"20:24
beaglestrown, ayoung, I was just running into this today after updating20:25
beaglestrown, ayoung I was able to resolve20:25
beagleslooking for what I did20:26
*** saneax is now known as saneax-_-|AFK20:27
EmilienMdprince: can you make auto triage when you file a bug? example with https://bugs.launchpad.net/tripleo/+bug/161848820:27
openstackLaunchpad bug 1618488 in tripleo "upload-swift-artifacts fails with: More than one endpoint exists with the name 'swift'." [Undecided,New]20:27
EmilienMis it for rc1 ?20:27
EmilienMcan it slip to ocata?20:27
EmilienMi'll let you change infos on the bug20:27
EmilienMshardy: same for you https://bugs.launchpad.net/tripleo/+bug/161844120:28
openstackLaunchpad bug 1618441 in tripleo "ovb job fails to find eth1 nic" [Undecided,New]20:28
EmilienMshardy: we have "ci" tag to track all CI bugs20:28
beaglesayoung, trown: openstack workbook update /usr/share/tripleo-common/workbooks/baremetal.yaml or some such20:28
openstackgerritHarry Rybacki proposed openstack/tripleo-quickstart: Add env. setup for virt deployments doc  https://review.openstack.org/36312020:28
ayoungbeagles, which issue did that resolve?20:29
beaglesayoung, Invalid input [name=tripleo.baremetal.v1.register_or_update, class=DirectWorkflowSpec, unexpected=[u'instance_boot_option']]20:30
EmilienMI need to remind everyone again to make auto triage when filing a bug20:30
EmilienMwe have so much new bugs every day not triaged :(20:30
beaglesayoung, or are you running into something else now20:30
ayoungbeagles, nah, that is the problem20:30
trownhmm shouldnt that happen as part of the undercloud install though20:31
ayoungbeagles, any idea of parameters? I have never seen the 'workbook' subcommand before20:31
*** hrybacki is now known as hrybacki|afk20:31
openstackgerritMerged openstack/tripleo-heat-templates: Added DPDK environment file and nic configs samples  https://review.openstack.org/36123820:32
beaglestrown, it should.. if there was a yum update after install it might pull it in20:32
beaglesayoung, it's a consequence of moving to mistral for some bits20:32
trownthat seems like a bug that we might be masking in tripleo-ci20:32
beaglestrown, sorry.. it might pull incompatible changes in.20:33
openstackgerritMathieu Mitchell proposed openstack/diskimage-builder: Introduce 'zuul' as a new source-repositories type  https://review.openstack.org/36301320:33
*** rcernin has quit IRC20:33
*** rcernin has joined #tripleo20:33
ayoungbeagles, OK, walk me through this.  I can list the workbooks, and I see the one you refernce20:34
ayoungopenstack workbook show tripleo.baremetal.v1  shows a very short summary20:34
*** trown is now known as trown|outtypewww20:35
ayoungI can read through the whole  /usr/share/tripleo-common/workbooks/baremetal.yaml  file, but how do I know what is actiely being executed?20:35
beaglesayoung, yep. tbh, ianae - I just pick up on tidbits20:35
beaglesayoung, oh that's a big question :)20:36
shardyEmilienM: I didn't self triage that because it was reported by someone else on IRC20:36
shardyI can do it though I suppose20:36
EmilienMshardy: thanks20:36
ayoungbeagles, help has a strang thing20:36
ayoung  workbook show definition20:36
shardyAll CI bugs can go to the ongoing milestone instead of rc1 I think20:36
ayoungbut...that is not an executable command20:36
EmilienMshardy: sorry, so many bugs20:36
ayoungopenstack workbook show definition20:36
ayoungWorkbook not found [workbook_name=definition]20:36
shardyEmilienM: Yeah, thanks for the help! :)20:37
beaglesayoung: should be openstack workbook show tripleo.baremetal.v1 or the UUID that shows up in the list20:37
ayoungno UUID20:38
ayoungbeagles, http://paste.openstack.org/show/564887/20:38
beaglesayoung, I might be confusing with openstack workflow list20:38
beaglesayoung,  but anyways the name should work (else a bug)20:39
ayoungbut... I see in workflow list...20:39
ayoung049ee948-2517-4b0d-9b8a-         | tripleo.baremetal.v1.register_or | fee8ef3e33254a4f9baad3f600d9b995 | <none> | nodes_json, remove=False,... | 2016-08-30 20:02:41 | None20:39
*** ianw_ is now known as ianw20:39
ayoungso that is the failing operation20:40
ayoung workflow show definition20:41
ayoung  also does not work20:41
beaglesayoung, there is a bug in the help text20:41
ayoungshould there be a hyphen or something?20:42
beaglesayoung, what I suspect has happened is that the workbooks were created before tripleo-common was updated or something like that so it is firing data into mistral that it does not know about and won't accept20:43
ayoungbeagles, so I need an updated  tripleo-common?20:43
beaglesayoung, so by updating the workbook you bring things up to sync. It's similar to updating code and having to run the whatevercomponent-db-sync so that the db tables are in sync with the cod20:44
beaglesayoung, actually what I said is that your tripleo-common is already up to date, mistral needs new data20:44
ayoungbeagles, so I ran the command you suggested but no change20:44
ayoungopenstack workbook update /usr/share/tripleo-common/workbooks/baremetal.yaml20:45
beaglesayoung, ah okay20:45
ayoungbeagles, doing that should load something into mistral, right>?20:45
ayoungand how do I see that something?20:45
*** myoung|bbl is now known as myoung20:45
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Add fluentd client service  https://review.openstack.org/35350620:46
ayoungshort of looking direcly in the database, which I am totally ready to do20:46
beaglesayoung, something should.. mmm. is the 'instance_boot_option' in that file20:46
beaglesayoung, ie. does "grep instance_boot_option /usr/share/tripleo-common/workbooks/baremetal.yaml" find anything?20:46
ayoung grep boot /usr/share/tripleo-common/workbooks/baremetal.yaml20:47
beaglesayoung: ah... yeah mmm. Seems like there is more amiss than what I was suggesting20:49
ayoungbeagles, is there a mistral CLI?20:49
beaglesayoung, either your working set of packages is out of sync or there is a bug in those packages20:50
beaglesayoung, sure, you were just using it I think20:50
*** adarazs has quit IRC20:50
ayoungmistral workbook-get-definition tripleo.baremetal.v1 | grep boot20:50
ayoungnothing there....20:50
* beagles nods20:51
*** adarazs has joined #tripleo20:52
beaglesayoung, if you are using a version of the tripleoclient that does have that option and mistral doesn't know about it then I could imagine there being issues20:53
ayoungbeagles, let me see where mistral is, then,....20:53
*** shardy has quit IRC20:53
beaglesayoung, check if your version of python-tripleoclient and tripleo-common are in sync, of similar source repos, etc.20:54
*** lblanchard has quit IRC20:54
beaglesayoung, outside of that .. I dunno. I just piece bits together as I go.20:54
ayoungbeagles, would that be mistral server that needs to be updated?20:55
*** yolanda has joined #tripleo20:55
ayoungpretty recent20:55
ayoungbut aweek old20:55
dprinceEmilienM: ack on auto-triage20:55
ayoungthey look in sync to me20:56
beaglesayoung, no I don't thinks so. Mistral should not care about this.. it would be a problem with the workbooks/workflows and whatever is trying to use them20:56
beaglesayoung, ie. the tripleoclient maybe?20:56
ayoung rpmquery -f /usr/share/tripleo-common/workbooks/baremetal.yaml20:57
*** chem has quit IRC20:57
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Availability monitoring agents support  https://review.openstack.org/25478821:01
openstackgerritBen Nemec proposed openstack/tripleo-heat-templates: De-bracket vncproxy_host in compute profile  https://review.openstack.org/36320421:01
*** egafford1 has quit IRC21:01
openstackgerritBen Nemec proposed openstack/tripleo-heat-templates: De-bracket vncproxy_host in compute profile  https://review.openstack.org/36320421:02
openstackgerritLars Kellogg-Stedman proposed openstack/tripleo-heat-templates: Add fluentd client service  https://review.openstack.org/35350621:03
*** radeks has joined #tripleo21:07
*** jcoufal has quit IRC21:08
*** dprince has quit IRC21:08
*** mlupton has quit IRC21:10
beagleslaunchpad is unhappy21:10
*** radeks has quit IRC21:11
*** jcoufal has joined #tripleo21:12
openstackgerritRonelle Landy proposed openstack/tripleo-quickstart: WIP: Add directory to store documentation files  https://review.openstack.org/36148321:12
*** rhallisey has quit IRC21:16
*** ayoung has quit IRC21:19
openstackgerritMerged openstack/os-collect-config: Merge deployments data for zaqar collectors  https://review.openstack.org/34383521:19
*** maeca1 has quit IRC21:22
*** rhefner has joined #tripleo21:23
*** mlupton has joined #tripleo21:26
*** morazi has quit IRC21:26
*** yolanda has quit IRC21:28
*** rlandy is now known as rlandy|brb21:28
*** dbecker has quit IRC21:29
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Last round of modern authtoken update  https://review.openstack.org/36318221:34
*** dsariel has quit IRC21:36
*** jeckersb is now known as jeckersb_gone21:36
*** jcoufal has quit IRC21:43
*** jcoufal has joined #tripleo21:46
openstackgerritJeff Peeler proposed openstack/tripleo-common: Include environments in capabilities output  https://review.openstack.org/35559821:47
*** mlupton has quit IRC21:48
*** mlupton has joined #tripleo21:48
*** ppradhan has left #tripleo21:53
*** rlandy|brb is now known as rlandy21:56
*** jcoufal has quit IRC21:58
*** jcoufal has joined #tripleo22:03
*** rajinir has quit IRC22:05
*** vijaykc4 has quit IRC22:12
*** akuznetsov has joined #tripleo22:13
*** cdearborn has quit IRC22:14
*** vijaykc4 has joined #tripleo22:18
*** ianw has quit IRC22:26
*** rlandy has quit IRC22:27
*** apetrich has quit IRC22:28
*** apetrich has joined #tripleo22:30
*** jrist has quit IRC22:32
*** akuznetsov has quit IRC22:33
*** jrist has joined #tripleo22:39
openstackgerritMerged openstack/instack-undercloud: configure heat with new authtoken parameters  https://review.openstack.org/36140922:42
openstackgerritMerged openstack/tripleo-heat-templates: Move ceph cluster config out of allNodesConfig  https://review.openstack.org/35506722:45
*** Ryjedo_ has joined #tripleo22:53
*** mpsairam has quit IRC22:54
*** Ryjedo has quit IRC22:54
*** Ryjedo_ is now known as Ryjedo22:54
*** mpsairam has joined #tripleo22:55
*** ianw has joined #tripleo22:57
openstackgerritRyan Brady proposed openstack/tripleo-common: Wire in jinja templating for custom roles  https://review.openstack.org/36246522:57
*** ayoung has joined #tripleo22:59
*** pradk has quit IRC23:00
*** hrybacki|afk is now known as hrybacki23:10
*** saneax-_-|AFK is now known as saneax23:12
*** abregman has quit IRC23:14
*** abregman has joined #tripleo23:16
*** mlupton has quit IRC23:28
*** rcernin has quit IRC23:35
*** jcoufal has quit IRC23:35
*** mlupton has joined #tripleo23:41
*** mlupton has quit IRC23:45

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