| openstackgerrit | Markos Chandras proposed openstack/diskimage-builder: lib: common-functions: Fix tmpfs umounting https://review.openstack.org/392002 | 00:09 |
|---|---|---|
| *** maticue has quit IRC | 00:15 | |
| *** artom_ has joined #tripleo | 00:17 | |
| *** artom has quit IRC | 00:20 | |
| *** noslzzp has quit IRC | 00:20 | |
| *** trozet has quit IRC | 00:20 | |
| *** trozet has joined #tripleo | 00:21 | |
| *** noslzzp has joined #tripleo | 00:21 | |
| *** gfidente|afk has quit IRC | 00:21 | |
| *** jrist has quit IRC | 00:21 | |
| *** jrist has joined #tripleo | 00:21 | |
| *** jrist has joined #tripleo | 00:21 | |
| *** gfidente|afk has joined #tripleo | 00:22 | |
| *** egafford has quit IRC | 00:24 | |
| *** egafford has joined #tripleo | 00:33 | |
| *** egafford has quit IRC | 00:38 | |
| *** akshai has quit IRC | 00:51 | |
| *** bana_k has quit IRC | 00:52 | |
| *** bana_k has joined #tripleo | 00:54 | |
| *** bana_k has quit IRC | 01:01 | |
| dhill_ | I don't understand why it fails : https://review.openstack.org/388920 ... any hints? | 01:08 |
| *** chlong has quit IRC | 01:09 | |
| *** rlandy has quit IRC | 01:17 | |
| openstackgerrit | Merged openstack/diskimage-builder: Don't set tracing in environment files https://review.openstack.org/388972 | 01:43 |
| *** bfournie1 has left #tripleo | 01:44 | |
| *** dmacpher has joined #tripleo | 01:45 | |
| *** dougbtv has joined #tripleo | 01:47 | |
| *** bfournie has joined #tripleo | 01:49 | |
| openstackgerrit | Merged openstack/diskimage-builder: Don't log datestamp by default in functional tests https://review.openstack.org/390126 | 01:55 |
| *** b00tcat has quit IRC | 02:01 | |
| *** limao has joined #tripleo | 02:17 | |
| *** ramishra has joined #tripleo | 02:40 | |
| *** rdopiera has quit IRC | 02:53 | |
| *** bkopilov has quit IRC | 02:58 | |
| *** bkopilov_ has quit IRC | 02:58 | |
| *** rdopiera has joined #tripleo | 03:02 | |
| *** rdopiera has quit IRC | 03:07 | |
| *** bana_k has joined #tripleo | 03:08 | |
| *** rdopiera has joined #tripleo | 03:09 | |
| *** ramishra has quit IRC | 03:21 | |
| openstackgerrit | RedHat RDO CI proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/365449 | 03:30 |
| *** bana_k has quit IRC | 03:43 | |
| *** links has joined #tripleo | 03:52 | |
| *** bana_k has joined #tripleo | 04:01 | |
| *** ramishra has joined #tripleo | 04:06 | |
| *** bana_k has quit IRC | 04:14 | |
| *** dmacpher has quit IRC | 04:15 | |
| *** jeckersb is now known as jeckersb_gone | 04:18 | |
| *** bkopilov has joined #tripleo | 04:19 | |
| *** bkopilov_ has joined #tripleo | 04:19 | |
| *** tzumainn has quit IRC | 04:25 | |
| *** dmacpher has joined #tripleo | 04:28 | |
| *** akshai has joined #tripleo | 04:35 | |
| *** akshai has quit IRC | 04:36 | |
| *** dsariel has quit IRC | 04:45 | |
| *** bana_k has joined #tripleo | 05:06 | |
| *** limao has quit IRC | 05:14 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add support for SUSE in dhcp-all-interfaces https://review.openstack.org/391772 | 05:15 |
| *** limao has joined #tripleo | 05:15 | |
| *** limao has quit IRC | 05:15 | |
| *** limao has joined #tripleo | 05:16 | |
| *** limao has quit IRC | 05:20 | |
| *** bana_k has quit IRC | 05:27 | |
| *** dsariel has joined #tripleo | 05:27 | |
| *** bana_k has joined #tripleo | 05:54 | |
| *** ebalduf has quit IRC | 05:55 | |
| *** ramishra_ has joined #tripleo | 06:00 | |
| *** flepied1 has quit IRC | 06:01 | |
| *** ramishra has quit IRC | 06:02 | |
| *** dsariel has quit IRC | 06:16 | |
| *** limao has joined #tripleo | 06:16 | |
| *** limao has quit IRC | 06:21 | |
| *** rcernin has joined #tripleo | 06:29 | |
| *** bana_k has quit IRC | 06:30 | |
| *** florianf has joined #tripleo | 06:30 | |
| *** florianf has quit IRC | 06:31 | |
| *** mcornea has joined #tripleo | 06:31 | |
| *** florianf has joined #tripleo | 06:31 | |
| *** limao has joined #tripleo | 06:44 | |
| openstackgerrit | Marios Andreou proposed openstack/tripleo-heat-templates: Add replacepkgs to the manual ovs upgrade workaround and fix a typo https://review.openstack.org/390826 | 06:58 |
| *** bvandenh has quit IRC | 07:03 | |
| *** florianf has quit IRC | 07:03 | |
| *** akuznetsov has joined #tripleo | 07:07 | |
| *** paramite has joined #tripleo | 07:08 | |
| openstackgerrit | Merged openstack/diskimage-builder: Changed the home-page of diskimage-builder in setup.cfg https://review.openstack.org/391496 | 07:11 |
| *** florianf has joined #tripleo | 07:12 | |
| *** jprovazn has joined #tripleo | 07:16 | |
| *** dmacpher has quit IRC | 07:16 | |
| paramite | Hi guys, can I do reverify on https://review.openstack.org/#/c/390802/ myself or somebody from core has to? | 07:16 |
| *** jtomasek_ has joined #tripleo | 07:16 | |
| paramite | meh, sorry for the noise, I saw +1 on place where it was not :) | 07:19 |
| paramite | But anyway, can we merge the patch? :) | 07:19 |
| *** hjensas has joined #tripleo | 07:19 | |
| *** hjensas has quit IRC | 07:19 | |
| *** hjensas has joined #tripleo | 07:19 | |
| *** tesseract has joined #tripleo | 07:21 | |
| *** tesseract is now known as Guest14381 | 07:21 | |
| *** bvandenh has joined #tripleo | 07:23 | |
| *** dmacpher has joined #tripleo | 07:31 | |
| *** akuznetsov has quit IRC | 07:31 | |
| *** akuznetsov has joined #tripleo | 07:31 | |
| *** jtomasek_ has quit IRC | 07:36 | |
| *** yamahata has quit IRC | 07:40 | |
| *** mhenkel has joined #tripleo | 07:43 | |
| *** ohamada has joined #tripleo | 07:59 | |
| *** ohamada has quit IRC | 08:00 | |
| d0ugal | paramite: ideally we need the ovb-ha job to pass too | 08:04 |
| *** chandankumar has joined #tripleo | 08:12 | |
| *** ebarrera has joined #tripleo | 08:13 | |
| *** hewbrocca_afk is now known as hewbrocca | 08:17 | |
| *** flepied has joined #tripleo | 08:17 | |
| *** akuznetsov has quit IRC | 08:18 | |
| *** zoli_gone-proxy is now known as zoliXXL | 08:19 | |
| *** florianf has quit IRC | 08:24 | |
| *** ccamacho has quit IRC | 08:24 | |
| *** ohamada has joined #tripleo | 08:28 | |
| *** zoliXXL is now known as zoli_gone-proxy | 08:30 | |
| *** aufi has joined #tripleo | 08:32 | |
| *** jtomasek_ has joined #tripleo | 08:34 | |
| *** zoli_gone-proxy is now known as zoliXXL | 08:37 | |
| *** florianf has joined #tripleo | 08:38 | |
| *** lucas-afk is now known as lucasagomes | 08:40 | |
| *** dixiaoli has joined #tripleo | 08:43 | |
| *** dixiaoli has quit IRC | 08:46 | |
| *** social has joined #tripleo | 08:54 | |
| *** athomas has joined #tripleo | 08:59 | |
| *** ohamada has quit IRC | 08:59 | |
| *** liverpooler has joined #tripleo | 09:00 | |
| *** flepied has quit IRC | 09:09 | |
| *** jaosorior has joined #tripleo | 09:15 | |
| *** jpich has joined #tripleo | 09:18 | |
| jaosorior | sshnaidm: ping | 09:23 |
| sshnaidm | jaosorior, hi | 09:23 |
| jaosorior | sshnaidm: hey dude, I saw that the job merged :D | 09:23 |
| sshnaidm | jaosorior, yeah, finally :) | 09:23 |
| sshnaidm | jaosorior, did you run experimental on your patch? | 09:24 |
| jaosorior | sshnaidm: well, I could but it won't do anything yet | 09:24 |
| *** jbadiapa has quit IRC | 09:24 | |
| sshnaidm | jaosorior, just for testing | 09:24 |
| jaosorior | sshnaidm: any idea how I can provision just one node to be | 09:25 |
| jaosorior | sshnaidm: ok done | 09:25 |
| jaosorior | sshnaidm: so yeah, one node so I can install FreeIPA there? | 09:25 |
| jaosorior | sshnaidm: I kinda confused the workflow with the multinode, where the node are provisioned already | 09:25 |
| jaosorior | so now I guess I have to do it via ironic also | 09:25 |
| *** dtantsur|afk is now known as dtantsur | 09:26 | |
| sshnaidm | jaosorior, yeah, in ovb it's with ironic | 09:26 |
| jaosorior | sshnaidm: fuck, I don't see the new job in the zuul queue | 09:26 |
| *** akrivoka has joined #tripleo | 09:27 | |
| *** panda|Zz is now known as panda|pto|PH | 09:28 | |
| *** milan has joined #tripleo | 09:28 | |
| sshnaidm | jaosorior, telnet://66.187.229.202:19885 | 09:29 |
| sshnaidm | jaosorior, everything is good :) | 09:29 |
| sshnaidm | jaosorior, it's experimental-tripleo pipeline | 09:29 |
| jaosorior | sshnaidm: sshnaidm couldn't access that address :/ | 09:30 |
| jaosorior | nor with telnet, nor with the browser, not sure what I'm supposed to do with it | 09:30 |
| *** abregman has joined #tripleo | 09:30 | |
| sshnaidm | jaosorior, running in console "telnet 66.187.229.202 19885" | 09:31 |
| sshnaidm | jaosorior, it's weird, I can connect and see job running | 09:31 |
| sshnaidm | jaosorior, anyway, it works, now you can start configuring it | 09:32 |
| jaosorior | sshnaidm: true that, thanks man! | 09:32 |
| sshnaidm | jaosorior, regarding ovb I'm not sure what is best here.. Don't you have heat template or separate role for this server? | 09:33 |
| jaosorior | sshnaidm: nope | 09:33 |
| jaosorior | I can get a heat template though | 09:33 |
| jaosorior | but no role | 09:33 |
| jaosorior | sshnaidm: it's a bit awkward, cause even the undercloud will be registered as an IPA client (well, FreeIPA client... now ironic-python-agent uses IPA too x_x) | 09:34 |
| jaosorior | the IPA acronym is overused | 09:34 |
| *** derekh has joined #tripleo | 09:35 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/tripleo-quickstart: Force the use of python/pip version 2 https://review.openstack.org/368013 | 09:35 |
| *** zoliXXL is now known as zoli|wfh | 09:36 | |
| sshnaidm | jaosorior, found now: https://adam.younglogic.com/2016/09/deploying-server-on-ironic-node-baseline/ | 09:36 |
| sshnaidm | jaosorior, seems like ayoung blog | 09:36 |
| jaosorior | sshnaidm: was thinking about using this https://github.com/resurrexit/freeipa-heat-template/blob/master/freeipa_server.yaml | 09:36 |
| sshnaidm | jaosorior, it seems nice | 09:38 |
| sshnaidm | jaosorior, either deploying server with ironic and then install there manually | 09:38 |
| jaosorior | sshnaidm: yeah, an intern in our team came up with it, I just have to tweak it and remove some hardcoded stuff | 09:38 |
| jaosorior | sshnaidm: any idea what the DNS is for the hypervisors in OVB? | 09:38 |
| sshnaidm | jaosorior, 8.8.8.8 and undercloud(?) | 09:39 |
| sshnaidm | jaosorior, let me check.. | 09:39 |
| jaosorior | I need the one for the undercloud | 09:40 |
| sshnaidm | jaosorior, 8.8.8.8 , 8.8.4.4 and some 10.1.8.10 - but for overcloud images only | 09:42 |
| jaosorior | any idea where the 10.1.8.10 comes from? | 09:43 |
| sshnaidm | jaosorior, rh1 cloud internal I suppose | 09:43 |
| sshnaidm | jaosorior, it's better to use google dns I think | 09:43 |
| jaosorior | sshnaidm: right, so I want to add a DNS entry, so the overcloud nodes know the address of the FreeIPA instance | 09:44 |
| jaosorior | sshnaidm: cause all they get is the hostname | 09:44 |
| sshnaidm | jaosorior, /etc/hosts ? | 09:45 |
| jaosorior | sshnaidm: yeah... I guess that's the only way at the moment | 09:46 |
| *** oshvartz has joined #tripleo | 09:47 | |
| *** jaosorior has quit IRC | 09:48 | |
| *** jaosorior has joined #tripleo | 09:48 | |
| *** limao has quit IRC | 09:50 | |
| *** limao has joined #tripleo | 09:50 | |
| *** limao has quit IRC | 09:55 | |
| panda|pto|PH | sshnaidm: I found the root cause, but I don't know to who I can talk to to fix it, and I don't know how to write a selinux policy (and where) | 09:55 |
| sshnaidm | panda|pto|PH, is it about running redis with systemd? | 09:56 |
| sshnaidm | panda|pto|PH, or just running redis..? | 09:56 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 09:57 |
| jaosorior | panda|pto|PH: gotta talk to rhallisey, but he's not online yet | 09:57 |
| jaosorior | panda|pto|PH: also, if you know the root cause and can reproduce it, you gotta run in permissive mode, then grep the audit log for the service you know that's causing issues, and pass those audit logs to rhallisey | 09:58 |
| jaosorior | he'll use audit2allow to create a policy | 09:58 |
| openstackgerrit | Merged openstack/instack-undercloud: Enable telemetry services by default on undercloud https://review.openstack.org/386702 | 09:58 |
| jaosorior | sshnaidm: how did you get the IP that I need to use to check the progress of that specific job? | 09:58 |
| sshnaidm | jaosorior, from http://status.openstack.org/zuul/ | 09:59 |
| sshnaidm | jaosorior, find you job running there and copy link | 09:59 |
| sshnaidm | jaosorior, it's in experimental-tripleo pipeline | 10:00 |
| jaosorior | sshnaidm: thanks dude! | 10:00 |
| *** owalsh has joined #tripleo | 10:00 | |
| *** mgould|afk is now known as mgould | 10:05 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Enable TLS in the internal network for Cinder API https://review.openstack.org/392090 | 10:07 |
| *** dsariel has joined #tripleo | 10:08 | |
| panda|pto|PH | jaosorior: already done the permissive part, redis is not allowed to create the socket when in enforcing mode, but in permissive is ok, I already have the denied in audit log | 10:08 |
| panda|pto|PH | jaosorior: https://bugs.launchpad.net/tripleo/+bug/1637961 | 10:09 |
| openstack | Launchpad bug 1637961 in tripleo "periodic HA master job pingtest times out" [High,Confirmed] - Assigned to Gabriele Cerami (gcerami) | 10:09 |
| jaosorior | panda|pto|PH: where's the denied audit log? | 10:11 |
| *** null_ref has quit IRC | 10:11 | |
| panda|pto|PH | jaosorior: http://logs.openstack.org/periodic/periodic-tripleo-ci-centos-7-ovb-ha/0ae3179/logs/overcloud-controller-0/var/log/audit/audit.txt.gz | 10:11 |
| panda|pto|PH | jaosorior: search for "denied" | 10:11 |
| *** dsariel_ has joined #tripleo | 10:12 | |
| jaosorior | panda|pto|PH: aah great, didn't see that in the laucnhpad bug | 10:12 |
| jaosorior | panda|pto|PH: alright, so now it's a matter of getting rhallisey to take a look at it | 10:12 |
| jaosorior | he'll add that to the openstack-selinux-rules package | 10:12 |
| panda|pto|PH | ah, that's what I weas looking for :) thanks. I have a name, and a package name :) | 10:13 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable internal TLS for Cinder API https://review.openstack.org/392095 | 10:15 |
| *** dsariel has quit IRC | 10:15 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add support for SUSE in dhcp-all-interfaces https://review.openstack.org/391772 | 10:16 |
| *** shardy has joined #tripleo | 10:18 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Enable TLS in the internal network for Nova API https://review.openstack.org/388495 | 10:18 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable internal TLS for Nova API https://review.openstack.org/388503 | 10:19 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Enable internal TLS for MySQL https://review.openstack.org/378471 | 10:19 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable TLS in the internal networkf or Mysql https://review.openstack.org/378537 | 10:20 |
| *** akrivoka has quit IRC | 10:22 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable internal TLS for Nova API https://review.openstack.org/388503 | 10:22 |
| *** thrash|g0ne is now known as thrash | 10:23 | |
| *** anton has quit IRC | 10:24 | |
| jistr | jaosorior: hi, when you have some time, can you review these please? https://review.openstack.org/#/q/topic:bug/1635606 | 10:26 |
| *** anton has joined #tripleo | 10:27 | |
| *** chandankumar has quit IRC | 10:28 | |
| *** chandankumar has joined #tripleo | 10:28 | |
| *** oshvartz has quit IRC | 10:30 | |
| jaosorior | jistr: sure thing | 10:30 |
| openstackgerrit | Steven Hardy proposed openstack-infra/tripleo-ci: Use parameter_merge_strategies for multinode env https://review.openstack.org/384618 | 10:32 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 10:32 |
| *** akrivoka has joined #tripleo | 10:36 | |
| *** rodrigods has quit IRC | 10:37 | |
| *** rodrigods has joined #tripleo | 10:37 | |
| *** anton has quit IRC | 10:38 | |
| *** anton has joined #tripleo | 10:38 | |
| *** akuznetsov has joined #tripleo | 10:41 | |
| *** chandankumar has quit IRC | 10:43 | |
| *** chandankumar has joined #tripleo | 10:43 | |
| jaosorior | shardy: hey dude, I remember it's possible to deploy a setup with only one controller, but that uses the pacemaker environment too, do you remember how to do it? | 10:45 |
| *** oshvartz has joined #tripleo | 10:45 | |
| shardy | jaosorior: you just include this environment I think: | 10:47 |
| shardy | https://github.com/openstack/tripleo-heat-templates/blob/master/environments/puppet-pacemaker.yaml | 10:47 |
| *** akuznetsov has quit IRC | 10:47 | |
| jaosorior | shardy: so basically I do the same as I would usually do to deploy an HA setup but don't specify the controller count? | 10:48 |
| shardy | jaosorior: yup, I believe that's right | 10:51 |
| jaosorior | shardy: thanks | 10:51 |
| *** dprince has joined #tripleo | 10:56 | |
| *** chlong has joined #tripleo | 10:56 | |
| *** oshvartz has quit IRC | 10:57 | |
| *** jkilpatr has quit IRC | 10:57 | |
| *** zoli|wfh is now known as zoli|lunch | 11:02 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable internal TLS for Cinder API https://review.openstack.org/392095 | 11:06 |
| *** liverpooler has quit IRC | 11:07 | |
| *** liverpooler has joined #tripleo | 11:07 | |
| *** bkopilov has quit IRC | 11:07 | |
| *** bkopilov_ has quit IRC | 11:07 | |
| *** liverpooler has quit IRC | 11:10 | |
| *** oshvartz has joined #tripleo | 11:11 | |
| jaosorior | if someone has time, could you review these commits? https://review.openstack.org/#/q/topic:barbican-integration they' | 11:15 |
| jaosorior | they're adding barbican to the overcloud | 11:15 |
| adarazs | lucasagomes: hi Lucas. re: https://review.openstack.org/#/c/368013/6 -- why do we need the options to specify python and pip if there's only one valid value for them? | 11:15 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Change nova ram_allocation_ratio to match puppet-nova https://review.openstack.org/392108 | 11:16 |
| lucasagomes | adarazs, well, it could be "python" for distros with python 2.7 as the default interpreter, and "python2.7" for distros with python3 as the default interpreter | 11:19 |
| lucasagomes | adarazs, my first patch I was trying to lookup for the right version in code but it was asked to add the options | 11:20 |
| adarazs | lucasagomes: oh. okay. followup question: if you create a virtenv with python2 won't the pip also be python2 based? or it might happen that there's no pip for python2 installed? | 11:21 |
| adarazs | lucasagomes: re: str(sys.version_info[0]) + str(sys.version_info[1]) == "27" -- it seems if you use sys.version_info >= (2,7) instead that works fine. not sure what kind of python black magic is that but it seems to give back the proper true/false for me with various values :) | 11:22 |
| openstackgerrit | Merged openstack/puppet-tripleo: NFS mounting for Glance file backend https://review.openstack.org/389284 | 11:23 |
| lucasagomes | adarazs, I think the pip should work depending on which version of python interpreter it's created, that's true | 11:24 |
| lucasagomes | adarazs, I can give it a go locally | 11:24 |
| lucasagomes | adarazs, oh, I can try to the version_info as well | 11:25 |
| *** liverpooler has joined #tripleo | 11:25 | |
| adarazs | I read it on stackoverflow so it must be true :) | 11:25 |
| lucasagomes | going to mark that patch as WIP and give it a go soon-ish | 11:25 |
| lucasagomes | adarazs, heh right :-D | 11:26 |
| lucasagomes | adarazs, mind adding those comments there ? | 11:26 |
| lucasagomes | there == in the patch | 11:26 |
| adarazs | sure. | 11:26 |
| lucasagomes | ty! | 11:26 |
| *** jeckersb_gone is now known as jeckersb | 11:27 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 11:28 |
| adarazs | done. | 11:29 |
| *** ohamada has joined #tripleo | 11:32 | |
| *** jkilpatr has joined #tripleo | 11:34 | |
| *** pradk has joined #tripleo | 11:35 | |
| *** dsariel_ has quit IRC | 11:36 | |
| jtomasek | shardy: is it ok to consider DefaultPasswords, ServiceNetMap, EndpointMap parameters internal and explicitly exclude them in GUI? | 11:37 |
| *** fultonj has joined #tripleo | 11:38 | |
| shardy | jtomasek: DefaultPasswords yes, but not ServiceNetMap, as that's the interface operators use to assign a service to a specific network | 11:39 |
| shardy | jtomasek: in network/service_net_map.j2.yaml we expose two parameters, ServiceNetMapDefaults (which mostly I would not expect operators to modify), and ServiceNetMap (which overrides ServiceNetMapDefaults) | 11:40 |
| jtomasek | shardy: ok, and EndpointMap? | 11:41 |
| shardy | jtomasek: I'd say yes for now ignore EndpointMap as it's generated by a script from endpoint_data.yaml | 11:41 |
| jaosorior | dprince: can you check these commits? They're for getting barbican to the overcloud https://review.openstack.org/#/q/topic:barbican-integration | 11:41 |
| shardy | so we'd need a new workflow to allow operators to modify that | 11:41 |
| dprince | jaosorior: ack | 11:41 |
| jtomasek | shardy: ok, thanks | 11:41 |
| shardy | jtomasek: I'm not sure the best way to handle ServiceNetMap - we pass that parameter around in a few places, but the UI probably wants to treat those as internal, but expose a ServiceNetMap interface _somewhere_ populated with the ServiceNetMapDefaults (or perhaps just a manual override that's pased via ServiceNetMap for now) | 11:43 |
| jaosorior | dprince: thanks | 11:47 |
| *** tobias_fiberdata has joined #tripleo | 11:48 | |
| jaosorior | shardy, jistr: if you have time :D https://review.openstack.org/#/q/topic:barbican-integration | 11:49 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: swift/proxy: remove swift::proxy::ceilometer::rabbit_host https://review.openstack.org/391864 | 11:49 |
| EmilienM | hello | 11:50 |
| jaosorior | it's not production ready but it's a start. We still need to make a DogTag profile to have a real backend for barbican. | 11:50 |
| jaosorior | EmilienM: hey dude, how's it going? | 11:50 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: nova/libvirt: add missing ports for live-migration https://review.openstack.org/391876 | 11:50 |
| EmilienM | jaosorior: excellent and you | 11:50 |
| jaosorior | EmilienM: a bit flu-ish. but drinking vitamins and coffee helps | 11:51 |
| EmilienM | looks like CI ovb/ha is flappy | 11:52 |
| *** anton is now known as aarapov | 11:52 | |
| EmilienM | if not broken | 11:53 |
| jtomasek | shardy: good point, I think this may be included in network configuration work. I'll keep displaying that parameter at each service parameters list for now, untill we have a better way to let user configure ServiceNetMap | 11:53 |
| *** aarapov is now known as anton | 11:53 | |
| EmilienM | ConnectFailure: resources.server1: Unable to establish connection to http://10.0.0.14:8774/v2.1/os-volumes_boot: ('Connection aborted.', BadStatusLine("''",)) | 11:53 |
| jaosorior | dafuq | 11:53 |
| EmilienM | someone already saw that ^ ? | 11:53 |
| jaosorior | EmilienM: is that going on in the pingtest? | 11:54 |
| *** dsariel has joined #tripleo | 11:54 | |
| EmilienM | jaosorior: yes | 11:55 |
| EmilienM | I'm investigating in logstash | 11:55 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Parameters listing updates https://review.openstack.org/391855 | 11:56 |
| jaosorior | EmilienM: can you pass a commit that shows that issue? gotta check the cinder logs | 11:56 |
| thrash | dprince: taking a look at the composable undercloud... Looks like the undercloud deploy needs to either process the jinja templates itself, or use the mistral action to do so | 11:56 |
| EmilienM | jaosorior: well, take a recent ha job on tripleo.org/cistatus.html | 11:56 |
| jaosorior | EmilienM: s/cinder/nova | 11:56 |
| dprince | thrash: hey, yes! exactly | 11:56 |
| EmilienM | sounds like it happens since a promotion | 11:57 |
| thrash | dprince: I spent some time yesterday investgating running mistral the same as we do heat. should be possible. | 11:57 |
| EmilienM | it might be something in Cinder | 11:57 |
| dprince | thrash: I meant to look into it last week in BCN but of course didn't get a chance. Would be happy for you to help right this... | 11:57 |
| dprince | thrash: write even | 11:57 |
| dprince | thrash: write it right | 11:57 |
| thrash | lol | 11:57 |
| thrash | dprince: I'll keep hacking at the mistral thing. I think that's the right way to go. | 11:58 |
| EmilienM | jaosorior: it happenned first on October 27th | 11:58 |
| jaosorior | EmilienM: during the summit O_O | 11:58 |
| dprince | thrash: actually, I was leaning towards a local version of the jinja template function | 11:58 |
| thrash | dprince: :D | 11:59 |
| shardy | I think a script that runs the j2 rendering would also be useful for t-h-t development | 11:59 |
| jtomasek | florianf: could you please check pending GUI patches so we manage to backport them before next newton tripleo release (tomorrow afternoon) | 11:59 |
| dprince | thrash: adding Mistral, while potentially useful, would better fall towards a later feature I think | 11:59 |
| florianf | jtomasek: yep, on it. | 12:00 |
| dprince | thrash: unless of course we had other things in mind to do w/ Mistral via the UC installer bits | 12:00 |
| EmilienM | jaosorior: I'm digging nova logs | 12:00 |
| jtomasek | florianf: thanks! | 12:00 |
| thrash | dprince: ok... either one shouldn't be too monumental. We can always refactor to the mistral action. | 12:00 |
| jaosorior | shardy: what does the jinja rendering at the moment? if that's done via mistral one could just run that task individually, no? | 12:01 |
| thrash | jaosorior: it is in mistral, but that requires mistral. :D | 12:01 |
| shardy | jaosorior: yes it's done by a mistral action, but in the composable undercloud case there is no mistral atm | 12:01 |
| shardy | and it's sometimes a little clearer when errors happen if you just run j2 locally | 12:01 |
| shardy | so a script that calls the exact same code as the mistral action would be useful, I think | 12:01 |
| openstackgerrit | Marios Andreou proposed openstack/tripleo-specs: Composable Service Upgrades https://review.openstack.org/392116 | 12:02 |
| marios | shardy: first pass ^^^ not much detail yet :) mainly cos we don't got it :) | 12:02 |
| shardy | marios: ack, thanks will check it out after lunch | 12:02 |
| thrash | shardy: a refactor of the jinja processing into a utility would certainly be useful. For the undercloud deploy and an external script | 12:03 |
| jaosorior | EmilienM: where did you see that BadStatus error? I haven't found it yet. | 12:03 |
| thrash | shardy: I'll start there. | 12:03 |
| *** rbowen has joined #tripleo | 12:03 | |
| *** fzdarsky has joined #tripleo | 12:03 | |
| *** pkovar has joined #tripleo | 12:03 | |
| shardy | thrash: ack sounds good - note rbrady did start splitting up the mistral actions, not sure if that's a useful starting point or not | 12:04 |
| shardy | https://review.openstack.org/#/c/370868/8 | 12:04 |
| thrash | shardy: I'll take a look. | 12:04 |
| EmilienM | jaosorior: http://logs.openstack.org/64/391064/1/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/2d15c11/console.html#_2016-10-31_10_19_54_304885 | 12:04 |
| *** fzdarsky is now known as fzdarsky|afk | 12:05 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Fix default for barbican documentation https://review.openstack.org/392119 | 12:09 |
| jaosorior | jistr: ^^ | 12:09 |
| fultonj | https://review.openstack.org/#/c/391064/ has gotten two +2s. Could someone +1 the workflow so it can get merged? | 12:10 |
| *** chandankumar has quit IRC | 12:10 | |
| jistr | jaosorior: wfm, +2+A | 12:11 |
| *** zaneb has joined #tripleo | 12:11 | |
| EmilienM | jaosorior: I see nothing in nova/cinder logs | 12:12 |
| EmilienM | looks like boot from volume doesn't work or is unstable | 12:12 |
| openstackgerrit | Merged openstack/puppet-tripleo: Fixes transparent binding to OpenDaylight in HA Proxy https://review.openstack.org/391660 | 12:12 |
| jaosorior | EmilienM: only intance of that I found is in the heat logs, but yeah, same here, I still find nothing in the nova or cinder volumes. I'm trying to search in the system.journal | 12:14 |
| jaosorior | shardy: would it be possible to get someone from the heat team to add support for httpd deployment of heat in tripleo? | 12:14 |
| jaosorior | shardy: would need the puppet-heat work and then to add in tripleo | 12:14 |
| *** tzumainn has joined #tripleo | 12:15 | |
| *** lucasagomes is now known as lucas-hungry | 12:16 | |
| *** fzdarsky|afk has quit IRC | 12:16 | |
| openstackgerrit | Lukas Bezdicka proposed openstack/tripleo-heat-templates: Update openstack-puppet-modules dependencies https://review.openstack.org/392123 | 12:17 |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: CI test - never merge https://review.openstack.org/392124 | 12:18 |
| *** pkovar has quit IRC | 12:21 | |
| EmilienM | jaosorior: I hold on a little bit debug now, I saw a recent promotion last night, I'll see if we still have this problem later | 12:21 |
| jaosorior | EmilienM: I still don't find anything... is it possible to to turn selinux off for the images? might want to test with that | 12:22 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 12:23 |
| *** noslzzp has quit IRC | 12:23 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Re-add NFS backend for Glance https://review.openstack.org/389285 | 12:23 |
| *** maeca1 has joined #tripleo | 12:23 | |
| *** jkilpatr_ has joined #tripleo | 12:23 | |
| *** jkilpatr has quit IRC | 12:24 | |
| *** noslzzp has joined #tripleo | 12:24 | |
| *** akuznetsov has joined #tripleo | 12:25 | |
| *** jayg|g0n3 is now known as jayg | 12:27 | |
| *** maticue has joined #tripleo | 12:29 | |
| *** abregman has quit IRC | 12:30 | |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: Re-add NFS backend for Glance https://review.openstack.org/392128 | 12:31 |
| *** zoli|lunch is now known as zoli|wfh | 12:31 | |
| *** tobias-fiberdata has joined #tripleo | 12:32 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: WIP - Deploy TripleO with Puppet 4 https://review.openstack.org/371209 | 12:33 |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: WIP - Deploy TripleO with Puppet 4 https://review.openstack.org/371209 | 12:33 |
| openstackgerrit | Jiri Stransky proposed openstack/puppet-tripleo: NFS mounting for Glance file backend https://review.openstack.org/392131 | 12:34 |
| *** lhinds has quit IRC | 12:35 | |
| *** tobias_fiberdata has quit IRC | 12:35 | |
| *** jcoufal has joined #tripleo | 12:36 | |
| *** pkovar has joined #tripleo | 12:37 | |
| *** jerrygb_ has quit IRC | 12:38 | |
| *** akshai has joined #tripleo | 12:39 | |
| *** lhinds has joined #tripleo | 12:39 | |
| *** rlandy has joined #tripleo | 12:41 | |
| *** lhinds has quit IRC | 12:42 | |
| *** lhinds has joined #tripleo | 12:43 | |
| *** chandankumar has joined #tripleo | 12:44 | |
| openstackgerrit | Lukas Bezdicka proposed openstack/tripleo-heat-templates: Update openstack-puppet-modules dependencies https://review.openstack.org/392123 | 12:45 |
| openstackgerrit | Merged openstack/diskimage-builder: Remove RedHat grub workaround install https://review.openstack.org/388580 | 12:49 |
| *** abregman has joined #tripleo | 12:51 | |
| *** jerrygb has joined #tripleo | 12:52 | |
| *** ramishra_ has quit IRC | 12:53 | |
| openstackgerrit | Tim Rozet proposed openstack/puppet-tripleo: Fixes transparent binding to OpenDaylight in HA Proxy https://review.openstack.org/392142 | 12:55 |
| *** derekh has quit IRC | 12:56 | |
| shadower | hey folks, what's the deal with specs and blueprints? When to use which? | 12:56 |
| *** akuznetsov has quit IRC | 12:57 | |
| shadower | I'd like to add two changes to the validations API (have them optionally run in the deployment workflows and allowing multiple sources of validations & their updating) | 12:57 |
| EmilienM | shadower: specs are useful when we need to discuss about code design or new policies | 12:57 |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: Rework gnocchi-upgrade to run in a separate upgrade step https://review.openstack.org/388649 | 12:57 |
| EmilienM | shadower: blueprints are usually some tasks that doesn't require much design, and allow developers to quickly implement new features | 12:58 |
| EmilienM | if you're creating a new API or system, a spec is better, as the team needs to agree on design | 12:58 |
| shadower | EmilienM: cool, thanks. Looks like I sholud write up a couple of specs, then | 12:58 |
| *** ayoung has quit IRC | 12:58 | |
| shadower | yeah | 12:58 |
| openstackgerrit | Merged openstack/puppet-tripleo: Add barbican profile https://review.openstack.org/352447 | 12:58 |
| openstackgerrit | Merged openstack/puppet-tripleo: Add barbican profile rspec testing https://review.openstack.org/388099 | 12:59 |
| *** akuznetsov has joined #tripleo | 13:00 | |
| jaosorior | EmilienM: damn dude, I'm quite stumped... haven't found what's up with CI :/ | 13:02 |
| jaosorior | EmilienM: have you found anything? | 13:02 |
| EmilienM | nothing | 13:03 |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-common: Stop trying to power off nodes right after registering https://review.openstack.org/392148 | 13:03 |
| dtantsur | what's the correct lp project to use for bugs: tripleo or tripleo-common? | 13:04 |
| EmilienM | dtantsur: tripleo | 13:04 |
| dtantsur | ack | 13:04 |
| * dtantsur fixes his bug | 13:04 | |
| *** akuznetsov has quit IRC | 13:04 | |
| *** ramishra has joined #tripleo | 13:06 | |
| *** fultonj has quit IRC | 13:06 | |
| *** fultonj has joined #tripleo | 13:08 | |
| *** abregman_ has joined #tripleo | 13:10 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Add replacepkgs to the manual ovs upgrade workaround and fix a typo https://review.openstack.org/390826 | 13:11 |
| *** rhallisey has joined #tripleo | 13:13 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: nova/libvirt: add missing ports for live-migration https://review.openstack.org/391876 | 13:13 |
| *** abregman has quit IRC | 13:14 | |
| *** links has quit IRC | 13:14 | |
| openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Undercloud upgrade for mitaka and newton https://review.openstack.org/381286 | 13:16 |
| *** lblanchard has joined #tripleo | 13:16 | |
| *** lucas-hungry is now known as lucasagomes | 13:18 | |
| *** zoli|wfh is now known as zoli|afk | 13:21 | |
| jaosorior | slagle: are you around? | 13:23 |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-common: Stop trying to power off nodes right after registering https://review.openstack.org/392148 | 13:25 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-specs: Propose TripleO Squads policy https://review.openstack.org/385201 | 13:27 |
| *** morazi has joined #tripleo | 13:28 | |
| EmilienM | meeting in 30 min, feel free to add off items topics https://etherpad.openstack.org/p/tripleo-meeting-items | 13:29 |
| jaosorior | dtantsur: I'm a bit confused with https://review.openstack.org/#/c/392148/ the commit message mentions that powering off the nodes is useless but then that's being done in the workflow, what's up? | 13:31 |
| slagle | jaosorior: yes | 13:31 |
| jaosorior | slagle: hey, do you happen to know what the current blocker is for getting multinode jobs to deploy net-iso? | 13:31 |
| dtantsur | jaosorior, people get confused when seeing the nodes in "power on" state :) | 13:31 |
| dtantsur | jaosorior, I'm fine with just dropping it otherwise | 13:31 |
| jaosorior | dtantsur: having that on the workflow is fine for me, I was just a bit confused on the commit message, and maybe you had some extra context for me to understand | 13:32 |
| *** egafford has joined #tripleo | 13:32 | |
| slagle | jaosorior: there is no blocker that i know of as of now | 13:32 |
| *** fragatina has joined #tripleo | 13:32 | |
| dtantsur | jaosorior, lemme try to rework the commit message | 13:32 |
| *** fragatina has quit IRC | 13:32 | |
| slagle | jaosorior: it's just a matter of someone doing it | 13:32 |
| *** fragatina has joined #tripleo | 13:33 | |
| jaosorior | slagle: and basically it would be single-nic-with-vlans, right? | 13:33 |
| jaosorior | slagle: or do we get multiple nics with multinode? | 13:33 |
| slagle | jaosorior: i'd probably start with single nics | 13:34 |
| EmilienM | jaosorior: right, we have found multinode job was sometimes not catching bugs we cought in ovb | 13:34 |
| slagle | jaosorior: and the jobs need to also be updated to actually use os-net-config | 13:34 |
| jaosorior | slagle: what do they use at the moment if os-net-config is not being used? | 13:34 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-tripleoclient: Updated from global requirements https://review.openstack.org/389945 | 13:35 |
| *** morazi has quit IRC | 13:35 | |
| slagle | jaosorior: they use a combination of a function from devstack-gate to create vxlan tunnels initially, and then on oc deployment, the nic templates just run an ip command to add the ip addresses to the already configured interfaces | 13:35 |
| *** absubram has joined #tripleo | 13:36 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-common: Power off new nodes when making them available, not right after enrolling https://review.openstack.org/392148 | 13:36 |
| dtantsur | jaosorior, this might be better ^^^ | 13:36 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/tripleo-validations: Updated from global requirements https://review.openstack.org/391170 | 13:36 |
| jaosorior | dtantsur: that works for me, thank you | 13:37 |
| jaosorior | slagle: oh, I see | 13:37 |
| openstackgerrit | Ana Krivokapic proposed openstack/python-tripleoclient: Only show deprecated args for overcloud deploy https://review.openstack.org/392166 | 13:39 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 13:40 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Fix default for barbican documentation https://review.openstack.org/392119 | 13:48 |
| *** morazi has joined #tripleo | 13:48 | |
| *** fragatina has quit IRC | 13:48 | |
| openstackgerrit | Merged openstack/tripleo-specs: Add promotion-blocker tag https://review.openstack.org/388094 | 13:53 |
| openstackgerrit | Merged openstack/tripleo-specs: Remove spec-lite tag https://review.openstack.org/388095 | 13:53 |
| *** ebalduf has joined #tripleo | 13:53 | |
| *** cdearborn has joined #tripleo | 13:55 | |
| *** chem has quit IRC | 13:56 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add support for ubuntu and InfiniBand indhcp-all-interfaces https://review.openstack.org/391772 | 13:56 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add element cloud-init-allow-password-authentication https://review.openstack.org/391765 | 13:56 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: redhat-common add option to select networking service NetworkManager/network https://review.openstack.org/392170 | 13:56 |
| *** derekh has joined #tripleo | 13:57 | |
| EmilienM | tripleo meeting is about to start on #openstack-meeting-alt, feel free to join ! | 13:58 |
| *** jerrygb_ has joined #tripleo | 14:00 | |
| openstackgerrit | Sagi Shnaidman proposed openstack/tripleo-quickstart: Fix tempest runner in devmode https://review.openstack.org/391312 | 14:02 |
| *** jerrygb has quit IRC | 14:02 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: redhat-common add option to select networking service NetworkManager/network https://review.openstack.org/392170 | 14:03 |
| *** liverpooler has quit IRC | 14:05 | |
| *** anton has quit IRC | 14:06 | |
| *** anton has joined #tripleo | 14:09 | |
| *** anton has quit IRC | 14:09 | |
| *** anton has joined #tripleo | 14:10 | |
| *** anton has quit IRC | 14:12 | |
| *** anton has joined #tripleo | 14:12 | |
| *** anton has quit IRC | 14:14 | |
| *** anton has joined #tripleo | 14:15 | |
| *** jprovazn has quit IRC | 14:16 | |
| *** zoli|afk is now known as zoli|wfh | 14:17 | |
| *** jprovazn has joined #tripleo | 14:18 | |
| lucasagomes | trown, hi there... what happened with tripleo-quickstart/config/config_general/devmode.yml ? Was it replaced by minimal.yml ? | 14:19 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: simplify ARCH param for rhel/centos param can be x86_64 and amd64 https://review.openstack.org/392176 | 14:20 |
| *** ayoung has joined #tripleo | 14:23 | |
| lucasagomes | trown, nvm, I found it https://review.openstack.org/#/c/369571/ | 14:23 |
| *** jkilpatr_ has quit IRC | 14:31 | |
| *** jerrygb has joined #tripleo | 14:31 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-common: Do not try "manage" actions on nodes that are not in "enroll" state https://review.openstack.org/392179 | 14:33 |
| *** jerrygb_ has quit IRC | 14:33 | |
| openstackgerrit | Merged openstack/tripleo-ui: Include nested parameters in service parameters https://review.openstack.org/390946 | 14:36 |
| openstackgerrit | Sven Anderson proposed openstack/puppet-tripleo: Implement Nova ec2api profile https://review.openstack.org/336562 | 14:38 |
| ansiwen | EmilienM: can we move forward here? https://review.openstack.org/#/c/336562/ | 14:39 |
| *** jkilpatr has joined #tripleo | 14:40 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/instack-undercloud: Prevent pxe_ilo driver from guessing the boot mode to use https://review.openstack.org/392181 | 14:43 |
| dtantsur | lucasagomes, fyi ^^^ | 14:43 |
| lucasagomes | dtantsur, damn, yes! | 14:44 |
| openstackgerrit | Merged openstack/tripleo-quickstart: Stop using deprecated network range https://review.openstack.org/343443 | 14:45 |
| jaosorior | EmilienM: there seems to be nothing in the audit logs, only strange thing is this from httpd (note that cinder runs over it) type=ANOM_ABEND msg=audit(1478006873.887:1495): auid=4294967295 uid=166 gid=166 ses=4294967295 subj=system_u:system_r:httpd_t:s0 pid=7701 comm="httpd" reason="memory violation" sig=11 | 14:46 |
| EmilienM | ansiwen: will look shortly | 14:48 |
| EmilienM | jaosorior: I'll look after mtg | 14:48 |
| *** jerrygb_ has joined #tripleo | 15:00 | |
| *** jerrygb has quit IRC | 15:03 | |
| weshay | anyone try the containerized compute deployment? wondering if there is info on registering the atomic image to ironic | 15:03 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Parameters listing updates https://review.openstack.org/391855 | 15:04 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Use mistral action to create new containers https://review.openstack.org/392189 | 15:04 |
| jrist | jtomasek: can you also look at https://review.openstack.org/#/c/390854/ ? | 15:06 |
| jrist | we need it for validations | 15:06 |
| jrist | shadower: any progress on ^ | 15:06 |
| jrist | jtomasek: haha I was caught up :) | 15:07 |
| jrist | jtomasek: then two new ones within the last 3 minutes | 15:07 |
| * jtomasek looks | 15:07 | |
| shadower | jrist: I've just replied to steve's comment. Will update the patch & retest | 15:07 |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-common: Power off new nodes when making them available, not right after enrolling https://review.openstack.org/392148 | 15:13 |
| *** paramite has quit IRC | 15:17 | |
| *** ramishra has quit IRC | 15:17 | |
| jkilpatr | EmilienM, do you know why bulk introspection is being depreciated? Most docs/people still use it. | 15:18 |
| *** yamahata has joined #tripleo | 15:18 | |
| *** paramite has joined #tripleo | 15:19 | |
| *** ramishra has joined #tripleo | 15:19 | |
| jtomasek | d0ugal: how does the CLI upload files to swift plan? | 15:20 |
| jpich | jkilpatr: openstack overcloud node introspect is meant to replace the bulk introspection command, AFAIUI | 15:21 |
| *** akuznetsov has joined #tripleo | 15:21 | |
| d0ugal | jtomasek: https://github.com/openstack/python-tripleoclient/blob/master/tripleoclient/workflows/plan_management.py#L22-L35 | 15:22 |
| d0ugal | jtomasek: tl;dr - it creates a tarball, uploads it and swift extracts it | 15:22 |
| jkilpatr | jpich, I can see that from the message, but the node introspect command means you are responsible for running it against each node, sure you can hack together a quick bash forloop. or move all nodes to manageable and use the --all-manageable option but its not a drop in replacement. | 15:22 |
| *** athomas has quit IRC | 15:22 | |
| *** chandankumar has quit IRC | 15:23 | |
| jtomasek | d0ugal: oh, I see, ok | 15:23 |
| d0ugal | jkilpatr: Doesn't the previous command only introspect manageable nodes? | 15:24 |
| jpich | jkilpatr: the background for moving away from bulk introspection was described in http://lists.openstack.org/pipermail/openstack-dev/2016-May/095456.html | 15:24 |
| jpich | d0ugal: It moves available nodes back into manageable | 15:24 |
| *** paramite has quit IRC | 15:24 | |
| d0ugal | jpich: ah | 15:25 |
| jkilpatr | jpich, thanks I was looking for reasoning like this. | 15:25 |
| jpich | jkilpatr: Hope that helps | 15:26 |
| *** milan has quit IRC | 15:28 | |
| *** zaneb has quit IRC | 15:28 | |
| *** athomas has joined #tripleo | 15:29 | |
| openstackgerrit | Merged openstack/tripleo-ui: Fix deployment success output https://review.openstack.org/389270 | 15:29 |
| jkilpatr | jpich, yes it does. I was looking to add scale features to these commands, namely --retires N and --batch N, one would retry failed introspection the other would batch N to perform at once. | 15:30 |
| jkilpatr | currently once you top 20 or so nodes bulk introspection is hopeless and it starts being problematic long before that. | 15:30 |
| *** akuznetsov has quit IRC | 15:31 | |
| *** absubram has quit IRC | 15:31 | |
| openstackgerrit | Merged openstack/tripleo-ui: Improve parameters loading https://review.openstack.org/391813 | 15:32 |
| *** kjw3 has joined #tripleo | 15:33 | |
| *** dhill_ has quit IRC | 15:34 | |
| *** dhill_ has joined #tripleo | 15:34 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 15:35 |
| shardy | shadower: hey did you have a patch that builds a dynamic inventory for ansible (for overcloud nodes) ? | 15:35 |
| *** absubram has joined #tripleo | 15:36 | |
| jtomasek | honza: you can use BaseHttpRequestErrorHandler to try to handle stuff there but again, problem is as you mentioned - apis behave inconsistently | 15:37 |
| jtomasek | honza: ApiServices are supposed to only return promises for api calls, not to do the api error handling | 15:38 |
| jtomasek | honza: I know that this is not very DRY on the actions side, but I'd prefer to keep it that way until the apis we use stabilize and then we could extract actions which trigger api calls into redux middleware | 15:40 |
| *** zaneb has joined #tripleo | 15:40 | |
| *** paramite has joined #tripleo | 15:41 | |
| jpich | jkilpatr: dtantsur is probably a good person to bounce ideas off wrt improving the introspection workflow | 15:42 |
| *** bana_k has joined #tripleo | 15:42 | |
| * dtantsur hides | 15:42 | |
| dtantsur | jkilpatr, jpich, note that "introspection start" and "introspection bulk start" are completely different commands, coming from different projects | 15:43 |
| dtantsur | however, batches support is built into ironic-inspector, but iirc only enabled for ssh drivers | 15:43 |
| *** pcaruana has joined #tripleo | 15:44 | |
| *** aufi has quit IRC | 15:44 | |
| jkilpatr | dtantsur, I'm aware took the time monday to map out the entire call path for the bulk commands vs normal introspection | 15:46 |
| dtantsur | heh, I can imagine | 15:47 |
| jkilpatr | dtantsur, I used the entire whiteboard and then some. Anyways I'm in contact with the ironic team too, trying to decide at what point in the chain I want to implement this and see how each project feels about the idea. | 15:48 |
| jkilpatr | also have to keep from hiding new flaws on the CI level, but provide robust end user commands that actually work. | 15:48 |
| *** dhill_ has quit IRC | 15:50 | |
| *** Guest14381 has quit IRC | 15:51 | |
| *** akshai has quit IRC | 15:54 | |
| *** akshai has joined #tripleo | 15:56 | |
| shadower | shardy: https://github.com/openstack/tripleo-validations/blob/master/scripts/tripleo-ansible-inventory | 15:56 |
| shardy | shadower: thanks! | 15:57 |
| *** dhill_ has joined #tripleo | 16:00 | |
| openstackgerrit | Merged openstack/puppet-tripleo: Fixes transparent binding to OpenDaylight in HA Proxy https://review.openstack.org/392142 | 16:03 |
| *** owalsh_ has joined #tripleo | 16:04 | |
| *** owalsh has quit IRC | 16:04 | |
| *** rcernin has quit IRC | 16:06 | |
| openstackgerrit | Adriano Petrich proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/365449 | 16:06 |
| *** milan has joined #tripleo | 16:07 | |
| *** artom__ has joined #tripleo | 16:07 | |
| *** pcaruana has quit IRC | 16:07 | |
| *** owalsh_ is now known as owalsh | 16:08 | |
| *** artom_ has quit IRC | 16:08 | |
| *** artom__ is now known as artom | 16:09 | |
| *** myoung is now known as myoung|food | 16:11 | |
| openstackgerrit | Attila Darazs proposed openstack/tripleo-quickstart: Fix documentation for accessing overcloud nodes https://review.openstack.org/392220 | 16:11 |
| weshay | bah.. can't find the containerized compute experimental job, /me looking for logs.. | 16:13 |
| *** chandankumar has joined #tripleo | 16:13 | |
| weshay | I just kicked off experimental but it's not in the list | 16:13 |
| *** akshai has quit IRC | 16:13 | |
| *** ebarrera has quit IRC | 16:13 | |
| *** akshai has joined #tripleo | 16:17 | |
| *** jaosorior has quit IRC | 16:17 | |
| EmilienM | weshay: can someone look on https://review.openstack.org/#/c/391209/ ? | 16:18 |
| weshay | aye | 16:19 |
| *** trozet has quit IRC | 16:21 | |
| *** trozet has joined #tripleo | 16:21 | |
| *** bana_k has quit IRC | 16:27 | |
| honza | jtomasek: If the the API call fail due to authentication, we redirect to the login page. For all other outcomes, we resolve the promise and "pass it on" to the specific API class. | 16:35 |
| honza | jtomasek: it's kind of like middleware | 16:35 |
| jtomasek | honza: I understand, my concern is that it skips whole api error handling infrastructure which we already have in place and adds another logic that does it, in a place which is not meant to do that. | 16:38 |
| honza | jtomasek: there is only thing that needs to happen when you encounter a 401 | 16:38 |
| honza | jtomasek: *shrug* | 16:39 |
| jtomasek | honza: you will have to put there a logic to figure out what api service we were calling and handle the api error specific to that api as apis are inconsistent. this is exactly what ApiErrorHandlers do | 16:39 |
| jtomasek | honza: you mentioned that mistral is different | 16:39 |
| honza | jtomasek: yes, you're right | 16:39 |
| jtomasek | honza: in general all you'd need to do is add dispatch(loginActions.logoutUser) to case 401 in ApiErrorHandlers | 16:40 |
| honza | jtomasek: sounds good | 16:41 |
| *** jerrygb has joined #tripleo | 16:41 | |
| jtomasek | honza: only ptoblem for now is that we'll need to pass dispatch as parameter to ApiErrorHandler to be able to dispatch action from there. once we move this to middlewares we'll have dispatch available without doing it | 16:42 |
| *** mcornea has quit IRC | 16:42 | |
| *** jerrygb_ has quit IRC | 16:43 | |
| *** jerrygb_ has joined #tripleo | 16:44 | |
| *** jerrygb has quit IRC | 16:46 | |
| *** mlupton has joined #tripleo | 16:52 | |
| *** myoung|food is now known as myoung | 16:54 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 16:58 |
| *** bana_k has joined #tripleo | 16:59 | |
| *** akshai has quit IRC | 16:59 | |
| *** jpich has quit IRC | 17:01 | |
| openstackgerrit | Merged openstack/tripleo-ui: Save the result of direct action calls in Mistral https://review.openstack.org/382277 | 17:03 |
| *** mlupton has quit IRC | 17:06 | |
| *** aufi has joined #tripleo | 17:16 | |
| *** derekh has quit IRC | 17:18 | |
| *** jerrygb has joined #tripleo | 17:25 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Add Barbican to the overcloud https://review.openstack.org/352458 | 17:27 |
| *** mlupton has joined #tripleo | 17:27 | |
| *** jerrygb_ has quit IRC | 17:28 | |
| *** akshai has joined #tripleo | 17:29 | |
| *** Goneri has joined #tripleo | 17:36 | |
| *** egafford has quit IRC | 17:39 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 17:40 |
| *** shardy has quit IRC | 17:44 | |
| *** hewbrocca is now known as hewbrocca_afk | 17:46 | |
| *** akuznetsov has joined #tripleo | 17:46 | |
| *** zoli|wfh is now known as zoli|gone | 17:49 | |
| *** zoli|gone is now known as zoli_gone-proxy | 17:50 | |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-docs: [WIP] Mistral API Documentation https://review.openstack.org/358685 | 17:52 |
| d0ugal | tzumainn: ^ I have done some more work on the docs | 17:52 |
| d0ugal | tzumainn: not complete yet, but updated the commit message with things I still want to add (some of these might be better added in a follow up to avoid this turning into a book-in-a-patch) | 17:53 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 17:53 |
| *** jtomasek_ has quit IRC | 17:53 | |
| d0ugal | tzumainn: feedback would be much appreciated :) | 17:53 |
| tzumainn | d0ugal, okay, I'll take a look soon - thanks! | 17:54 |
| *** akshai has quit IRC | 17:55 | |
| *** egafford has joined #tripleo | 17:56 | |
| *** mgould is now known as mgould|afk | 18:00 | |
| *** athomas has quit IRC | 18:02 | |
| *** akuznetsov has quit IRC | 18:02 | |
| openstackgerrit | Lukas Bezdicka proposed openstack/tripleo-heat-templates: Ensure we update ceph and composable nodes https://review.openstack.org/392259 | 18:04 |
| *** akuznetsov has joined #tripleo | 18:05 | |
| openstackgerrit | Lukas Bezdicka proposed openstack/tripleo-heat-templates: Ensure we update ceph and composable nodes https://review.openstack.org/392260 | 18:06 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 18:07 |
| *** mhenkel has quit IRC | 18:07 | |
| tzumainn | d0ugal, left some comments, looks great overall! | 18:07 |
| *** lucasagomes is now known as lucas-afk | 18:09 | |
| *** dtantsur is now known as dtantsur|afk | 18:13 | |
| *** akshai has joined #tripleo | 18:15 | |
| *** aufi has quit IRC | 18:16 | |
| slagle | EmilienM: do we know what's wrong with the ovb-ha job? | 18:22 |
| EmilienM | slagle: I was investigating it | 18:22 |
| EmilienM | so this morning I woke up and saw the job mostly red everywhere | 18:22 |
| EmilienM | I think it's a bad promotion | 18:22 |
| EmilienM | slagle: I see 2 errors with ping test | 18:23 |
| EmilienM | they happen randomly | 18:23 |
| EmilienM | 1) pingtest stack timeout | 18:24 |
| EmilienM | 2) pingtest stack error when creating the Nova::Service from volume | 18:24 |
| EmilienM | Nova::Server sorry | 18:24 |
| *** abregman_ has quit IRC | 18:24 | |
| EmilienM | slagle: I looked in logstack and 2) started on October 27th | 18:24 |
| EmilienM | looking at logstash again, looks like 1) started yesterday | 18:25 |
| *** anton has quit IRC | 18:26 | |
| slagle | are there bugs filed for both issues? | 18:26 |
| EmilienM | nothing in logstash showed me that was due to a patch we merged | 18:26 |
| *** anton has joined #tripleo | 18:26 | |
| EmilienM | I think it's a promotion, and a bug in OpenStack somewhere | 18:26 |
| EmilienM | slagle: not yet, as I'm still debugging the what is failing | 18:26 |
| EmilienM | slagle: I'm creating a launchpad now | 18:27 |
| EmilienM | I also see failures on non-ha | 18:27 |
| EmilienM | ClientException: resources.volume1: Gateway Time-out (HTTP 504) | 18:28 |
| EmilienM | ok that's a 3) | 18:28 |
| EmilienM | they might be all related | 18:28 |
| EmilienM | all 3 are in pingtest, and related to nova/cinder somehow | 18:28 |
| EmilienM | I'm filing a single bug now | 18:28 |
| EmilienM | slagle: https://bugs.launchpad.net/tripleo/+bug/1638350 | 18:32 |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 18:32 |
| EmilienM | ovb-nonha looks pretty stable though | 18:36 |
| EmilienM | I saw some failures but not a lot | 18:36 |
| EmilienM | while ovb-ha is really in bad shape | 18:36 |
| slagle | it might be a heat/keystone error | 18:37 |
| slagle | looking at http://logs.openstack.org/86/381286/11/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/f17ac81/ | 18:37 |
| slagle | the heat-engine.log on controller 2 shows: | 18:38 |
| slagle | Domain admin client authentication f ailed | 18:38 |
| slagle | http://logs.openstack.org/86/381286/11/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/f17ac81/logs/overcloud-controller-2/var/log/heat/heat-engine.txt.gz#_2016-11-01_15_07_47_203 | 18:38 |
| *** akuznetsov has quit IRC | 18:39 | |
| EmilienM | Domain admin client authentication failed | 18:39 |
| EmilienM | indeed | 18:39 |
| EmilienM | did we have it before? /me checks in old logs | 18:39 |
| openstackgerrit | Ryan Hallisey proposed openstack/tripleo-heat-templates: Add a heat output describing where roles are located https://review.openstack.org/327822 | 18:40 |
| *** maeca1 has quit IRC | 18:40 | |
| *** maeca1 has joined #tripleo | 18:40 | |
| slagle | EmilienM: in the keystone.log: Could not find domain: heat_stack | 18:41 |
| EmilienM | I check 2 weeks ago, we didn't have this log | 18:42 |
| EmilienM | (in heat-engine) | 18:42 |
| EmilienM | slagle: link? | 18:42 |
| EmilienM | 2016-11-01 15:07:47.189 28358 WARNING keystone.auth.controllers [req-5e7fad87-3dfa-47e1-aa82-02b651d0e86d - - - - -] Could not find domain: heat_stack | 18:43 |
| slagle | EmilienM: http://logs.openstack.org/86/381286/11/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/f17ac81/logs/overcloud-controller-2/var/log/keystone/keystone.txt.gz#_2016-11-01_15_07_47_189 | 18:43 |
| EmilienM | ok I got it | 18:43 |
| EmilienM | I'm going to logstash it | 18:43 |
| slagle | at the same time as the error in heat-engine.log | 18:43 |
| EmilienM | how I'm afraid logstash won't find it | 18:44 |
| EmilienM | because it doesn't look in this file I think | 18:44 |
| EmilienM | damn | 18:45 |
| EmilienM | puppet should create the domain | 18:46 |
| *** egafford has quit IRC | 18:48 | |
| *** egafford has joined #tripleo | 18:49 | |
| *** pkovar has quit IRC | 18:55 | |
| *** nyechiel has joined #tripleo | 18:55 | |
| EmilienM | Nov 1 15:01:51 localhost os-collect-config: #033[mNotice: /Stage[main]/Heat::Keystone::Auth/Keystone_role[heat_stack_user]/ensure: created | 18:58 |
| EmilienM | 2016-11-01 15:07:47.189 28358 WARNING keystone.auth.controllers [req-5e7fad87-3dfa-47e1-aa82-02b651d0e86d - - - - -] Could not find domain: heat_stack | 18:59 |
| EmilienM | oh wait, heat_stack_user not heat_stack domain | 18:59 |
| *** chandankumar has quit IRC | 19:00 | |
| slagle | are we configuring heat with the wrong domain name? | 19:02 |
| EmilienM | perhaps | 19:02 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Undercloud upgrade for mitaka and newton https://review.openstack.org/381286 | 19:02 |
| EmilienM | we did some changes recently | 19:02 |
| EmilienM | me and Steve were fixing a bug about composable services when heat is separated from keystone | 19:03 |
| EmilienM | I'm looking into the patch again | 19:03 |
| slagle | well, this is strange | 19:03 |
| slagle | heat.conf on controller 0 is different from the others | 19:03 |
| slagle | controller 1 and 2 have: | 19:04 |
| slagle | stack_user_domain_name = heat_stack | 19:04 |
| slagle | stack_domain_admin = heat_stack_domain_admin | 19:04 |
| slagle | but controller 0 does not have those 2 lines | 19:04 |
| slagle | this might explain why the failure is transient, it depens on which heat-engine it gets picked up by in ha | 19:05 |
| EmilienM | stack_user_domain_name etc are missing | 19:05 |
| EmilienM | problem could be here: https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/keystone.pp#L154-L173 | 19:06 |
| EmilienM | mwhahaha: on top of your head, could you check that code? | 19:06 |
| EmilienM | specially https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/keystone.pp#L159 | 19:06 |
| *** Goneri has quit IRC | 19:06 | |
| mwhahaha | hmm | 19:07 |
| EmilienM | launchpad 1631130 | 19:08 |
| openstack | Launchpad bug 1631130 in tripleo "Deployment fails when Keystone and Heat services are deployed on different roles nodes" [High,Fix released] https://launchpad.net/bugs/1631130 - Assigned to Emilien Macchi (emilienm) | 19:08 |
| EmilienM | so we tried to fix that ^ | 19:08 |
| mwhahaha | might be an ordering thing | 19:08 |
| EmilienM | so we did two things: https://review.openstack.org/#/c/386118/ and https://review.openstack.org/#/c/386115/ | 19:09 |
| *** ooolpbot has joined #tripleo | 19:10 | |
| ooolpbot | URGENT TRIPLEO TASKS NEED ATTENTION | 19:10 |
| ooolpbot | https://bugs.launchpad.net/tripleo/+bug/1638350 | 19:10 |
| *** ooolpbot has quit IRC | 19:10 | |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 19:10 |
| mwhahaha | EmilienM: so the os_service_default would result in the missing config items, so i guess the question is when does the base heat profile get included | 19:11 |
| EmilienM | mwhahaha: in CI, we shouldn't be in the "else" | 19:12 |
| EmilienM | since keystone & heat are collocated | 19:12 |
| mwhahaha | but it's a parse order thing | 19:12 |
| mwhahaha | or at least it could be | 19:13 |
| mwhahaha | EmilienM: Nov 1 15:01:51 localhost os-collect-config: #033[mNotice: /Stage[main]/Heat::Keystone::Domain/Heat_config[DEFAULT/stack_user_domain_name]/ensure: removed#033[0m | 19:15 |
| mwhahaha | it got removed after it was created | 19:15 |
| mwhahaha | http://logs.openstack.org/86/381286/11/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/f17ac81/logs/overcloud-controller-0/var/log/messages | 19:15 |
| mwhahaha | Nov 1 14:29:22 localhost os-collect-config: #033[mNotice: /Stage[main]/Heat::Keystone::Domain/Heat_config[DEFAULT/stack_domain_admin]/ensure: created#033[0m | 19:16 |
| EmilienM | wtf | 19:16 |
| mwhahaha | it was removed 30 mins later on a subseqent run | 19:16 |
| mwhahaha | so we do hit that code or hit something else that removed it | 19:16 |
| EmilienM | so created at step 3 | 19:17 |
| EmilienM | and removed at step 5 | 19:18 |
| mwhahaha | so here's the thing | 19:19 |
| EmilienM | so we have 2 problems, it should be created at step 5 (it's created earlier) and it's removed at step 5 (while it should be created) | 19:19 |
| mwhahaha | tipleo::profile::base::heat doest not include heat::ekystone::domain for any steps except 3 and 4 | 19:19 |
| *** egafford has quit IRC | 19:20 | |
| mwhahaha | https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/heat.pp#L56-L62 | 19:20 |
| mwhahaha | so we will always be in that else | 19:20 |
| mwhahaha | er that else would never conflict | 19:20 |
| mwhahaha | this code makes my brain hurt | 19:21 |
| EmilienM | well no | 19:21 |
| EmilienM | we check if tripleo::profile::base::heat is defined | 19:21 |
| EmilienM | which should be the case | 19:21 |
| EmilienM | mwhahaha: sorry, I wrote that thing :( | 19:22 |
| mwhahaha | EmilienM: i'm just thinking because of the include vs class | 19:22 |
| EmilienM | the include should create the resource since manage_* are true in puppet-heat by defeault | 19:23 |
| mwhahaha | if tripleo::profile::base::heat actually was already defined and it included class { '::heat::keystone::domain': } the include would be a noop in profile base keystone | 19:23 |
| mwhahaha | we really need to just do the user creation seperate from the configuration update | 19:24 |
| mwhahaha | and make the code right | 19:24 |
| EmilienM | mwhahaha: could it be an ordering issue where the class is defined *after* keystone? | 19:24 |
| mwhahaha | EmilienM: yes | 19:24 |
| EmilienM | so if defined doesn't work | 19:24 |
| EmilienM | ok, so I missed that part | 19:25 |
| mwhahaha | if defined is dangerous | 19:25 |
| EmilienM | my bad :( | 19:25 |
| mwhahaha | we should really not use it for anything | 19:25 |
| mwhahaha | because it is dependent on catalog ordering | 19:25 |
| EmilienM | mwhahaha: safest way would be to create new parameters | 19:25 |
| mwhahaha | and since our catalogs may be dynamic in nature it's hard to ensure ordering | 19:25 |
| EmilienM | right | 19:25 |
| EmilienM | in fact, I was just looking for a way to detect if heat & keystone were collocated | 19:26 |
| mwhahaha | user creation should be it's own class and we should just include them all with keystone | 19:26 |
| mwhahaha | the configuration should be done seperately | 19:26 |
| EmilienM | right but we only need this user etc if we have heat deployed in the cloud | 19:27 |
| mwhahaha | short term, just add a manage_config to heat::keystone::domain | 19:27 |
| mwhahaha | long term, user creation should be it's own class | 19:27 |
| mwhahaha | similar to heat::keystone::auth | 19:28 |
| *** bnemec has joined #tripleo | 19:28 | |
| mwhahaha | or just disable the manage_* and do that in tripleo | 19:28 |
| mwhahaha | -^ probably more realistic | 19:28 |
| EmilienM | mwhahaha: outside puppet you mean? | 19:29 |
| mwhahaha | no do it in puppet, but do it in tripleo classes | 19:29 |
| EmilienM | ah | 19:29 |
| EmilienM | makes sense | 19:29 |
| mwhahaha | so there's no reason we can't just do a keyston_user in tripleo profiles | 19:29 |
| EmilienM | mwhahaha: in our case, how do you detect that both keystone & heat are collocated though? | 19:30 |
| *** dsariel has quit IRC | 19:30 | |
| mwhahaha | who cares | 19:30 |
| mwhahaha | just disable the user creation in heat | 19:30 |
| mwhahaha | and create the user in the keystone profile | 19:30 |
| mwhahaha | if heat is enabled | 19:31 |
| mwhahaha | why does it matter if they are collocated? | 19:31 |
| mwhahaha | the goal was to not leak the password right? | 19:31 |
| EmilienM | mwhahaha: yes | 19:31 |
| mwhahaha | if you disable user creation, there's no need to call keystone::heat::domain in the keystone profile | 19:32 |
| mwhahaha | sec i'll propose patch | 19:32 |
| *** ohamada has quit IRC | 19:33 | |
| *** egafford has joined #tripleo | 19:33 | |
| *** paramite has quit IRC | 19:39 | |
| *** rcernin has joined #tripleo | 19:41 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: gnocchi statsd should be able to send data to port 8125 https://review.openstack.org/392287 | 19:43 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-tripleo: Create heat user in keystone profile https://review.openstack.org/392288 | 19:44 |
| mwhahaha | EmilienM: -^ | 19:44 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-tripleo: Create heat user in keystone profile https://review.openstack.org/392288 | 19:45 |
| EmilienM | mwhahaha: looking | 19:46 |
| mwhahaha | EmilienM: so either something like that or we add in a manage_config to heat::keystone::domain and use that instead. I'm not sure which is best | 19:47 |
| EmilienM | mwhahaha: mhh this thing works for me | 19:50 |
| *** mlupton has quit IRC | 19:51 | |
| EmilienM | config is handled at step 3 and 4 anyway, right? | 19:51 |
| mwhahaha | yea | 19:51 |
| EmilienM | actually I'm wondering if we should remove this thing and configure it up to step 3 and more | 19:51 |
| mwhahaha | but shouldn't we be configuring the user before starting the services? | 19:51 |
| mwhahaha | :D | 19:51 |
| EmilienM | true, but it should not be a big deal until we actually use heat | 19:52 |
| EmilienM | in your patch, I would also change manifests/profile/base/heat.pp | 19:52 |
| EmilienM | to configure declare ::heat::keystone::domain in "if $step >= 4" maybe | 19:53 |
| EmilienM | it sounds dangerous to me to ignore it at step 5 | 19:53 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 19:54 |
| EmilienM | mwhahaha: wdyt? | 19:56 |
| mwhahaha | yea | 19:57 |
| *** Goneri has joined #tripleo | 19:57 | |
| mwhahaha | EmilienM: so there's also https://review.openstack.org/#/c/392292/ | 19:57 |
| mwhahaha | if we want to go that rought | 19:57 |
| mwhahaha | er route | 19:57 |
| mwhahaha | i think that might be cleaner but not sure the scope of this change (if it needs backporting) | 19:58 |
| EmilienM | mwhahaha: I prefer puppet-heat patch for longterm support | 19:59 |
| EmilienM | mwhahaha: I think logic should go in upstream modules | 19:59 |
| EmilienM | but that's me | 19:59 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 20:00 |
| mwhahaha | so here's the sticking point | 20:00 |
| mwhahaha | when heat & keystone are combined you have to know | 20:00 |
| mwhahaha | since it's the same class we can't use them together | 20:00 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 20:00 |
| mwhahaha | unless we calculate the manage_* flags via hiera and know that heat and keystone are combined | 20:00 |
| EmilienM | mwhahaha: tc meeting, brb | 20:02 |
| *** mlupton has joined #tripleo | 20:05 | |
| *** florianf has quit IRC | 20:06 | |
| *** ooolpbot has joined #tripleo | 20:10 | |
| ooolpbot | URGENT TRIPLEO TASKS NEED ATTENTION | 20:10 |
| ooolpbot | https://bugs.launchpad.net/tripleo/+bug/1638350 | 20:10 |
| *** ooolpbot has quit IRC | 20:10 | |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 20:10 |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: gnocchi statsd should be able to send data to port 8125 https://review.openstack.org/392287 | 20:11 |
| *** dprince has quit IRC | 20:12 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add element cloud-init-allow-password-authentication https://review.openstack.org/391765 | 20:13 |
| *** nyechiel has quit IRC | 20:17 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add support for ubuntu and InfiniBand indhcp-all-interfaces https://review.openstack.org/391772 | 20:18 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: add support for SUSE in dhcp-all-interfaces https://review.openstack.org/391772 | 20:19 |
| *** jayg is now known as jayg|g0n3 | 20:19 | |
| *** maeca1 has quit IRC | 20:20 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Perform package install outside of debootstrap https://review.openstack.org/392253 | 20:21 |
| openstackgerrit | wes hayutin proposed openstack/tripleo-specs: update the index for ocata https://review.openstack.org/388208 | 20:21 |
| *** yamahata has quit IRC | 20:22 | |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: simplify ARCH param for rhel/centos param can be x86_64 and amd64 https://review.openstack.org/392176 | 20:23 |
| *** paramite has joined #tripleo | 20:23 | |
| *** nyechiel has joined #tripleo | 20:27 | |
| openstackgerrit | Alex Schultz proposed openstack/puppet-tripleo: Create heat user in keystone profile https://review.openstack.org/392288 | 20:32 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: simplify ARCH param for rhel/centos param can be x86_64 and amd64 https://review.openstack.org/392176 | 20:34 |
| openstackgerrit | Noam Angel proposed openstack/diskimage-builder: redhat-common add option to select networking service NetworkManager/network https://review.openstack.org/392170 | 20:35 |
| openstackgerrit | Ana Krivokapic proposed openstack/python-tripleoclient: Remove endpoints creation from post-deployment config https://review.openstack.org/392302 | 20:38 |
| *** egafford has quit IRC | 20:38 | |
| *** nyechiel has quit IRC | 20:40 | |
| *** rhallisey has quit IRC | 20:40 | |
| *** jkilpatr has quit IRC | 20:41 | |
| *** jprovazn has quit IRC | 20:42 | |
| *** mlupton has quit IRC | 20:47 | |
| *** oneswig has joined #tripleo | 20:54 | |
| openstackgerrit | Ana Krivokapic proposed openstack/python-tripleoclient: Remove endpoints creation from post-deployment config https://review.openstack.org/392302 | 20:55 |
| EmilienM | mwhahaha: back! | 20:56 |
| mwhahaha | k waiting to see if ci passes (after i fixed the user role instead of role) | 20:56 |
| EmilienM | mwhahaha: I guess your first solution with puppet-tripleo is simplier | 20:57 |
| EmilienM | https://review.openstack.org/#/c/392288/ | 20:57 |
| EmilienM | you also changed the step | 20:57 |
| EmilienM | mwhahaha: should we make heat_admin_password required in hiera? | 20:58 |
| EmilienM | you put 'changeme' if undef | 20:58 |
| mwhahaha | well that's the default of the puppet-heat class | 20:59 |
| mwhahaha | so if it's not in hiera it gets changeme | 20:59 |
| EmilienM | oh well, default in puppet-heat is same | 20:59 |
| EmilienM | right | 20:59 |
| mwhahaha | we probably should make it required :D but i want to see if this works first | 20:59 |
| EmilienM | mwhahaha: do we need Class['::keystone::roles::admin'] before creating the domain/user? | 20:59 |
| mwhahaha | no just the user role i believe | 20:59 |
| EmilienM | k | 21:00 |
| *** paramite has quit IRC | 21:00 | |
| *** dsariel has joined #tripleo | 21:00 | |
| dtrainor | When I have a deployment that fails because I have a Controller that runs in to the infamous "No valid host was found" error, will that leave any other node that depends on a Controller, like a Compute, in the CREATE_ABORTED status? Is that what aborts the Compute deploy, or would I have seen a separate error for the Compute deployment failure? | 21:01 |
| dtrainor | The reason I ask is because I'm looking at the CREATE_ABORTED status of this Compute node, and the Compute resource has a required_by of a pretty inclusive list of resources | 21:02 |
| openstackgerrit | Ana Krivokapic proposed openstack/python-tripleoclient: Remove endpoints creation from post-deployment config https://review.openstack.org/392302 | 21:05 |
| EmilienM | slagle: can you add https://review.openstack.org/#/c/392288 in your backlog? if it's loog to you, I'll approve it tonight if CI pass consistently and if heat config is good again on all controller nodes. | 21:06 |
| EmilienM | s/loog/looks good/ - this is me re-inventing english | 21:07 |
| slagle | EmilienM: honestly, i could only review that for syntax | 21:08 |
| slagle | EmilienM: maybe capture what it fixes somewhere | 21:08 |
| slagle | i dont understand what the root cause ended up being | 21:09 |
| slagle | were we creating the user, then deleting it? | 21:09 |
| *** ooolpbot has joined #tripleo | 21:10 | |
| ooolpbot | URGENT TRIPLEO TASKS NEED ATTENTION | 21:10 |
| ooolpbot | https://bugs.launchpad.net/tripleo/+bug/1638350 | 21:10 |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 21:10 |
| *** ooolpbot has quit IRC | 21:10 | |
| EmilienM | slagle: I made a mistake by using "if Defined" in Puppet which is a bad thing. We're having an orchestration issue where users is randomly managed | 21:12 |
| *** absubram has quit IRC | 21:12 | |
| *** morazi has quit IRC | 21:13 | |
| EmilienM | slagle: what mwhahaha does in https://review.openstack.org/#/c/392288/ is to separate heat config and keystone resource in a better way I did | 21:13 |
| *** Goneri has quit IRC | 21:14 | |
| *** mlupton has joined #tripleo | 21:18 | |
| *** mlupton has quit IRC | 21:25 | |
| openstackgerrit | James Slagle proposed openstack/python-tripleoclient: Fix nodes count check on stack-update with custom roles https://review.openstack.org/392313 | 21:25 |
| openstackgerrit | James Slagle proposed openstack/python-tripleoclient: Fix nodes count check on stack-update with custom roles https://review.openstack.org/392313 | 21:26 |
| *** mlupton has joined #tripleo | 21:27 | |
| *** mlupton has quit IRC | 21:27 | |
| *** mlupton has joined #tripleo | 21:29 | |
| *** abregman has joined #tripleo | 21:29 | |
| *** lblanchard has quit IRC | 21:31 | |
| *** akrivoka has quit IRC | 21:31 | |
| openstackgerrit | James Slagle proposed openstack/python-tripleoclient: Fix nodes count check on stack-update with custom roles https://review.openstack.org/392313 | 21:38 |
| *** isq has joined #tripleo | 21:41 | |
| *** jerrygb has quit IRC | 21:46 | |
| *** abregman has quit IRC | 21:49 | |
| *** jcoufal has quit IRC | 21:52 | |
| *** jkilpatr has joined #tripleo | 21:55 | |
| *** thrash is now known as thrash|g0ne | 21:57 | |
| *** jerrygb has joined #tripleo | 22:04 | |
| *** oneswig has quit IRC | 22:05 | |
| *** akshai has quit IRC | 22:05 | |
| openstackgerrit | Dan Sneddon proposed openstack/tripleo-heat-templates: Add Neutron network type and VLAN ranges to network-environment.yaml https://review.openstack.org/392325 | 22:05 |
| *** jerrygb has quit IRC | 22:06 | |
| tdasilva | EmilienM: hi! wondering if you could help me out figure out something about swift configuration in tripleo. in my overcloud deployment i see the option object_post_as_copy = false | 22:06 |
| tdasilva | but here https://github.com/openstack/puppet-swift/blob/master/manifests/proxy/copy.pp it is set as true | 22:07 |
| tdasilva | where else could it be set as false? | 22:07 |
| *** mlupton has quit IRC | 22:07 | |
| tdasilva | EmilienM: the other place I looked was here: https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/services/swift-proxy.yaml but it doesn't list any copy options there at all | 22:09 |
| *** ooolpbot has joined #tripleo | 22:10 | |
| ooolpbot | URGENT TRIPLEO TASKS NEED ATTENTION | 22:10 |
| ooolpbot | https://bugs.launchpad.net/tripleo/+bug/1638350 | 22:10 |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 22:10 |
| *** ooolpbot has quit IRC | 22:10 | |
| *** mlupton has joined #tripleo | 22:20 | |
| openstackgerrit | Merged openstack/diskimage-builder: elements: simple-init: Remove SUSE interfaces https://review.openstack.org/388158 | 22:20 |
| *** rcernin has quit IRC | 22:31 | |
| *** yamahata has joined #tripleo | 22:32 | |
| *** dsariel has quit IRC | 22:32 | |
| mwhahaha | tdasilva: is that the default in the packaging maybe? | 22:32 |
| mwhahaha | tdasilva: cause i don't see a reference to swift::proxy::copy anywhere in the heat templates | 22:33 |
| tdasilva | mwhahaha: the default is set to true in swift, but i just found this https://github.com/openstack/puppet-swift/blob/master/spec/classes/swift_proxy_copy_spec.rb | 22:37 |
| mwhahaha | tdasilva: that's just a unit test | 22:37 |
| EmilienM | I'm back | 22:37 |
| EmilienM | mwhahaha: looks like ha job failed | 22:37 |
| tdasilva | mwhahaha: lol, as you can see i know nothing about puppet | 22:38 |
| mwhahaha | EmilienM: 503 for something, looking | 22:38 |
| EmilienM | http://logs.openstack.org/88/392288/3/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/7db9856/console.html#_2016-11-01_22_13_43_986545 | 22:38 |
| EmilienM | 401 Unauthorized: This server could not verify that you are authorized to access the document you requested. Either you supplied the wrong credentials (e.g., bad password), or your browser does not understand how to supply the credentials required. (HTTP 401) | 22:38 |
| EmilienM | during glance image upload | 22:38 |
| mwhahaha | EmilienM: what creds are used for the glance upload? | 22:39 |
| *** jkilpatr has quit IRC | 22:40 | |
| *** jkilpatr has joined #tripleo | 22:40 | |
| tdasilva | mwhahaha: oh, sorry, you said packaging and I understood upstream source. so yeah, the default in upstream source is true, but it is possible that the packaging changes that to false | 22:41 |
| mwhahaha | tdasilva: https://github.com/rdo-packages/swift-distgit/blob/rpm-master/proxy-server.conf#L82 | 22:43 |
| tdasilva | mwhahaha: perfect, thanks! | 22:43 |
| mwhahaha | tdasilva: packaging sets it as false and tripleo never corrects it. you could do a custom template to trigger a swift::proxy::copy invocation for your deploy if you want it true | 22:43 |
| tdasilva | mwhahaha: suppose i wanted to add another option to the copy filter? how would i got about it? | 22:44 |
| mwhahaha | tdasilva: as an extra thing or by default? I'm not a swift expert so what exactly do you need to add? | 22:45 |
| tdasilva | mwhahaha: well..i was just using copy as an example, but what i need to add is an option to staticweb middleware for this bug: https://bugzilla.redhat.com/show_bug.cgi?id=1383305 | 22:47 |
| openstack | bugzilla.redhat.com bug 1383305 in openstack-tripleo "Please configure the url_base /etc/swift/proxy-server.conf" [Unspecified,Assigned] - Assigned to thiago | 22:47 |
| mwhahaha | and i just had a conversation about swift configs today heh | 22:47 |
| tdasilva | so i tried to add that here: https://github.com/openstack/puppet-swift/blob/master/manifests/proxy/staticweb.pp and here https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/services/swift-proxy.yaml | 22:48 |
| mwhahaha | so for tripleo, that means update puppet-swift to do whatever you want and then include it in tripleo | 22:48 |
| mwhahaha | tdasilva: if you're looking at adding something to the staicweb filter then yea that'd be the right spot | 22:49 |
| *** [1]cdearborn has joined #tripleo | 22:50 | |
| tdasilva | mwhahaha: the way i went about it was to hack directly in my undercloud and do a overcloud deploy to test it, but i don't see the changes in the overcloud once it is deployed | 22:50 |
| mwhahaha | tdasilva: did you build the overcloud images with your change included? | 22:51 |
| *** panda|pto|PH is now known as panda|Zz | 22:51 | |
| mwhahaha | there's some environmental vars you can use to do that but that's kinda outside my expertise | 22:52 |
| tdasilva | mwhahaha: no...i just made the changes directly to /etc/puppet and /usr/share/openstack-tripleo-heat-templates/ | 22:53 |
| mwhahaha | tdasilva: i think what you need is to add the url_base to swift::proxy::staticweb, and update the THT with hieradata to set the url_base | 22:53 |
| mwhahaha | tdasilva: after you deployed or before? i just know to get those puppet changes in you have to include them when you build the overcloud images | 22:53 |
| tdasilva | mwhahaha: yeah, i think that's what i did, just wanted to actually test it out before sending git-review | 22:53 |
| tdasilva | before i deployed, but i didn't do any image rebuilding | 22:54 |
| mwhahaha | yea that's probably why it didn't end up on the overcloud | 22:54 |
| tdasilva | mwhahaha: just did this: http://tripleo.org/basic_deployment/basic_deployment_cli.html#redeploy-the-overcloud | 22:55 |
| mwhahaha | tdasilva: http://hardysteven.blogspot.com/2016/08/tripleo-deploy-artifacts-and-puppet.html | 22:55 |
| tdasilva | mwhahaha: cool, i will check that out! | 22:56 |
| mwhahaha | good luck, if you need help with the puppet-swift updates just ask over in #puppet-openstack | 22:56 |
| EmilienM | mwhahaha: overcloudrc iirc | 23:02 |
| EmilienM | I'm doing recheck | 23:02 |
| *** cdearborn has quit IRC | 23:03 | |
| mwhahaha | EmilienM: unrecognized fernet token or something | 23:03 |
| EmilienM | mwhahaha: I'm looking heat config now | 23:04 |
| *** mlupton has quit IRC | 23:04 | |
| EmilienM | params are good !! | 23:04 |
| EmilienM | on controller0 and 1 | 23:04 |
| EmilienM | so it seems like we have good config now | 23:04 |
| mwhahaha | EmilienM: http://logs.openstack.org/88/392288/3/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/7db9856/logs/overcloud-controller-2/var/log/glance/api.txt.gz | 23:04 |
| mwhahaha | looks like we have some floating keystone issue or something | 23:05 |
| *** dmacpher has quit IRC | 23:05 | |
| mwhahaha | EmilienM: looks like glance is doing v2.0 auth and other stuff is doing v3, is this expected? | 23:07 |
| EmilienM | mwhahaha: no idea | 23:09 |
| EmilienM | looks like another issue (again) | 23:09 |
| mwhahaha | oh this must be for the glance middleware | 23:09 |
| mwhahaha | cause it's glance posting for a token | 23:09 |
| *** ooolpbot has joined #tripleo | 23:10 | |
| ooolpbot | URGENT TRIPLEO TASKS NEED ATTENTION | 23:10 |
| ooolpbot | https://bugs.launchpad.net/tripleo/+bug/1638350 | 23:10 |
| openstack | Launchpad bug 1638350 in tripleo "pingtest failing on OVB jobs to create Cinder volume and Nova server" [Critical,In progress] - Assigned to Emilien Macchi (emilienm) | 23:10 |
| *** ooolpbot has quit IRC | 23:10 | |
| mwhahaha | there's a whole bunch of 401s and 404s to /v3/auth/tokens on controller-0 | 23:10 |
| mwhahaha | meh these ligsd | 23:10 |
| mwhahaha | s/ligsd/logs | 23:10 |
| mwhahaha | EmilienM: aren't the fernet keys supposed to be the same on each controller? | 23:15 |
| mwhahaha | cause they aren't | 23:15 |
| *** dmacpher has joined #tripleo | 23:18 | |
| mwhahaha | EmilienM: we enabled fernet keys by default in puppet-keystone, are we not properly configuring fernet in tripleo? | 23:21 |
| * mwhahaha facepalms | 23:23 | |
| EmilienM | mwhahaha: what??? | 23:26 |
| mwhahaha | so we enabled fernet by default in puppet-keystone which is fine for single node controllers. I thought tripleo handled fernet key generation. | 23:27 |
| mwhahaha | it doesn't | 23:27 |
| EmilienM | yes we do | 23:27 |
| mwhahaha | not in the hieradata | 23:27 |
| EmilienM | we generate with tripleoclient | 23:27 |
| mwhahaha | there are no fernet keys | 23:27 |
| EmilienM | I wrote that | 23:27 |
| mwhahaha | its not working | 23:27 |
| EmilienM | it was working | 23:27 |
| EmilienM | I can show you logs of when it worked, keys were same | 23:27 |
| mwhahaha | well it doesn't work anymore :D | 23:27 |
| mwhahaha | all i've got is credential keys in the heiradata | 23:27 |
| mwhahaha | no fernet keys | 23:28 |
| mwhahaha | and the fernet keys are all different | 23:28 |
| mwhahaha | on the nodes | 23:28 |
| mwhahaha | http://paste.openstack.org/show/587591/ | 23:28 |
| EmilienM | wait | 23:28 |
| EmilienM | I'm stupid | 23:28 |
| EmilienM | I did credentials work | 23:29 |
| EmilienM | not fernets | 23:29 |
| EmilienM | ok I need to sleep | 23:29 |
| mwhahaha | k so i proposed puppet-keystone revert | 23:29 |
| EmilienM | I'm approving your revert... | 23:29 |
| EmilienM | if you don't mind | 23:29 |
| mwhahaha | that's fine | 23:29 |
| EmilienM | k | 23:29 |
| EmilienM | ayoung did work for fernet | 23:29 |
| EmilienM | but seems like it's not finished | 23:29 |
| * mwhahaha gets the bus out and prepares to drive over the world | 23:29 | |
| ayoung | For multi node controllers we can't just generate keys in the Keystone server | 23:30 |
| EmilienM | mwhahaha: https://review.openstack.org/#/c/383956/ | 23:30 |
| ayoung | the Tripleo approach is to generate them on the undercloud and push them out to the overcloud | 23:30 |
| mwhahaha | ayoung: yea the problem is i thought we were already doing that | 23:31 |
| ayoung | not yet | 23:31 |
| mwhahaha | it's ocata, that stuff landed in mitaka :D | 23:31 |
| ayoung | mwhahaha, I really don;t know how to solve that problem in the general Puppet case | 23:31 |
| ayoung | not the puppet changes they didn't | 23:31 |
| mwhahaha | ayoung: they did today, cause i thought tripleo was doing fernet key generation in tripleo client already | 23:32 |
| ayoung | sorry | 23:32 |
| mwhahaha | ayoung: there's the disconnect and we gate the puppet-keystone a single controller so we didn't know | 23:32 |
| ayoung | mwhahaha, looking for a spec that might be relevant one sec | 23:32 |
| *** cwolferh has quit IRC | 23:32 | |
| ayoung | http://git.openstack.org/cgit/openstack/keystone-specs/tree/specs/keystone/ocata/fernet-key-store.rst | 23:33 |
| EmilienM | why don't we do like what I did for credentials for now? | 23:33 |
| ayoung | the goal there is to use an external key manager | 23:33 |
| EmilienM | I thought we agreed on that ayoung ? | 23:33 |
| ayoung | EmilienM, yeah, and I thought I posted that... | 23:33 |
| ayoung | https://review.openstack.org/#/c/371247/ | 23:33 |
| EmilienM | ocata is super short, we won't make your keystone blueprint + tripleo integration | 23:33 |
| mwhahaha | ayoung: can we just do an MVP to get on fernet and work on expanding that later like everyone else? | 23:34 |
| ayoung | MVP? | 23:34 |
| mwhahaha | minimal viable product | 23:34 |
| mwhahaha | aka do the simple stuff first and expand | 23:34 |
| ayoung | ah, I thought you meant most valuable player. We have one of those. It is EmilienM | 23:34 |
| mwhahaha | do like we do for credentials and then expand to add rotation and such | 23:34 |
| ayoung | mwhahaha, so, yeah, that review should be the basis. What happened today? | 23:35 |
| mwhahaha | ayoung: we landed fernet by default in puppet-keystone | 23:35 |
| mwhahaha | and since we're not managing fernet keys in tripleo at all, it blew up | 23:35 |
| ayoung | and what happened? THings break? | 23:35 |
| mwhahaha | yea cause you can't auth on a multiple controller cluster when all the keys are different | 23:35 |
| mwhahaha | tokens only work 1/3 the time | 23:35 |
| ayoung | well, that should be a solvable problem | 23:36 |
| ayoung | mwhahaha, I think we have things out of order | 23:36 |
| ayoung | before we can make Fernet default, we need synchronized keys | 23:36 |
| mwhahaha | ayoung: but we have a method for doing that via tripleoclient | 23:36 |
| ayoung | Welll, sort of. | 23:37 |
| mwhahaha | why is that not in place already is kinda my question | 23:37 |
| ayoung | Were we using it? | 23:37 |
| mwhahaha | no | 23:37 |
| mwhahaha | that's how we got here | 23:37 |
| ayoung | mwhahaha, maybe cuz I didn;'t write it yet | 23:37 |
| ayoung | let me see... | 23:37 |
| mwhahaha | sorry i'm confused cause this was stuff that landed elsewhere in mitaka | 23:38 |
| ayoung | https://review.openstack.org/#/c/383956/ | 23:38 |
| mwhahaha | so i wasn't aware tripleo was lacking this bit | 23:38 |
| ayoung | Hmmm that looks wrongish | 23:38 |
| mwhahaha | also because we dont' explicitly define uuid, we got nailed when upstream changed | 23:38 |
| ayoung | Ah | 23:38 |
| ayoung | OK so we need that one one...is it sufficient, though... | 23:39 |
| ayoung | I seem to recall needing a tripleoclient change too | 23:39 |
| mwhahaha | anyway we can solve this tomorrow | 23:39 |
| mwhahaha | now we know that tripleo doesn't do fernet :D | 23:39 |
| ayoung | EmilienM, when you did credentials, did that require a python-tripleoclient change, or was that just something else I am misrememberating? | 23:40 |
| * mwhahaha wanders off for a bit | 23:40 | |
| EmilienM | ayoung: tripleoclient | 23:41 |
| EmilienM | I have to go now or she kills me and we don't want that | 23:41 |
| openstackgerrit | James Slagle proposed openstack/python-tripleoclient: Fix nodes count check on stack-update with custom roles https://review.openstack.org/392313 | 23:44 |
| *** kjw3 has quit IRC | 23:47 | |
| *** dmacpher has quit IRC | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!