| *** panda has quit IRC | 00:06 | |
| *** pradk has quit IRC | 00:06 | |
| *** panda has joined #tripleo | 00:06 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Prune old branches when updating cache https://review.openstack.org/266086 | 00:32 |
|---|---|---|
| *** shakamunyi has quit IRC | 00:35 | |
| *** shakamunyi has joined #tripleo | 00:35 | |
| *** dmacpher-afk has quit IRC | 00:48 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: WIP - Deploy Aodh services, replacing Ceilometer Alarm https://review.openstack.org/241408 | 00:54 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: WIP - Deploy Aodh services, replacing Ceilometer Alarm https://review.openstack.org/241408 | 00:55 |
| openstackgerrit | Merged openstack/diskimage-builder: deploy-ironic: Fix syntax error when checking for root device hints https://review.openstack.org/264736 | 00:55 |
| *** barra204 has joined #tripleo | 00:56 | |
| *** shakamunyi has quit IRC | 00:57 | |
| *** eil397 has quit IRC | 01:00 | |
| *** eil397 has joined #tripleo | 01:05 | |
| *** alop has quit IRC | 01:09 | |
| *** barra204 has quit IRC | 01:14 | |
| *** shakamunyi has joined #tripleo | 01:15 | |
| *** trozet has joined #tripleo | 01:27 | |
| *** barra204 has joined #tripleo | 01:31 | |
| *** shakamunyi has quit IRC | 01:32 | |
| *** shakamunyi has joined #tripleo | 01:36 | |
| *** dmacpher has joined #tripleo | 01:36 | |
| *** trozet has quit IRC | 01:37 | |
| *** barra204 has quit IRC | 01:38 | |
| *** barra204 has joined #tripleo | 01:38 | |
| *** shakamunyi has quit IRC | 01:39 | |
| *** thrash is now known as thrash|g0ne | 01:39 | |
| *** shakamunyi has joined #tripleo | 01:44 | |
| *** barra204 has quit IRC | 01:45 | |
| *** shakamunyi has quit IRC | 01:48 | |
| *** shakamunyi has joined #tripleo | 01:48 | |
| *** shakamunyi has quit IRC | 01:54 | |
| *** rbrady has quit IRC | 01:58 | |
| *** eil397 has quit IRC | 02:07 | |
| *** shakamunyi has joined #tripleo | 02:08 | |
| *** shakamunyi has quit IRC | 02:10 | |
| *** shakamunyi has joined #tripleo | 02:12 | |
| slagle | if there are any cores around, https://review.openstack.org/#/q/I3c1ecadceb9623fa03edaf87226297834448d161,n,z removes our stable/liberty pin | 02:17 |
| *** tzumainn has quit IRC | 02:17 | |
| *** shakamunyi has quit IRC | 02:22 | |
| *** shakamunyi has joined #tripleo | 02:23 | |
| *** cwolferh has quit IRC | 02:32 | |
| *** shakamunyi has quit IRC | 02:32 | |
| *** sthillma has quit IRC | 02:38 | |
| *** rbrady has joined #tripleo | 02:45 | |
| openstackgerrit | Reedip proposed openstack/instack-undercloud: Remove Py33 support https://review.openstack.org/266112 | 02:52 |
| openstackgerrit | Reedip proposed openstack/instack: Remove Py33 support https://review.openstack.org/266116 | 03:05 |
| *** spredzy has quit IRC | 03:07 | |
| *** spredzy has joined #tripleo | 03:10 | |
| *** trozet has joined #tripleo | 03:14 | |
| *** cwolferh has joined #tripleo | 03:46 | |
| *** prometheanfire has quit IRC | 03:57 | |
| *** Marga_ has joined #tripleo | 03:58 | |
| *** Marga_ has quit IRC | 03:59 | |
| *** Marga_ has joined #tripleo | 03:59 | |
| *** cwolferh has quit IRC | 04:03 | |
| *** prometheanfire has joined #tripleo | 04:08 | |
| *** yamahata has quit IRC | 04:10 | |
| *** rhallisey has quit IRC | 04:19 | |
| *** links has joined #tripleo | 04:26 | |
| *** Marga_ has quit IRC | 04:26 | |
| *** coolsvap|away is now known as coolsvap | 04:29 | |
| *** morazi has joined #tripleo | 04:30 | |
| *** teju has joined #tripleo | 04:38 | |
| *** trozet has quit IRC | 04:42 | |
| *** cwolferh has joined #tripleo | 04:42 | |
| teju | Hi, I need help with this question - https://ask.openstack.org/en/question/87294/tripleo-heat-stack-failed/ | 04:47 |
| *** tserong has quit IRC | 05:00 | |
| *** stendulker has joined #tripleo | 05:01 | |
| *** dmacpher has quit IRC | 05:05 | |
| *** masco has joined #tripleo | 05:06 | |
| *** tserong has joined #tripleo | 05:07 | |
| *** rlandy has quit IRC | 05:12 | |
| *** dmacpher has joined #tripleo | 05:22 | |
| *** oshvartz has quit IRC | 05:24 | |
| *** dmacpher_ has joined #tripleo | 05:38 | |
| *** dmacpher has quit IRC | 05:42 | |
| *** jaosorior has joined #tripleo | 05:47 | |
| teju | Hi, I am installing RHEL OS P7. I ran the command to deploy overcloud and it failed. I see few deployments failed. I fixed few infra related issues. Now, can I resume the stack to start where it left off last time it failed? | 05:50 |
| *** yamahata has joined #tripleo | 05:51 | |
| *** dmacpher_ has quit IRC | 06:28 | |
| *** ryansb has quit IRC | 06:31 | |
| *** dsneddon has quit IRC | 06:37 | |
| *** jaosorior has quit IRC | 06:48 | |
| *** akuznetsov has joined #tripleo | 06:52 | |
| *** akuznetsov has quit IRC | 07:02 | |
| *** akuznetsov has joined #tripleo | 07:06 | |
| *** jaosorior has joined #tripleo | 07:09 | |
| *** dshulyak_ has joined #tripleo | 07:09 | |
| *** bvandenh has joined #tripleo | 07:12 | |
| *** ukalifon1 has joined #tripleo | 07:15 | |
| *** oshvartz has joined #tripleo | 07:18 | |
| *** bvandenh has quit IRC | 07:22 | |
| *** ryansb has joined #tripleo | 07:24 | |
| *** ryansb has joined #tripleo | 07:24 | |
| *** akuznetsov has quit IRC | 07:40 | |
| *** jprovazn has joined #tripleo | 07:49 | |
| *** jtomasek has joined #tripleo | 07:57 | |
| *** akrivoka has joined #tripleo | 07:58 | |
| *** rasca has joined #tripleo | 07:58 | |
| *** panda has quit IRC | 08:06 | |
| *** rasca has quit IRC | 08:06 | |
| *** jistr has joined #tripleo | 08:06 | |
| *** rasca has joined #tripleo | 08:06 | |
| *** panda has joined #tripleo | 08:06 | |
| *** jistr is now known as jistr|doc | 08:07 | |
| *** hewbrocca has joined #tripleo | 08:13 | |
| *** bvandenh has joined #tripleo | 08:14 | |
| *** aufi has joined #tripleo | 08:15 | |
| *** athomas has joined #tripleo | 08:15 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder: Add dib element to generate logical volumes https://review.openstack.org/252041 | 08:17 |
| *** devvesa has joined #tripleo | 08:28 | |
| *** shardy has joined #tripleo | 08:28 | |
| *** akuznetsov has joined #tripleo | 08:30 | |
| *** liverpooler has joined #tripleo | 08:35 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack-infra/tripleo-ci: Keep py3.X compatibility for urllib https://review.openstack.org/261182 | 08:36 |
| *** mcornea has joined #tripleo | 08:41 | |
| *** yamahata has quit IRC | 08:51 | |
| *** yamahata has joined #tripleo | 08:51 | |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Convert port cidr splitting to str_split https://review.openstack.org/266246 | 08:57 |
| *** ifarkas has joined #tripleo | 08:57 | |
| *** dtantsur|afk is now known as dtantsur | 08:57 | |
| shardy | If anyone is regularly testing network isolation, it'd be good to get feedback on ^^ given we don't test it in the gate | 08:57 |
| shardy | I've done basic local testing and it seems to work, but additional eyes would be good :) | 08:58 |
| *** jistr|doc has quit IRC | 08:59 | |
| *** mbound has joined #tripleo | 09:01 | |
| *** Marga_ has joined #tripleo | 09:02 | |
| *** jdob has quit IRC | 09:02 | |
| hewbrocca | boy I would love it if we could get that gate going | 09:05 |
| *** jdob has joined #tripleo | 09:07 | |
| *** nico_auv has joined #tripleo | 09:09 | |
| *** gfidente has joined #tripleo | 09:11 | |
| *** gfidente has quit IRC | 09:11 | |
| *** gfidente has joined #tripleo | 09:11 | |
| cmyster | shardy: re ^, I am having mysql issues as well. I wonder if its the same thing... | 09:15 |
| cmyster | (usually here it was RAM, DB loves RAM) | 09:16 |
| cmyster | as for exit code 1, thats really none informative. I'de want to see here an stdout of running the same command with the same arguments manually. | 09:17 |
| *** openstackgerrit has quit IRC | 09:17 | |
| *** openstackgerrit has joined #tripleo | 09:17 | |
| shardy | hewbrocca: Yeah, I think we can/should just add it to one of the exiting jobs | 09:20 |
| shardy | cmyster: My patch is a change to the net-iso port templates, so probably not related to DB issues | 09:20 |
| *** jistr has joined #tripleo | 09:20 | |
| hewbrocca | I thought that there was some actual rework of the network in the rack that needed to happen | 09:21 |
| hewbrocca | No idea | 09:21 |
| shardy | hewbrocca: Hmm, Ok not sure about that, will have to check w/dprince | 09:21 |
| cmyster | shardy: once the DB is down all hell breaks loose so unless we want to make sure a DB is up and running every call and report back (probably not) then there is nothing we can do here. I'll answer here as it doesn't seem to me to be net-iso related in this case. | 09:24 |
| *** yamahata has quit IRC | 09:27 | |
| *** Marga_ has quit IRC | 09:30 | |
| cmyster | shardy: BTW regarding what I asked with using all communications with ctrlplane and only having the controllers use an external net, who did you say can help me there? slagle? | 09:32 |
| *** paramite has joined #tripleo | 09:32 | |
| *** olap_ has joined #tripleo | 09:34 | |
| *** derekh has joined #tripleo | 09:35 | |
| *** athomas has quit IRC | 09:41 | |
| *** athomas has joined #tripleo | 09:46 | |
| shardy | cmyster: I'd chat to dsneddon about it when he wakes up | 09:47 |
| *** slagle has quit IRC | 09:52 | |
| *** morazi has quit IRC | 09:52 | |
| *** derekh has quit IRC | 09:54 | |
| *** slagle has joined #tripleo | 09:54 | |
| *** derekh has joined #tripleo | 09:55 | |
| *** social has quit IRC | 09:55 | |
| *** social has joined #tripleo | 09:55 | |
| *** morazi has joined #tripleo | 09:56 | |
| *** mkovacik has joined #tripleo | 09:58 | |
| *** mgould has joined #tripleo | 10:02 | |
| cmyster | ack | 10:03 |
| *** devvesa has quit IRC | 10:12 | |
| *** devvesa has joined #tripleo | 10:12 | |
| *** morazi has quit IRC | 10:15 | |
| *** derekh has quit IRC | 10:16 | |
| *** derekh has joined #tripleo | 10:19 | |
| shardy | lucas-dinner or dtantsur: Hey, is there any way for data in the ironic db to be provided via a nova config drive? | 10:19 |
| *** morazi has joined #tripleo | 10:19 | |
| shardy | I remember there was a patch to enable injection of the discovery data when we stored it in the extra column | 10:20 |
| dtantsur | shardy, hi, that patch was not accepted by nova | 10:20 |
| shardy | I need a way to provide mac->nic mappings, is there any other way you can think of? | 10:20 |
| * dtantsur thinks | 10:20 | |
| shardy | dtantsur: ah, yeah I seem to recall that now | 10:20 |
| shardy | dtantsur: one thing dprince and I were discussing is the discovery/inspector data in swift | 10:21 |
| shardy | e.g, is there some way (tempurl?) we can provide that data to each instance | 10:21 |
| shardy | it's currently stored under the inspector keystone account, are there plans to expose that data via the inspector API? | 10:21 |
| shardy | Access to that data may be another way to generate this mapping | 10:22 |
| dtantsur | shardy, it is exposed via API in inspector as of liberty, however it's only exposed in inspector-client in Mitaka | 10:22 |
| dtantsur | (which should not be a too big deal) | 10:22 |
| shardy | dtantsur: Ok - I'm really looking for a way to consume it at runtime inside each node | 10:22 |
| shardy | where we won't necessarily have credentials to access the inspector API | 10:22 |
| dtantsur | well, you still need credentials for this particular API... | 10:23 |
| shardy | similar to what we could do previously with that nova config drive patch applied | 10:23 |
| shardy | dtantsur: Yeah, that's why I'm wondering about a pre-signed URL per-node or something | 10:23 |
| dtantsur | shardy, can't we build a configdrive in tripleoclient? | 10:23 |
| shardy | it's tricky since we don't know the ironic ID inside the node do we? | 10:23 |
| dtantsur | shardy, I think we don't | 10:23 |
| shardy | dtantsur: Hmm, that's an interesting idea, possibly yes | 10:24 |
| shardy | so, we'd retrieve the swift data, do some processing on it, then attach another config drive to the instance? | 10:24 |
| shardy | we already specify force_config_drive=True, so there's a config drive for each node w/the normal nova user/meta-data | 10:25 |
| dtantsur | at least it's more or less straightforward and is under our control | 10:26 |
| shardy | I guess we'd need to ensure the "admin" user/tenant can access the inspector swift data directly | 10:27 |
| devvesa | Hi. I'm testing TripleO stable/liberty and I find grub errors when overcloud images are spawned | 10:27 |
| shardy | as inside tripleoclient we can't really switch to the IRONIC_INSPECTOR_PASSWORD etc | 10:28 |
| shardy | maybe we just need to configure the container acl's etc to enable that | 10:28 |
| devvesa | error: attempt to read or write outside of partition. | 10:28 |
| devvesa | grub rescue> | 10:28 |
| devvesa | anyone else has experienced this error? | 10:28 |
| dtantsur | shardy, from tripleoclient you can use inspector API to get this data | 10:28 |
| shardy | dtantsur: ah, OK, so we no longer require switching to the inspector credentials to read the data from swift? | 10:29 |
| shardy | http://docs.openstack.org/developer/tripleo-docs/advanced_deployment/profile_matching.html | 10:29 |
| shardy | that still talks about using the inspector password, so I guess it needs an update | 10:29 |
| dtantsur | whatever credentials you use for accessing inspector API will work | 10:29 |
| dtantsur | shardy, this documentation is completely outdated, don't believe a single work from it :) | 10:30 |
| *** paramite is now known as paramite|afk | 10:30 | |
| shardy | dtantsur: Ok, thanks very much for the info, I'll give it a try | 10:30 |
| dtantsur | shardy, speaking of which, when you have some time, may I ask you to review the profile matching patch https://review.openstack.org/#/c/250405/ ? | 10:30 |
| shardy | dtantsur: sure, will do | 10:30 |
| *** weshay_xchat has joined #tripleo | 10:34 | |
| *** electrofelix has joined #tripleo | 10:36 | |
| openstackgerrit | Merged openstack/diskimage-builder: Add pip-and-virtualenv element https://review.openstack.org/239561 | 10:37 |
| *** dsneddon has joined #tripleo | 10:43 | |
| *** dsneddon has quit IRC | 10:50 | |
| *** dsneddon has joined #tripleo | 10:51 | |
| *** dsneddon has quit IRC | 10:51 | |
| *** dsneddon has joined #tripleo | 10:51 | |
| *** dsneddon has quit IRC | 10:52 | |
| *** dsneddon has joined #tripleo | 10:52 | |
| *** jaosorior has quit IRC | 10:55 | |
| *** jaosorior has joined #tripleo | 10:55 | |
| *** weshay_xchat has quit IRC | 10:56 | |
| *** coolsvap is now known as coolsvap|away | 10:56 | |
| *** stendulker has quit IRC | 10:57 | |
| openstackgerrit | Merged openstack/python-tripleoclient: Install python-hardware package on the agent image by default https://review.openstack.org/263225 | 10:59 |
| *** weshay_xchat has joined #tripleo | 11:14 | |
| *** openstackgerrit has quit IRC | 11:17 | |
| *** openstackgerrit has joined #tripleo | 11:17 | |
| *** jcoufal has joined #tripleo | 11:26 | |
| *** lucas-dinner is now known as lucasagomes | 11:27 | |
| *** jaosorior has quit IRC | 11:49 | |
| *** jaosorior has joined #tripleo | 11:50 | |
| *** pblaho has joined #tripleo | 12:01 | |
| *** Tristitia has quit IRC | 12:05 | |
| *** panda has quit IRC | 12:05 | |
| *** panda has joined #tripleo | 12:07 | |
| *** jcoufal has quit IRC | 12:07 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/python-tripleoclient: Fail the introspection command if introspection has failed https://review.openstack.org/257855 | 12:09 |
| *** cmyster has quit IRC | 12:10 | |
| *** cmyster has joined #tripleo | 12:12 | |
| *** cmyster has quit IRC | 12:12 | |
| *** cmyster has joined #tripleo | 12:12 | |
| mgould | morning folks! Can you please take a look at https://review.openstack.org/#/c/266320/ ? | 12:19 |
| d0ugal | mgould: it might be worth asking in #openstack-infra (I *think* that is where the requirements cores are usually) | 12:20 |
| d0ugal | mgould: you may want to use depends-on in the tripleoclient patch btw to link them | 12:21 |
| *** paramite|afk is now known as paramite | 12:24 | |
| dtantsur | mgould, right #openstack-infra is a better place (do not forget to explicitly mention that this is a requirements update, not all infra cores are requirements cores) | 12:26 |
| mgould | dtantsur, will do - thanks! | 12:27 |
| openstackgerrit | Miles Gould proposed openstack/python-tripleoclient: Remove tripleoclient.baremetal wrapper https://review.openstack.org/265336 | 12:28 |
| *** jcoufal has joined #tripleo | 12:29 | |
| d0ugal | mgould: Thanks :) The nice thing about a depends-on is that I can +2 it without worrying it will merge too early | 12:30 |
| mgould | d0ugal, nifty! And thanks for reviewing :-) | 12:31 |
| d0ugal | mgould: Thank you for doing one of my todos! | 12:31 |
| mgould | d0ugal, np :-) | 12:31 |
| *** masco has quit IRC | 12:31 | |
| *** d0ugal has quit IRC | 12:42 | |
| *** d0ugal has joined #tripleo | 12:43 | |
| *** d0ugal is now known as Guest58385 | 12:43 | |
| openstackgerrit | Marek Aufart proposed openstack/tripleo-common: Add environment capabilities https://review.openstack.org/232534 | 12:43 |
| openstackgerrit | Marek Aufart proposed openstack/tripleo-common: Add environment capabilities https://review.openstack.org/232534 | 12:45 |
| *** Guest58385 is now known as d0ugal | 12:45 | |
| *** d0ugal has quit IRC | 12:45 | |
| *** d0ugal has joined #tripleo | 12:45 | |
| *** dprince has joined #tripleo | 12:58 | |
| *** shakamunyi has joined #tripleo | 13:04 | |
| *** jprovazn has quit IRC | 13:06 | |
| *** jprovazn has joined #tripleo | 13:07 | |
| *** jcoufal has quit IRC | 13:07 | |
| *** akuznetsov has quit IRC | 13:08 | |
| *** panda has quit IRC | 13:11 | |
| *** panda has joined #tripleo | 13:11 | |
| *** derekh has quit IRC | 13:11 | |
| *** derekh has joined #tripleo | 13:16 | |
| *** thrash|g0ne is now known as thrash | 13:20 | |
| *** jayg|g0n3 is now known as jayg | 13:22 | |
| EmilienM | I have a patch in puppet-elements, passing all jobs and is here for long time, and just need +A. Can someone approve https://review.openstack.org/#/c/256640/ please? | 13:25 |
| *** jprovazn has quit IRC | 13:25 | |
| *** Marga_ has joined #tripleo | 13:26 | |
| *** aufi has quit IRC | 13:26 | |
| *** rhallisey has joined #tripleo | 13:27 | |
| *** trown|outttypeww is now known as trown | 13:28 | |
| *** akuznetsov has joined #tripleo | 13:36 | |
| *** liverpooler has quit IRC | 13:38 | |
| openstackgerrit | Merged openstack/tripleo-puppet-elements: Add logging components https://review.openstack.org/256640 | 13:39 |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Deploy Logging on the undercloud with Puppet https://review.openstack.org/258653 | 13:44 |
| EmilienM | shardy: see my comment on | 13:46 |
| EmilienM | https://review.openstack.org/#/c/258653/ | 13:46 |
| shardy | EmilienM: ack, thanks! :) | 13:47 |
| openstackgerrit | Ryan Hallisey proposed openstack/tripleo-heat-templates: Allow container template to recognize an update https://review.openstack.org/265368 | 13:47 |
| *** dmacpher has joined #tripleo | 13:54 | |
| *** eggmaster has joined #tripleo | 13:54 | |
| *** tzumainn has joined #tripleo | 13:57 | |
| *** akuznetsov has quit IRC | 14:00 | |
| dprince | meeting time | 14:01 |
| *** rlandy has joined #tripleo | 14:03 | |
| *** Goneri has joined #tripleo | 14:04 | |
| *** rebrego has joined #tripleo | 14:06 | |
| *** regebro is now known as Guest53711 | 14:06 | |
| *** rebrego is now known as regebro | 14:06 | |
| *** jcoufal has joined #tripleo | 14:10 | |
| *** jprovazn has joined #tripleo | 14:10 | |
| *** julim has joined #tripleo | 14:14 | |
| *** masco has joined #tripleo | 14:17 | |
| *** trozet has joined #tripleo | 14:18 | |
| *** akuznetsov has joined #tripleo | 14:21 | |
| *** masco has quit IRC | 14:23 | |
| *** shardy is now known as shardy_mtg | 14:27 | |
| openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Use HAProxy 'transparent' bind option instead of sysctl ip_nonlocal https://review.openstack.org/265309 | 14:28 |
| *** shardy_mtg has quit IRC | 14:32 | |
| *** shardy_mtg has joined #tripleo | 14:34 | |
| *** regebro has quit IRC | 14:35 | |
| *** lblanchard has joined #tripleo | 14:35 | |
| *** regebro has joined #tripleo | 14:36 | |
| openstackgerrit | Brad P. Crochet proposed openstack/instack-undercloud: Clean out os-refresh-config on every run https://review.openstack.org/266404 | 14:39 |
| openstackgerrit | Merged openstack/tripleo-specs: Add spec for TripleO UI https://review.openstack.org/239056 | 14:39 |
| *** jaosorior has quit IRC | 14:40 | |
| openstackgerrit | Brad P. Crochet proposed openstack/instack-undercloud: Clean out os-refresh-config on every run https://review.openstack.org/266404 | 14:42 |
| *** kbyrne has quit IRC | 14:46 | |
| *** Marga_ has quit IRC | 14:50 | |
| *** links has quit IRC | 14:50 | |
| *** Marga_ has joined #tripleo | 14:50 | |
| *** rpothier has joined #tripleo | 14:53 | |
| *** liverpooler has joined #tripleo | 14:54 | |
| dprince | tzumainn, jdob: I would ask that you keep an open mind I guess. Like I said I think it would be grand if Heat API proper supported deploying templates that resided in a swift container | 15:01 |
| dprince | tzumainn, jdob: but I think if heatclient does this (which it already intended to do with --template-object which has been in there for quite some time) then it works for us as is | 15:01 |
| jdob | i'm not having a closed mind about it, I agree it'd be good | 15:02 |
| tzumainn | dprince, that's not what I'm hesitant about - I'm hesitant that we put all this business logic in openstack clients for the sake of speed | 15:02 |
| jdob | right, what tzumainn said | 15:02 |
| dprince | tzumainn: we are putting logic in a libarary, which will be behind an API | 15:02 |
| dprince | tzumainn: my preference is simply rather than adding custom code to tripleo-common, lets fix things in the root project. In this case Heat | 15:03 |
| tzumainn | dprince, but we're doing that in a way different from every other openstack api, by saying that an openstack client acts as a library for Mistral | 15:03 |
| *** jrist has quit IRC | 15:03 | |
| dprince | tzumainn: so... rather than have tripleo-common go and download files from Swift so we can essentially call heatclient locally, I just fixed heatclient | 15:03 |
| dprince | tzumainn: My view was this was just a bug | 15:04 |
| tzumainn | dprince, I think it's a bug if we're fixing things in heat that can't be accessed through the standard Heat API | 15:04 |
| dprince | tzumainn: https://bugs.launchpad.net/python-heatclient/+bug/1532326 | 15:04 |
| openstack | Launchpad bug 1532326 in python-heatclient "--template-object doesn't support nested stacks" [High,Fix released] - Assigned to Dan Prince (dan-prince) | 15:04 |
| tzumainn | dprince, that's what led to the GUI/CLI divergence in the first place | 15:05 |
| dprince | tzumainn: there is no divergence here | 15:05 |
| dprince | tzumainn: the same API request is sent to Heat in either case | 15:05 |
| *** teju has quit IRC | 15:06 | |
| *** lucasagomes is now known as lucas-hungry | 15:06 | |
| dprince | tzumainn: in one case the client downloads the files locally and processes them. In the other case the client downloads them from Swift and processes them | 15:06 |
| tzumainn | dprince, if you put that swift functionality in the heat cli, then cli users can do something that api users can't do? or am I missing something | 15:06 |
| dprince | no, both will use exactly the same path | 15:07 |
| dprince | tzumainn: because we will always upload to a swift container in either case | 15:07 |
| tzumainn | how does the heat api access the heat client logic? | 15:07 |
| dprince | tzumainn: exactly the same path, for CLI and GUI | 15:07 |
| *** egafford has joined #tripleo | 15:08 | |
| dprince | tzumainn: heat API doesn't need the heatclient at all | 15:08 |
| dprince | tzumainn: it is just an API | 15:08 |
| tzumainn | dprince, okay, so I'm completely confused now | 15:08 |
| tzumainn | dprince, I thought you were advocating implementing feature X in heat client, and if so, my question is how the heat API allows access to feature X | 15:08 |
| egafford | EmilienM: ping | 15:09 |
| dprince | tzumainn: I'm advocating just a simple thing. Lets fix heatclients --template-object support (and associated --environment-object) so that it allows users to deploy directly from Swift | 15:09 |
| dprince | tzumainn: this gives us both a CLI and library | 15:09 |
| tzumainn | dprince, I don't see how a client is a library for an API? | 15:10 |
| dprince | tzumainn: which we can easily put behind an API to do the things we need to do | 15:10 |
| EmilienM | egafford: in upstream meeting now, but please let me know what's up, I'll catchup later | 15:10 |
| dprince | tzumainn: right, I'm going to create a Mistral action that does this for us | 15:10 |
| dprince | tzumainn: as a prototype, and if people like it we'll see where it goes | 15:10 |
| tzumainn | dprince, okay, so I completely disagree now | 15:10 |
| egafford | Cool; no problem. Just pinging about what you'd like to do about https://review.openstack.org/#/c/233229/. I'm open to literally anything that provides an accurate deprecation path. :) | 15:10 |
| tzumainn | dprince, you're advocating that in Heat, we put functionality in heatclient, and then provide API access *not* through the Heat API but through Mistral? | 15:11 |
| *** Marga_ has quit IRC | 15:11 | |
| egafford | EmilienM: ^ totally at your convenience; thanks. | 15:11 |
| EmilienM | egafford: will look in a few, thanks for the headsup | 15:11 |
| dprince | tzumainn: the functionality I want already exists (mostly) I'm just advocating we make use of it | 15:11 |
| EmilienM | also ping gchamoul please | 15:11 |
| dprince | tzumainn: We still still be using Heat API just like we always did | 15:12 |
| *** Marga_ has joined #tripleo | 15:12 | |
| tzumainn | dprince, but you're putting feature X in Heat client and then saying we should use Mistral for API access feature X in Heat, correct? | 15:12 |
| dprince | tzumainn: I'm exposing a client library capability via Mistral. Yes | 15:13 |
| dprince | tzumainn: to be fair I think that is totally reasonable | 15:13 |
| tzumainn | dprince, hm, okay | 15:13 |
| tzumainn | dprince, I think we should see what others thing : ) | 15:13 |
| tzumainn | er, think : ) | 15:13 |
| dprince | tzumainn: jtomasek asked directly if we could do "local disk" things on the email thread. Same thing I think | 15:13 |
| dprince | tzumainn: I would also create a Mistral action to upload "stock" templates from /usr/share I think | 15:14 |
| dprince | tzumainn: and if the end user doesn't supply those we could simply upload that... | 15:14 |
| gchamoul | egafford: just need to pick up my daugther at school and I will try finish that one! but tbh I am quite confused about it now :D | 15:14 |
| tzumainn | dprince, to be honest these seem like solutions of convenience rather than ones that fit the openstack model | 15:14 |
| gchamoul | egafford: will be back in ~20min | 15:14 |
| dprince | tzumainn: I am trying to simplify things, yes | 15:15 |
| *** jrist has joined #tripleo | 15:15 | |
| egafford | gchamoul: Ack; thanks. | 15:15 |
| dprince | tzumainn: as for OpenStack model I'm not sure what rules I'm breaking | 15:15 |
| tzumainn | dprince, the expectation that feature X in project Y is accessible through project Y's API, rather than Mistral's API | 15:16 |
| *** jrist has quit IRC | 15:16 | |
| dprince | tzumainn: the essence of TripleO is to use OpenStack to deploy OpenStack. | 15:16 |
| *** jrist has joined #tripleo | 15:16 | |
| *** jrist has joined #tripleo | 15:16 | |
| dprince | tzumainn: FWIW, once project Y gets the capabilities we need I'm game for using it | 15:16 |
| jtomasek | dprince, tzumainn: I think we agree that problem is that some feature implemented in heatclient should reside in Heat directly. It does not really matter if we use tripleo-common+api or mistral action to call heatclient, both seems to have the same original issue | 15:17 |
| dprince | tzumainn: I don't think we have the luxury of waiting for these things in all cases though... | 15:17 |
| tzumainn | dprince, yes, but what you're describing sounds like a somewhat hackish solution | 15:17 |
| shardy_mtg | tzumainn: The feature under discussion is a workflow to deploy heat stacks | 15:17 |
| shardy_mtg | Upload some heat templates to swift, run some template validation on those templates, deploy those templates, poll heat until completion | 15:18 |
| shardy_mtg | that is a workflow | 15:18 |
| dprince | tzumainn: so yes, perhaps there is a "release valve" sort of nature to what I'm suggesting. We you rather I just used Ansible Tower... because I mean there are people who use that and don't think it dirty at all | 15:18 |
| shardy_mtg | tzumainn: the fact that some convenience functions exist in heatclient is incidental really | 15:18 |
| dprince | jtomasek: yes | 15:19 |
| jtomasek | shardy_mtg: from GUI POW it is a multiple workflows because GUI needs to add inputs at certain parts of the overal workflow | 15:19 |
| dprince | I actually think there are many things we agree on here | 15:19 |
| shardy_mtg | e.g they just avoid folks having to dowload the files from swift (heatclient does it, same as for HTTP URLs or local files) | 15:19 |
| *** shardy_mtg is now known as shardy | 15:19 | |
| slagle | +2 plz :), https://review.openstack.org/#/q/I3c1ecadceb9623fa03edaf87226297834448d161,n,z | 15:20 |
| tzumainn | shardy, is that accurate? I thought that dprince was saying that we should implement all of that within heat client, and then use mistral to access it from an API perspective | 15:20 |
| tzumainn | shardy, wait, let me simplify | 15:21 |
| tzumainn | shardy, I asked dprince if it was correct that he was 'putting feature X in Heat client and then saying we should use Mistral for API access feature X in Heat, correct?' and he agreed | 15:21 |
| dprince | tzumainn: what shardy says is correct I think | 15:21 |
| tzumainn | shardy, and I'm saying that seems wrong | 15:21 |
| dprince | tzumainn: I think you are just looking at the wrong "feature" perhaps | 15:22 |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Changes for configuring Nuage https://review.openstack.org/260399 | 15:22 |
| dprince | tzumainn: the Heat API already has the feature we really need, namely to deploy a stack | 15:22 |
| dprince | tzumainn: we are really just arguing about where the code to deal with local vs. Swift template storage goes | 15:23 |
| shardy | tzumainn: the ability to download templates from swift already exists in heatclient, it's just a bit old and incomplete | 15:23 |
| tzumainn | shardy, dprince, why doesn't that code belong in Heat outside of the heatclient? | 15:23 |
| shardy | tzumainn: It's possible a future heat API version might also enable directly referincing swift, but this capability exists now | 15:23 |
| dprince | tzumainn: and given... that today heatclient already attempts to deal with all this... a small refinement and fix is all we need in heatclient I think to get us in a good place to develop our workflow around say a Swift storage backend | 15:24 |
| shardy | tzumainn: because in general it's better not to allow users to DoS heat by passing random URLs to it, it's safer to require the user to create a request containing all the artefacts that define a stack | 15:24 |
| shardy | tzumainn: that's one reason heat doesn't download the files internally | 15:25 |
| shardy | it may be possible in a future version, but the workflow of 1. download templates from $somewhere and 2. Call heat with said templates, already works | 15:25 |
| tzumainn | shardy, so you think it's logical to say 'heat won't provide API access to this functionality, so it's just in the heatclient; but it's okay that we proxy that API access through Mistral'? | 15:25 |
| shardy | tzumainn: it's not proxying that API, it's defining a workflow to store templates in swift then deploy them | 15:26 |
| shardy | the API is dowload from swift, then call heat | 15:26 |
| dprince | tzumainn: I think it is okay. But do note that I would use this same path regardless of CLI or UI. All of our workflows would use it the same | 15:26 |
| shardy | tzumainn: what I think you would prefer is if heat accepted the swift object identifier directly | 15:27 |
| shardy | which may be possible, but the DoS concerns remain, and then we'd be proxying user requests to swift | 15:27 |
| *** Tristitia has joined #tripleo | 15:27 | |
| shardy | I can see both sides of the argument, but I think it's not accurate to say the mistral thing is a stealth heat API | 15:28 |
| shardy | it's defining a series of actions which interact with heat, and swift, and potentially other services in future | 15:28 |
| *** Marga_ has quit IRC | 15:28 | |
| shardy | tzumainn: the same series of actions will be required inside the tripleO API | 15:28 |
| dprince | shardy: exactly, Mistral is no more a stealth Heat API than Heat is a stealth Nova API | 15:28 |
| *** kbyrne has joined #tripleo | 15:29 | |
| tzumainn | dprince, shardy, does the Heat API call the Nova CLI? | 15:29 |
| dprince | it is just a different way to organize the workflows | 15:29 |
| *** kbyrne has quit IRC | 15:29 | |
| openstackgerrit | Jaume Devesa proposed openstack/tripleo-heat-templates: Make enabling of controller services configurable. https://review.openstack.org/260413 | 15:29 |
| dprince | tzumainn: Heat uses novaclient, yes :) | 15:29 |
| shardy | tzumainn: Yes, it does, but it calls it to instantiate an object defined in the declarative HOT model | 15:29 |
| *** kbyrne has joined #tripleo | 15:30 | |
| shardy | tzumainn: it also calls swift, via resources defined in templates e.g to create a container | 15:30 |
| shardy | what it can't do, currently, is download the pieces to create a stack directly from swift | 15:30 |
| shardy | heatclient can do that tho, as a convenience thing | 15:30 |
| shardy | you could do exactly the same with a small script doing the download from swift to the local filesystem, then a stack-create | 15:31 |
| dprince | Again, If heat grows this capability then great. If not I can live with it. Given we have client work arounds I'm not sure "Heat API" is our biggest problem ATM | 15:31 |
| *** bvandenh has quit IRC | 15:32 | |
| tzumainn | dprince, that's what I don't like; the idea of putting in a bunch of client workarounds and then using mistral instead of waiting for individual projects to express the features through their own API | 15:33 |
| tzumainn | but meh, may as well see how this discussion shakes out in the mailing list :P | 15:33 |
| dprince | tzumainn: these aren't client workarounds | 15:33 |
| dprince | tzumainn: these are client features, that stand on their own | 15:33 |
| dprince | tzumainn: regardless of whether we use them the patches stand | 15:34 |
| dprince | tzumainn: and that I think is what sways me to like this sort of solution... it encourages us to add to existing clients (heat, nova, neutron, etc.) rather than build our own thing in tripleo-common | 15:35 |
| dprince | tzumainn: we will require some wrappers in TripleO common, sure to combine things in the right places | 15:35 |
| tzumainn | dprince, myabe I misstated my objection; it feels like Mistral is a workaround for having a feature in heatclient that can't be expressed through the project's API | 15:36 |
| dprince | tzumainn: Yeah. I don't think so. All I'm doing is re-working some of the tripleo-common/python-tripleoclient workflows so that we have less code in TripleO | 15:37 |
| dprince | tzumainn: and ultimately less code, but exactly the same workflow | 15:37 |
| shardy | tzumainn: it can be expressed via the projects APIs though, 1. download template from swift, 2. heat stack-create | 15:37 |
| tzumainn | dprince, shardy, if you told me that heatclient was also using mistral, then I think I'd be less up in arms; both the heatclient and our API access would go directly through mistral | 15:37 |
| dprince | tzumainn: Mistral is just the workflow API, nothing more really | 15:37 |
| shardy | that can be done regardless of anything in heatclient | 15:38 |
| dprince | shardy: exactly, with more code mind you | 15:38 |
| tzumainn | dprince, shardy, but what I think is being suggested here, is that heatclient doesn't use mistral, and to gain the exact same feature through an API we'd use mistral | 15:38 |
| dprince | shardy: but perhaps I should stop confusing people with heatclient things here | 15:38 |
| shardy | dprince: maybe it'd be simpler to discuss a more general "deployment workflow" | 15:39 |
| dprince | shardy: agree | 15:39 |
| *** aufi has joined #tripleo | 15:43 | |
| dprince | on a similar note... the pattern I'm really arguing here is lets add our "library" code to the right places | 15:44 |
| dprince | we take a similar tone with Puppet for example | 15:44 |
| dprince | like rather than adding our own puppet-tripleo composition classes for a Nova feature... just add them to puppet-nova instead if it makes sense | 15:45 |
| dprince | so in this case heatclient (the library) gets a couple of fixes to deploy from Swift, rather than us adding it to our own tripleo-common library like we currently do | 15:45 |
| dprince | how that all fits into our grand deployment workflow is a related... but different issue. I'm sorry we've intermixed these things in the conversation today | 15:48 |
| dprince | tzumainn: ^^ | 15:48 |
| *** kbyrne has quit IRC | 15:50 | |
| *** mkovacik has quit IRC | 15:51 | |
| *** mcornea has quit IRC | 15:52 | |
| *** dshulyak_ has quit IRC | 15:53 | |
| *** dshulyak_ has joined #tripleo | 15:54 | |
| *** bvandenh has joined #tripleo | 15:55 | |
| *** dshulyak_ has quit IRC | 15:55 | |
| shardy | dprince: Hey, I looked into the configdrive vs deploy artefacts thing | 15:57 |
| openstackgerrit | Dan Prince proposed openstack/instack-undercloud: Set default Glance backend to Swift https://review.openstack.org/248492 | 15:57 |
| shardy | turns out there is no way to pass the data needed via the configdrive, because the nova patch enabling ironic to inject config drive data was rejected | 15:58 |
| *** masco has joined #tripleo | 15:58 | |
| shardy | so my options are either a new deployment resource (which runs before NetworkDeployment), put the mapping in swift and generate a tempurl, or move the deploy artefacts deployment to before NetworkDeployment | 15:59 |
| shardy | it's a shame, because adding this data to ironic would really be ideal | 15:59 |
| *** rpothier_ has joined #tripleo | 16:00 | |
| dprince | shardy: I wonder if it is worth a mailing list discussion about the Nova config drive support | 16:00 |
| dprince | shardy: lucas-hungry's patch was only a couple of Nova lines... | 16:00 |
| dprince | shardy: and quite simple | 16:00 |
| shardy | dprince: maybe, it'd certainly make things simpler | 16:00 |
| *** liverpooler has quit IRC | 16:00 | |
| dprince | shardy: anyways, If you need to proceed with other options we don't have to block on that. I'd just be keen to have this in the right spot and it is sad that Nova -> Ironic is limited in this case | 16:01 |
| *** rpothier has quit IRC | 16:03 | |
| *** lucas-hungry is now known as lucasagomes | 16:05 | |
| lucasagomes | dprince, hi there | 16:05 |
| lucasagomes | dprince, yeah, it was 1 line change actually (w/o tests) | 16:05 |
| *** panda has quit IRC | 16:05 | |
| lucasagomes | dprince, but there other problems with the approach I remember was pointed out, for e.g, configdrive for Ironic would be different than the config drive for other drivers | 16:06 |
| *** julim has quit IRC | 16:06 | |
| *** panda has joined #tripleo | 16:06 | |
| lucasagomes | dprince, the metadata service would have different information than the configdrive (unless we also updated the configdrive) | 16:06 |
| shardy | Do we have a reference for the simplest possible network-isolation setup? E.g I have a 2 node nonha environment, do I need to configure --baremetal-bridge-names etc to make it work? | 16:06 |
| lucasagomes | (the metadata)* | 16:06 |
| *** julim has joined #tripleo | 16:07 | |
| trozet | EmilienM: there has been some talk that the puppet-aodh module isn't stable. Is that true or does it work just fine with liberty? | 16:08 |
| *** kbyrne has joined #tripleo | 16:09 | |
| EmilienM | trozet: what? puppet-aodh is very stable and working | 16:10 |
| trozet | EmilienM: thanks figured it was just rumors | 16:11 |
| EmilienM | it's just OPM master which was not built | 16:11 |
| EmilienM | but should be fix now | 16:11 |
| trozet | EmilienM: ok cool, do you still need help on the tht portion?\ | 16:14 |
| EmilienM | I'm trying to deploy aodh on the ovecloud | 16:14 |
| trozet | EmilienM: ok there is a Doctor project on OPNFV that uses aodh, they are asking if we can add support by our release in 2.5 weeks | 16:15 |
| openstackgerrit | Jaume Devesa proposed openstack/tripleo-heat-templates: MidoNet heat templates https://review.openstack.org/260415 | 16:15 |
| trozet | EmilienM: so if I can get past a few other bugs I can help you either code/test it | 16:16 |
| *** masco has quit IRC | 16:16 | |
| EmilienM | trozet: well, I'm working on it, but very had to push for new features lately | 16:16 |
| *** ayoung has quit IRC | 16:16 | |
| trozet | EmilienM: we can even manually patch in your patch to our build | 16:16 |
| trozet | EmilienM: and test it | 16:16 |
| EmilienM | trozet: undercloud has aodh | 16:16 |
| EmilienM | trozet: see https://review.openstack.org/241408 | 16:16 |
| trozet | EmilienM: yeah my opendaylight patch in tripleO has been sitting there for months, so we manually patch it into tht | 16:16 |
| trozet | during our build | 16:16 |
| trozet | EmilienM: we only need for overcloud | 16:16 |
| *** aufi has quit IRC | 16:20 | |
| *** fgimenez has joined #tripleo | 16:20 | |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: TEST: this will help testing major upgrades https://review.openstack.org/260456 | 16:24 |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: Add resources for major upgrade in Pacemaker scenario https://review.openstack.org/253276 | 16:24 |
| openstackgerrit | Merged openstack/tripleo-common: Adds file deletion to plan manager https://review.openstack.org/255366 | 16:26 |
| *** sparkyco_ has joined #tripleo | 16:26 | |
| *** athomas has quit IRC | 16:28 | |
| *** mkovacik has joined #tripleo | 16:31 | |
| *** athomas has joined #tripleo | 16:33 | |
| *** olap_ has quit IRC | 16:33 | |
| *** dcain has joined #tripleo | 16:35 | |
| tzumainn | dprince, sorry, just saw your comment - that broadly makes sense to me, actually | 16:38 |
| dprince | tzumainn: cool | 16:39 |
| dprince | egafford: +2 on the sahara patches BTW | 16:39 |
| dprince | egafford: if we can get someone else to review them it would be nice to have them land soo | 16:39 |
| dprince | n | 16:39 |
| dprince | egafford: one of your Trove patches was still WIP? | 16:40 |
| openstackgerrit | Merged openstack/tripleo-common: Add the missing plan name to the PlanAlreadyExistsError https://review.openstack.org/257343 | 16:41 |
| *** pradk has joined #tripleo | 16:42 | |
| openstackgerrit | Ben Nemec proposed openstack/tripleo-heat-templates: Configure keystone public_endpoint https://review.openstack.org/265922 | 16:42 |
| openstackgerrit | Ben Nemec proposed openstack/tripleo-heat-templates: Enable TLS in loadbalancer if cert path is detected https://review.openstack.org/266469 | 16:42 |
| *** rasca has quit IRC | 16:44 | |
| *** dtantsur is now known as dtantsur|brb | 16:45 | |
| egafford | dprince: Oh my goodness that's wonderful. | 16:45 |
| *** derekh has quit IRC | 16:46 | |
| egafford | dprince: Yeah, I fear that I haven't gotten enough expert testing on the heat templates patch just yet; happily the patch is structurally extremely similar to Sahara. | 16:46 |
| egafford | Any change to trove-integration is likely to come in the fairly minor block of actually pushing properties to puppet-trove, and should be easy to review, though, and I know that my patch does get trove up, running, and responding, so if we're willing to accept an potentially intermediary and not utterly tested patch at this stage, with the idea that fix patches will be comparatively tiny, I'd love to remove the WIP., | 16:48 |
| egafford | I just also don't want to introduce any issues into your product. | 16:48 |
| egafford | dprince: ^ Happy to go with your preferences on that. | 16:48 |
| dprince | egafford: okay, I will review it too then, thanks | 16:49 |
| *** akuznetsov has quit IRC | 16:49 | |
| *** sparkyco_ has quit IRC | 16:52 | |
| egafford | dprince: You have pre-emptively made my year, and it's only January. | 16:54 |
| dprince | egafford: yeah, well you patches were posted early last fall I think... so we've kept you waiting a good bit I think | 16:55 |
| egafford | shardy: Any chance you can take a look @ https://review.openstack.org/220863 ? dprince has 1 +2; it'd be awesome to merge it if we can. | 16:55 |
| egafford | dprince: True; then again, priorities. I get it. I'm just really excited to put the main patches and subsequent rebase cycles behind me. :) | 16:56 |
| shardy | egafford: Hey, looks good, but can you confirm how I disable Sahara? | 17:00 |
| shardy | I commented to that effect previously and I don't see the boolean to control it | 17:00 |
| egafford | dprince, shardy: What's our decision on this requirement? I'm getting conflicting signals. | 17:01 |
| shardy | egafford: if the answer is, we're deferring that to composable roles, that's OK, I just want to be clear what the plan is | 17:02 |
| shardy | egafford: I did ask this question back in September on the review | 17:02 |
| shardy | otherwise lgtm :) | 17:03 |
| egafford | shardy: That was my last definition of the plan. Yeah, totally; I talked with you and dprince a bit about concerns about hanging uncontrolled concerns handled by other repos (like os-cloud-config for keystone). | 17:03 |
| egafford | It's an utterly valid concern, but in recent convos with dprince I think the sense is that we are waiting on roles to actualize that, as the infrastructure isn't fully there to disable services across all of tripleo (though doing so in just the heat templates is reasonably possible.) | 17:04 |
| dprince | shardy: I could go either way on the option. Before the end of this cycle I would like to have composable roles done... but still getting buy on that | 17:05 |
| shardy | egafford: Ok, I guess I was expecting to see e.g a ControllerEnableSahara, which fed into a boolean in the puppet manifest | 17:05 |
| dprince | shardy: if we end up not getting things composable then perhaps we add a boolean option later... | 17:06 |
| shardy | but if that's too hard due to cross-repo concerns, I'm OK with saying we have to wait until composable roles | 17:06 |
| dprince | shardy: good questions though | 17:06 |
| shardy | I guess my question then is, if a boolean is too hard, will composable roles actually solve the problem? | 17:06 |
| shardy | given that it's really just a fancy way of doing booleans, with a lot of flexibility | 17:06 |
| *** devvesa has quit IRC | 17:07 | |
| shardy | I guess stuff like the os-cloud-config thing & keystone endpoints is solved by moving to puppet for keystone configuration | 17:07 |
| shardy | egafford: but are there other things we need to be thinking about? | 17:07 |
| shardy | if so it'd be good to highlight them | 17:08 |
| egafford | shardy: Yeah, moving into heat for all resources is important for encapsulation. Honestly, you're Way More Of An Authority on whether there are other things to think about. | 17:08 |
| shardy | egafford: Yeah, I was just looking for evidence if you actually tried to wire in the boolean and stuff broke :) | 17:09 |
| shardy | cos the same stuff will probably still break w/composable roles in t-h-t | 17:09 |
| shardy | anyway, I'm not blocking on this, just wanted to discuss it | 17:09 |
| egafford | My fundamental concern with building the first set of optional roles is precisely that I am an outsider to your product. Well, I did absolutely try back in the day and see that os-cloud-config was still building endpoints. | 17:09 |
| egafford | After I saw that, it became a disprove-a-universal-negative-or-deeply-learn-literally-every-line-of-code-in-tripleo proposition for me. | 17:10 |
| *** penick has quit IRC | 17:16 | |
| *** alop has joined #tripleo | 17:21 | |
| openstackgerrit | Merged openstack/tripleo-puppet-elements: overcloud-controller: install python-networking-odl https://review.openstack.org/261256 | 17:24 |
| openstackgerrit | Ethan Gafford proposed openstack/tripleo-heat-templates: Trove Integration https://review.openstack.org/233240 | 17:25 |
| *** dtantsur|brb is now known as dtantsur|afk | 17:27 | |
| *** fgimenez has quit IRC | 17:30 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Sahara Integration https://review.openstack.org/220863 | 17:31 |
| shardy | egafford: ^^ | 17:31 |
| shardy | As noted on the review, I'd prefer it to be optional, but hopefully we can work that out later via composable roles | 17:31 |
| egafford | shardy: That's amazing; thanks. | 17:32 |
| egafford | I'd be happy to try to push an optionalization patch now that the main patch is in, though I fear that I Just Will Not Absolutely Know if there are resources elsewhere (outside heat/puppet) that are still being created. | 17:33 |
| *** oshvartz has quit IRC | 17:40 | |
| *** mbound has quit IRC | 17:41 | |
| *** hewbrocca has quit IRC | 17:44 | |
| *** eil397 has joined #tripleo | 17:46 | |
| *** dshulyak_ has joined #tripleo | 17:48 | |
| *** paramite has quit IRC | 17:49 | |
| *** dshulyak_ has quit IRC | 17:51 | |
| *** athomas has quit IRC | 17:51 | |
| *** athomas has joined #tripleo | 17:59 | |
| *** trown is now known as trown|lunch | 18:02 | |
| *** jistr has quit IRC | 18:03 | |
| *** lblanchard has quit IRC | 18:08 | |
| *** lblanchard has joined #tripleo | 18:10 | |
| *** gfidente is now known as gfidente|afk | 18:11 | |
| *** akuznetsov has joined #tripleo | 18:12 | |
| *** mgould has quit IRC | 18:12 | |
| *** yamahata has joined #tripleo | 18:15 | |
| *** Marga_ has joined #tripleo | 18:17 | |
| *** regebro has quit IRC | 18:19 | |
| egafford | dprince: Who else can +2 https://review.openstack.org/#/c/221418/ ? | 18:20 |
| dprince | egafford: anyone on tripleo-core really | 18:22 |
| dprince | egafford: I think steve was going to look at it from earlier... | 18:23 |
| egafford | dprince: I'd also like to ask where the best place to add some very light testing (probably just to ensure Sahara can still respond to an empty cluster-list call) would be. | 18:23 |
| egafford | dprince: It'd be great to enforce just a tiny bit of "this can still work at all." | 18:24 |
| egafford | I don't want to bog down your CI building half-hour long CDH clusters, but. :) | 18:24 |
| dprince | egafford: perhaps a simple test could be added onto this (if it is quick) | 18:25 |
| dprince | egafford: https://review.openstack.org/#/c/241167/ | 18:26 |
| dprince | egafford: quick... as it it doesn't cost much in terms of CI wall time | 18:26 |
| dprince | egafford: < a minute lets say | 18:26 |
| egafford | Cool. I literally just want to call sahara cluster-list. | 18:26 |
| egafford | And verify a 200 and an empty list. | 18:26 |
| *** jaosorior has joined #tripleo | 18:27 | |
| *** ukalifon1 has quit IRC | 18:32 | |
| *** ukalifon has joined #tripleo | 18:32 | |
| openstackgerrit | Ryan Hallisey proposed openstack/tripleo-common: Deploy Docker compute role with tripleo.sh https://review.openstack.org/264933 | 18:40 |
| *** ifarkas has quit IRC | 18:49 | |
| *** ukalifon has quit IRC | 18:51 | |
| *** ayoung has joined #tripleo | 18:52 | |
| *** rickflare has joined #tripleo | 18:53 | |
| rickflare | anyone know what would cause this error | 18:53 |
| rickflare | Failed to start Hiera: RuntimeError: Config file /etc/puppetlabs/code/hiera.yaml not found | 18:53 |
| *** sthillma has joined #tripleo | 18:58 | |
| *** jaosorior has quit IRC | 19:00 | |
| *** jaosorior has joined #tripleo | 19:00 | |
| *** sthillma has quit IRC | 19:03 | |
| *** electrofelix has quit IRC | 19:04 | |
| *** julim has quit IRC | 19:06 | |
| *** trown|lunch is now known as trown | 19:06 | |
| *** julim has joined #tripleo | 19:08 | |
| *** ayoung has quit IRC | 19:08 | |
| *** athomas has quit IRC | 19:11 | |
| *** athomas has joined #tripleo | 19:19 | |
| *** nico_auv has quit IRC | 19:19 | |
| *** akrivoka has quit IRC | 19:21 | |
| *** sthillma has joined #tripleo | 19:29 | |
| *** jaosorior has quit IRC | 19:31 | |
| *** jaosorior has joined #tripleo | 19:32 | |
| *** julim has quit IRC | 19:40 | |
| *** ayoung has joined #tripleo | 19:41 | |
| *** julim has joined #tripleo | 19:41 | |
| *** cwolferh has quit IRC | 19:42 | |
| *** oshvartz has joined #tripleo | 19:50 | |
| stevebaker | dprince: hey, I had a thought https://review.openstack.org/#/c/265479/2/heatclient/common/template_utils.py | 20:03 |
| dprince | stevebaker: okay, lets have it :) | 20:03 |
| stevebaker | dprince: the thought is on line #174 | 20:04 |
| dprince | stevebaker: even better, looking | 20:04 |
| *** panda has quit IRC | 20:05 | |
| dprince | stevebaker: okay, this would work I think | 20:06 |
| dprince | stevebaker: and how would we expose it via the CLI? and potentially eventually in openstackclient too? | 20:06 |
| *** panda has joined #tripleo | 20:06 | |
| dprince | stevebaker: the only downside of this approach I think is it wouldn't allow you to mix the two approaches, local files and those in swift for example | 20:07 |
| dprince | stevebaker: agree that the ordering issue could be a problem though with my initial approach. alternate there might be to track the ordering via a ordering_hints option too.... getting to be to many options though so what you propose might be a reasonable compromise | 20:09 |
| stevebaker | dprince: I've been thinking about that. There could be an option like --object-path-prefix which can be specified multiple times. Any env path which matches the prefix is treated as an object. So its like my object-store catalog approach but explicit instead of implicit (this would handle your localhost case). And the cli env_path_is_object implementation could always fallback to the implicit catalog | 20:10 |
| stevebaker | approach, so in the common case the user wouldn't even need to specify the extra argument | 20:10 |
| *** gfidente|afk has quit IRC | 20:11 | |
| dprince | stevebaker: okay, want me to run these pasted shardy? or have at it? | 20:13 |
| dprince | stevebaker: or, would you like to take these over? | 20:13 |
| stevebaker | dprince: we can continue discussing the cli options at leisure, but I think you can safely make the lib change | 20:14 |
| *** lucasagomes is now known as lucas-dinner | 20:14 | |
| dprince | stevebaker: yeah, I think so too | 20:14 |
| dprince | stevebaker: and the CLI could even wait until openstackclient when that happens I guess | 20:14 |
| dprince | stevebaker: because you mentioned a freeze right? | 20:14 |
| stevebaker | dprince: yes, and shell.py can always have a env_path_is_object which only does the implicit approach, so it still works without having to add a cli option | 20:15 |
| stevebaker | everybody wins! | 20:15 |
| *** dshulyak_ has joined #tripleo | 20:15 | |
| dprince | stevebaker: okay, lets go with it then | 20:16 |
| dprince | stevebaker: thanks for the updates, I'll try to get those applied | 20:16 |
| stevebaker | cool | 20:16 |
| *** dprince has quit IRC | 20:23 | |
| *** dshulyak_ has quit IRC | 20:27 | |
| *** dcain1 has joined #tripleo | 20:32 | |
| *** dcain has quit IRC | 20:35 | |
| *** akuznetsov has quit IRC | 20:39 | |
| *** shivrao has joined #tripleo | 20:40 | |
| *** akuznetsov has joined #tripleo | 20:41 | |
| *** jaosorior has quit IRC | 20:44 | |
| shardy | If anyone has a moment https://review.openstack.org/#/c/260144/ needs a final +2 | 20:52 |
| *** shivrao has quit IRC | 20:53 | |
| *** shardy has quit IRC | 20:55 | |
| *** rlandy has quit IRC | 21:03 | |
| *** jayg is now known as jayg|g0n3 | 21:07 | |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Deploy Logging on the undercloud with Puppet https://review.openstack.org/258653 | 21:10 |
| *** ayoung has quit IRC | 21:14 | |
| *** jtomasek has quit IRC | 21:18 | |
| *** rhallisey has quit IRC | 21:18 | |
| *** rlandy has joined #tripleo | 21:22 | |
| *** jprovazn has quit IRC | 21:24 | |
| *** dcain1 has quit IRC | 21:25 | |
| *** dcain has joined #tripleo | 21:26 | |
| *** bvandenh has quit IRC | 21:26 | |
| *** oshvartz has quit IRC | 21:29 | |
| *** sthillma has quit IRC | 21:33 | |
| *** sthillma has joined #tripleo | 21:34 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder: Initial add of gentoo support for diskimage-builder https://review.openstack.org/263349 | 21:35 |
| prometheanfire | SpamapS: hi, let me know what you think of that new patch | 21:36 |
| *** yamahata has quit IRC | 21:36 | |
| *** cwolferh has joined #tripleo | 21:38 | |
| prometheanfire | SpamapS: gpgv doesn't seem to work for that use, so sticking with gpg --verify | 21:44 |
| *** julim has quit IRC | 21:49 | |
| *** trown is now known as trown|outttypeww | 21:53 | |
| SpamapS | prometheanfire: oh, what does it do? | 21:59 |
| *** ayoung has joined #tripleo | 22:00 | |
| *** eil397 has quit IRC | 22:00 | |
| prometheanfire | not work | 22:01 |
| prometheanfire | :P | 22:01 |
| prometheanfire | it doesn't like the format | 22:01 |
| prometheanfire | gpgv takes a specific format for the keyring | 22:01 |
| *** shivrao has joined #tripleo | 22:01 | |
| prometheanfire | it would add another step | 22:01 |
| prometheanfire | gpg --verify is good as I use it because it is using a keyring that has a single key | 22:01 |
| prometheanfire | so I'm not doing the extra step to export the key I just imported into a specific format | 22:02 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Manage keystone initialization directly in t-h-t manifests https://review.openstack.org/244162 | 22:02 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: endpoint_map: bump nova api from 2.0 to 2.1 https://review.openstack.org/263366 | 22:03 |
| *** akuznetsov has quit IRC | 22:03 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: WIP - Deploy Aodh services, replacing Ceilometer Alarm https://review.openstack.org/241408 | 22:03 |
| SpamapS | prometheanfire: eh, not sure what you mean | 22:03 |
| SpamapS | prometheanfire: it works fine here.. as I wrote the original command | 22:03 |
| SpamapS | prometheanfire: gpgv reads the exact same keyring format as gpg. | 22:04 |
| prometheanfire | it didn't work for me | 22:05 |
| SpamapS | http://paste.openstack.org/show/483668/ | 22:06 |
| SpamapS | prometheanfire: thats what I did (just re-exporting it from the thing I manually pasted the echo into) | 22:06 |
| prometheanfire | SpamapS: I'm gonna generate a kbx file, we can commit and use that instead | 22:08 |
| prometheanfire | no need for gpgdir at all then | 22:08 |
| *** davidlenwell has quit IRC | 22:09 | |
| SpamapS | what's a kbx file? | 22:09 |
| prometheanfire | what gpgv needs | 22:09 |
| prometheanfire | it's a format | 22:09 |
| *** lblanchard has quit IRC | 22:11 | |
| *** lblanchard has joined #tripleo | 22:11 | |
| *** davidlenwell has joined #tripleo | 22:11 | |
| *** ChanServ sets mode: +v davidlenwell | 22:11 | |
| SpamapS | prometheanfire: that is.. um. Not right. | 22:12 |
| SpamapS | prometheanfire: it needs no such thing. :-P | 22:12 |
| prometheanfire | ok | 22:13 |
| prometheanfire | tested your way, it worked | 22:13 |
| *** jcoufal has quit IRC | 22:13 | |
| prometheanfire | didn't work in locally, but I do odd things with gpg | 22:13 |
| *** yamahata has joined #tripleo | 22:14 | |
| SpamapS | prometheanfire: are you maybe on a newer version that has changed behavior? (I hate things that do that) | 22:14 |
| prometheanfire | 2.1 | 22:14 |
| prometheanfire | .10 even | 22:14 |
| prometheanfire | a friend is an upstream dev of gpg | 22:15 |
| SpamapS | actually I just tried with 2.0.22 and it worked fine too | 22:15 |
| *** eil397 has joined #tripleo | 22:16 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder: Initial add of gentoo support for diskimage-builder https://review.openstack.org/263349 | 22:16 |
| prometheanfire | new review | 22:16 |
| prometheanfire | has gpgv use | 22:16 |
| SpamapS | prometheanfire: appreciate your patience on this. I am a bit of a stickler for getting the chain of custody right. :-P | 22:16 |
| prometheanfire | that's fine, I tend to be as well | 22:17 |
| *** lblanchard has quit IRC | 22:17 | |
| prometheanfire | this last pass was more cosmetic though | 22:17 |
| SpamapS | yeah, I might have just +2'd that had I not started wondering if maybe gpgv2 borked the interface or something. | 22:18 |
| *** michchap has joined #tripleo | 22:18 | |
| prometheanfire | :P, anyway, it's good now | 22:18 |
| SpamapS | Yeah, +2'd | 22:20 |
| prometheanfire | no | 22:20 |
| prometheanfire | actually... | 22:20 |
| prometheanfire | you have a system with gpg 2.1? | 22:20 |
| prometheanfire | gpgv --keyring "${GPGDIR}"/gentookeys.gpg stage4-amd64-cloud-nomultilib-20160107.tar.bz2.DIGESTS.asc | 22:20 |
| prometheanfire | gpgv: Ohhhh jeeee: ... this is a bug (keyid.c:342:keystrlen) | 22:20 |
| prometheanfire | Aborted | 22:21 |
| prometheanfire | I'll bug that guy I know | 22:21 |
| prometheanfire | https://lists.gnupg.org/pipermail/gnupg-devel/2016-January/030674.html | 22:21 |
| prometheanfire | so, I'd say stick with gpgv, 2.1 (still considered unstable) will get the fix | 22:22 |
| *** ayoung has quit IRC | 22:23 | |
| *** Goneri has quit IRC | 22:24 | |
| trozet | dsneddon: question for you on defining nic heat templates | 22:26 |
| dsneddon | trozet, Shoot | 22:26 |
| trozet | dsneddon: https://gerrit.opnfv.org/gerrit/gitweb?p=apex.git;a=blob;f=build/nics/controller.yaml;h=336b34a12eb35dbbc1cc7d43321ba698ebfe2720;hb=HEAD | 22:28 |
| trozet | dsneddon: can you see that? | 22:28 |
| dsneddon | trozet, I see it | 22:28 |
| *** jpeeler has quit IRC | 22:28 | |
| trozet | dsneddon: my question is (I talked to you about network isolation yesterday), you can see there is one nic and br-ex | 22:28 |
| trozet | dsneddon: we have 4 nics | 22:28 |
| trozet | dsneddon: when I ran overcloud deploy the 2nd and 4th nic failed to get an IP address, cause i dont have them in the template | 22:29 |
| dsneddon | trozet, Just put them in, with just type: interface, name: nicX, use_dhcp: false | 22:29 |
| trozet | dsneddon: so adding them to the template will get it to pass I think, but my question is if I disable those 2 networks | 22:29 |
| trozet | dsneddon: but use the same template, will it just ignore the NICs that do not have values for the IP addresses | 22:30 |
| *** cwolferh has quit IRC | 22:30 | |
| *** rpothier_ has quit IRC | 22:30 | |
| dsneddon | trozet, If you have the NICs with use_dhcp: false, you don't need to put anything for IP addresses. Those interfaces will be unused. | 22:30 |
| dsneddon | trozet, If you want to use them for provider networks, then you have to put them on a bridge (still with no IP), and include them in the --bridge-mappings. | 22:31 |
| trozet | dsneddon: well we want to put internal api traffic on them | 22:32 |
| trozet | dsneddon: so the VIPs will be created, but we dont need any IP on the NICs themselves? | 22:32 |
| dsneddon | trozet, All you have in those templates is a control plane and an External network. Only those two VIPs should be created. If you have other VIPs, then you should configure other networks. | 22:33 |
| dsneddon | trozet, Do you need to see what a multiple interface configuration looks like? | 22:33 |
| dsneddon | trozet, Check this out: https://code.engineering.redhat.com/gerrit/#/c/65267/ | 22:34 |
| dsneddon | trozet, Is that what you are trying to do? | 22:34 |
| dsneddon | trozet, Or do you really just want two NICs enabled with just control plane and External? | 22:35 |
| openstackgerrit | James Slagle proposed openstack/tripleo-heat-templates: Add Satellite 5 support https://review.openstack.org/266602 | 22:36 |
| trozet | dsneddon: no we are trying to cover multiple cases | 22:37 |
| trozet | dsneddon: so a user running deploy can specify the network layout | 22:38 |
| dsneddon | trozet, Well, in any case, you have to put the IPs somewhere in the templates, not the VIPs, but the VIPs will go on the interfaces with the matching IPs. | 22:38 |
| trozet | dsneddon: exactly! | 22:38 |
| dsneddon | trozet, So if you don't assign the IPs, then the VIPs won't be assigned either. | 22:38 |
| trozet | dsneddon: right | 22:39 |
| dsneddon | trozet, So use the example I posted above to assign the Internal API to the right NIC, and tenant, etc. | 22:39 |
| trozet | dsneddon: so lets say maximum i have 4 networks | 22:39 |
| trozet | dsneddon: but a user could decide at deploy time to only use 2 | 22:39 |
| dsneddon | trozet, Are you assuming the user doesn't have the ability to use VLANs? | 22:39 |
| dsneddon | trozet, Or bonds? | 22:40 |
| prometheanfire | SpamapS: question about dib, once it's in, do images get generated and uploaded to https://www.openstack.org/marketplace/distros/ ? | 22:40 |
| trozet | dsneddon: my question is, if i put all 4 NICs into the nic/controller.yaml, but only run a deployment with 2 networks, will heat fail because {get_param: InternalAPIp} would be blank | 22:40 |
| trozet | dsneddon: we havent gotten to vlans or bond parsing yet | 22:41 |
| dsneddon | trozet, os-net-config would fail, because you would have an interface with a stanza that said ip_netmask: <blank> | 22:41 |
| trozet | dsneddon: ok so then what I will do is make a separate nic file for each scenario, then replace that one in the resource registry at deploy time | 22:41 |
| dsneddon | trozet, I have an idea how to create these using an interface. | 22:42 |
| prometheanfire | maybe I just need to point http://docs.openstack.org/image-guide/obtain-images.html to gentoo's mirrors | 22:42 |
| trozet | dsneddon: ok, what is it? | 22:42 |
| SpamapS | prometheanfire: no they do not. dib is just a tool. :) | 22:43 |
| dsneddon | trozet, First, you ask the user how many interfaces they have, and fill an array with [nic1:nic<num_ifaces>] | 22:43 |
| SpamapS | prometheanfire: but dib would be a nice way to build images for the marketplace. | 22:43 |
| prometheanfire | k | 22:43 |
| prometheanfire | asking my docs person who I need to contact about adding to http://docs.openstack.org/image-guide/obtain-images.html | 22:43 |
| dsneddon | trozet, Then you ask them which interface is used for provisioning, and assign that one to the control plane interface. Now you have X unused interfaces. | 22:43 |
| dsneddon | trozet, You also have the user select which of the available networks they want (all are optional, really). | 22:44 |
| trozet | dsneddon: we have something like that but more intelligent | 22:44 |
| trozet | dsneddon: https://gerrit.opnfv.org/gerrit/gitweb?p=apex.git;a=blob;f=config/deploy/network/network_settings.yaml;h=2560cc5d2bbc91bee69626ed0b4f7facdae82a99;hb=HEAD | 22:44 |
| SpamapS | prometheanfire: believe it or not, this is the first time I've seen that page. :) | 22:44 |
| trozet | dsneddon: that is all a user needs to define his tripleO network | 22:44 |
| trozet | dsneddon: he can even leave out values and we auto-detect them on the network | 22:44 |
| trozet | dsneddon: for virtual deployments we can also auto-generate missing values | 22:45 |
| dsneddon | trozet, OK, I see what you are going for. | 22:45 |
| trozet | dsneddon: so you see how in that settings file, private_network and storage_network are disabled | 22:45 |
| trozet | dsneddon: my patch provides the code to allow you to turn them on, but overcloud deploy failed cause the nics are missing | 22:45 |
| trozet | dsneddon: so you can see there is yaml there to indicate VLAN or bond, but we dont currently parse those yet | 22:46 |
| dsneddon | trozet, I'd have to think about this some more. I feel like this is the wrong schema to represent the common scenarios. | 22:47 |
| trozet | dsneddon: its tailored to OPNFV, which only uses 4 networks | 22:47 |
| trozet | dsneddon: admin, private, public, storage | 22:48 |
| dsneddon | trozet, In that case, I think you need to define different templates for 2, 3, or 4 NICs in use, or produce the templates from another template, that only includes the required nics. | 22:48 |
| trozet | dsneddon: but it makes deploying much easier when we handle all the translation for the user so he doesnt need to know heat or tripleO | 22:48 |
| trozet | dsneddon: exactly | 22:48 |
| trozet | dsneddon: so thanks for the info! I will go do that | 22:49 |
| dsneddon | trozet, Just ask them for number of interfaces, and retrieve nic_template_<num_ifaces>.yaml | 22:49 |
| trozet | dsneddon: yup | 22:49 |
| dsneddon | trozet, I have toyed with the idea of making that a recoverable error in os-net-config | 22:50 |
| *** egafford has quit IRC | 22:50 | |
| dsneddon | trozet, As long as you are using 4 networks, you might as well reuse the provisioning network as admin. Is that what you do? | 22:51 |
| trozet | dsneddon: what do you mean as "admin"? | 22:51 |
| trozet | dsneddon: oh you mean our admin_network? | 22:52 |
| dsneddon | trozet, You just said: "dsneddon: admin, private, public, storage" | 22:52 |
| trozet | dsneddon: yeah that is the provisioning network | 22:52 |
| trozet | dsneddon: private carries tenant and internal api | 22:52 |
| trozet | dsneddon: public is the external net, and storage carries storage and storagemgmt (set all that in the ServiceNetMap) | 22:52 |
| dsneddon | trozet, I might actually argue that admin should carry provisioning and internal api, leaving a dedicated interface for tenant. | 22:52 |
| dsneddon | trozet, I generally recommend against combining tenant and internal api, because they are both busy networks, and you don't want tenants impacting operational services. | 22:53 |
| trozet | dsneddon: yeah its true | 22:53 |
| trozet | dsneddon: this was decided way back in OPNFV, but perhaps we should change this | 22:53 |
| dsneddon | trozet, Think about it. Unless the bulk of the traffic is on public, and tenant internal traffic is minimal. | 22:54 |
| dsneddon | trozet, By think about it, I mean consider it. I think it would be a good change for scale. | 22:55 |
| *** alop has quit IRC | 22:56 | |
| *** cwolferh has joined #tripleo | 22:56 | |
| trozet | dsneddon: we arent dealing with a large scale deployment, its really something one person sets up and tests NFV | 22:56 |
| trozet | dsneddon: but we should do it the right way and I agree with you, but its not my decision alone to make | 22:56 |
| *** alop has joined #tripleo | 22:57 | |
| *** dcain has quit IRC | 22:57 | |
| trozet | dsneddon: standard was setup by OPNFV so I would need to convince other people, will try to do it after this release | 22:57 |
| dsneddon | trozet, OK, let me know if you need backup. | 22:57 |
| trozet | dsneddon: thanks...and thanks for the help! | 22:58 |
| prometheanfire | lol | 22:58 |
| trozet | dsneddon: also why i have you, any chance on you reviewing that opendaylight patch soon? | 22:58 |
| *** thrash is now known as thrash|g0ne | 23:00 | |
| *** sthillma has quit IRC | 23:04 | |
| *** sthillma has joined #tripleo | 23:05 | |
| *** Goneri has joined #tripleo | 23:07 | |
| *** pradk has quit IRC | 23:08 | |
| *** davidlenwell has quit IRC | 23:17 | |
| *** davidlenwell has joined #tripleo | 23:18 | |
| *** ChanServ sets mode: +v davidlenwell | 23:18 | |
| *** weshay_xchat has quit IRC | 23:37 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder: Initial add of gentoo support for diskimage-builder https://review.openstack.org/263349 | 23:38 |
| prometheanfire | SpamapS: updated with changes from robbat2 | 23:38 |
| *** weshay_xchat has joined #tripleo | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!