| *** aufi has quit IRC | 00:14 | |
| *** dmarlin has joined #tripleo | 00:17 | |
| *** thrash is now known as thrash|g0ne | 00:18 | |
| *** bfournie has quit IRC | 00:23 | |
| *** limao has joined #tripleo | 00:35 | |
| rook | bnemec: hey - ovb question | 00:44 |
|---|---|---|
| rook | so i have deployed my OOO overcloud (baremetal) and I am going through the doc and it mentions "Apply the Nova pxe boot patch file in the patches directory to the host cloud Nova." is this on the overcloud controller? | 00:45 |
| rook | ah *compute nodes* | 00:49 |
| rook | bnemec: however, watching your youtube video i got the sense that I didn't need to apply anything to the overcloud? | 00:50 |
| *** ctayal has joined #tripleo | 00:54 | |
| *** yolanda has joined #tripleo | 00:56 | |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 01:13 |
| *** penick has joined #tripleo | 01:14 | |
| *** penick_ has joined #tripleo | 01:15 | |
| openstackgerrit | Alex Schultz proposed openstack/python-tripleoclient: Implement overcloud delete command https://review.openstack.org/401024 | 01:18 |
| *** ctayal has quit IRC | 01:19 | |
| *** ctayal has joined #tripleo | 01:19 | |
| *** penick has quit IRC | 01:19 | |
| *** penick_ is now known as penick | 01:19 | |
| *** ansmith has joined #tripleo | 01:20 | |
| *** newmember has quit IRC | 01:23 | |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Stop using documentation network range https://review.openstack.org/395529 | 01:24 |
| *** ctayal has quit IRC | 01:28 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Enables auto-detection for VIP interfaces https://review.openstack.org/390401 | 01:29 |
| *** bfournie has joined #tripleo | 01:32 | |
| rook | hey bnemec - seeing this https://gist.github.com/jtaleric/42eebb9d0891c3f2faeca1bdd09f5aab -- might be something I fubared? | 01:49 |
| *** ctayal has joined #tripleo | 01:59 | |
| *** Goneri has quit IRC | 02:00 | |
| *** cwolferh has quit IRC | 02:04 | |
| *** achadha_ has quit IRC | 02:09 | |
| *** Goneri has joined #tripleo | 02:10 | |
| *** achadha has joined #tripleo | 02:12 | |
| *** tzumainn has quit IRC | 02:13 | |
| *** achadha has quit IRC | 02:17 | |
| *** achadha has joined #tripleo | 02:19 | |
| *** achadha has quit IRC | 02:24 | |
| *** ctayal has quit IRC | 02:24 | |
| *** cdearborn has joined #tripleo | 02:25 | |
| *** bana_k has quit IRC | 02:25 | |
| *** [2]cdearborn has joined #tripleo | 02:30 | |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 02:30 |
| *** [4]cdearborn has joined #tripleo | 02:35 | |
| *** [1]cdearborn has quit IRC | 02:39 | |
| *** [1]cdearborn has joined #tripleo | 02:39 | |
| *** cwolferh has joined #tripleo | 02:42 | |
| *** cdearborn has quit IRC | 02:44 | |
| *** cdearborn has joined #tripleo | 02:48 | |
| *** [2]cdearborn has quit IRC | 02:50 | |
| *** fzdarsky_ has joined #tripleo | 02:50 | |
| *** [3]cdearborn has joined #tripleo | 02:51 | |
| *** fragatina has quit IRC | 02:53 | |
| *** fzdarsky|afk has quit IRC | 02:54 | |
| *** [4]cdearborn has quit IRC | 02:54 | |
| *** noslzzp has quit IRC | 02:55 | |
| *** brault has quit IRC | 03:04 | |
| *** [1]cdearborn has quit IRC | 03:04 | |
| *** midl has quit IRC | 03:05 | |
| *** ansmith has quit IRC | 03:06 | |
| *** cdearborn has quit IRC | 03:06 | |
| *** achadha has joined #tripleo | 03:08 | |
| *** [3]cdearborn has quit IRC | 03:09 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 03:11 |
| *** dmacpher has joined #tripleo | 03:14 | |
| *** bkopilov has quit IRC | 03:15 | |
| *** rhefner has quit IRC | 03:22 | |
| *** zoli|gone has quit IRC | 03:23 | |
| *** cschwede has quit IRC | 03:23 | |
| *** myoung has quit IRC | 03:24 | |
| *** myoung has joined #tripleo | 03:25 | |
| *** marios has quit IRC | 03:26 | |
| *** marios has joined #tripleo | 03:26 | |
| *** chandankumar has joined #tripleo | 03:27 | |
| *** zoliXXL has joined #tripleo | 03:28 | |
| *** cschwede has joined #tripleo | 03:28 | |
| *** yamahata has quit IRC | 03:29 | |
| *** cwolferh has quit IRC | 03:29 | |
| openstackgerrit | RedHat RDO CI proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/360618 | 03:30 |
| openstackgerrit | RedHat RDO CI proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/365449 | 03:30 |
| *** udesale has joined #tripleo | 03:34 | |
| openstackgerrit | Brent Eagles proposed openstack/tripleo-heat-templates: Run os-net-config before restarting cluster on update https://review.openstack.org/400960 | 03:40 |
| *** links has joined #tripleo | 03:42 | |
| *** fragatina has joined #tripleo | 03:51 | |
| *** fragatina has quit IRC | 03:52 | |
| *** fragatina has joined #tripleo | 03:52 | |
| *** Guest83036 has joined #tripleo | 03:56 | |
| *** sudipto_ has joined #tripleo | 03:56 | |
| *** Goneri has quit IRC | 04:11 | |
| *** misty-cloud has joined #tripleo | 04:16 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 04:22 |
| *** yamahata has joined #tripleo | 04:29 | |
| *** bkopilov has joined #tripleo | 04:37 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements https://review.openstack.org/400776 | 04:40 |
| *** Guest83036 has quit IRC | 04:41 | |
| *** sudipto_ has quit IRC | 04:41 | |
| *** yamahata has quit IRC | 04:43 | |
| openstackgerrit | Aparna proposed openstack/diskimage-builder: Update hpssacli to ssacli in proliant-tools element https://review.openstack.org/396504 | 04:52 |
| *** achadha has quit IRC | 04:58 | |
| *** achadha has joined #tripleo | 05:02 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 05:05 |
| *** yamahata has joined #tripleo | 05:05 | |
| *** newmember has joined #tripleo | 05:08 | |
| *** pgadiya has joined #tripleo | 05:13 | |
| *** masco has joined #tripleo | 05:15 | |
| *** Vijayendra_ has joined #tripleo | 05:15 | |
| *** achadha has quit IRC | 05:16 | |
| *** prateek has joined #tripleo | 05:18 | |
| *** achadha has joined #tripleo | 05:21 | |
| *** achadha has quit IRC | 05:30 | |
| *** achadha has joined #tripleo | 05:35 | |
| *** achadha has quit IRC | 05:49 | |
| *** sudipto_ has joined #tripleo | 05:50 | |
| *** Guest83036 has joined #tripleo | 05:50 | |
| *** skramaja has joined #tripleo | 05:50 | |
| *** achadha has joined #tripleo | 05:54 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 05:58 |
| *** achadha has quit IRC | 05:58 | |
| *** pkovar has joined #tripleo | 05:59 | |
| *** chandankumar has quit IRC | 05:59 | |
| *** dmacpher has quit IRC | 06:03 | |
| *** ctayal has joined #tripleo | 06:03 | |
| *** lmiccini has joined #tripleo | 06:17 | |
| *** bana_k has joined #tripleo | 06:20 | |
| *** achadha has joined #tripleo | 06:25 | |
| *** bana_k has quit IRC | 06:27 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 06:30 |
| *** florianf has joined #tripleo | 06:31 | |
| *** bana_k has joined #tripleo | 06:33 | |
| *** achadha has quit IRC | 06:35 | |
| *** abregman has joined #tripleo | 06:36 | |
| *** achadha has joined #tripleo | 06:39 | |
| *** dsariel has joined #tripleo | 06:42 | |
| *** karthiks has quit IRC | 06:42 | |
| *** oshvartz has joined #tripleo | 06:47 | |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: OVN plugin configuration fixes https://review.openstack.org/397674 | 06:47 |
| *** achadha has quit IRC | 06:49 | |
| *** bana_k has quit IRC | 06:49 | |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: OVN plugin configuration fixes https://review.openstack.org/397674 | 06:49 |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: OVN plugin configuration fixes https://review.openstack.org/397674 | 06:55 |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: Disable Neutron agents with OVN. https://review.openstack.org/388088 | 06:55 |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: Split OVN northd and ml2 plugin https://review.openstack.org/387940 | 06:55 |
| *** karthiks has joined #tripleo | 06:56 | |
| *** abehl has joined #tripleo | 06:57 | |
| *** bvandenh has joined #tripleo | 06:59 | |
| *** iranzo has joined #tripleo | 07:07 | |
| *** iranzo has joined #tripleo | 07:07 | |
| *** hjensas has quit IRC | 07:11 | |
| *** achadha has joined #tripleo | 07:12 | |
| *** tesseract has joined #tripleo | 07:12 | |
| *** rasca has joined #tripleo | 07:12 | |
| *** tesseract is now known as Guest59528 | 07:13 | |
| *** aufi has joined #tripleo | 07:13 | |
| *** achadha_ has joined #tripleo | 07:16 | |
| *** achadha has quit IRC | 07:16 | |
| *** numans has joined #tripleo | 07:17 | |
| *** ccamacho has joined #tripleo | 07:17 | |
| *** mhenkel has joined #tripleo | 07:18 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 07:18 |
| openstackgerrit | Merged openstack/tripleo-specs: Move tripleo-repos spec from policy to ocata https://review.openstack.org/397976 | 07:21 |
| *** ealcaniz has joined #tripleo | 07:21 | |
| *** jprovazn has joined #tripleo | 07:23 | |
| ccamacho | morning guys! | 07:25 |
| *** tremble has joined #tripleo | 07:25 | |
| openstackgerrit | Merged openstack/tripleo-specs: Composable HA services https://review.openstack.org/378035 | 07:27 |
| *** chandankumar has joined #tripleo | 07:31 | |
| *** iranzo has quit IRC | 07:32 | |
| *** tobias_fiberdata has joined #tripleo | 07:33 | |
| *** iranzo has joined #tripleo | 07:36 | |
| *** iranzo has joined #tripleo | 07:36 | |
| *** tobias-fiberdata has quit IRC | 07:36 | |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: OVN plugin configuration fixes https://review.openstack.org/397674 | 07:37 |
| *** b00tcat has joined #tripleo | 07:38 | |
| *** brault has joined #tripleo | 07:39 | |
| *** yamahata has quit IRC | 07:39 | |
| *** ctayal_ has joined #tripleo | 07:40 | |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 07:42 |
| *** aufi has quit IRC | 07:43 | |
| *** ctayal has quit IRC | 07:44 | |
| *** ctayal_ has quit IRC | 07:53 | |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 07:54 |
| *** gchamoul is now known as gchamoul|afk | 07:55 | |
| *** hjensas has joined #tripleo | 08:00 | |
| *** tobias-fiberdata has joined #tripleo | 08:00 | |
| *** gchamoul|afk is now known as gchamoul | 08:00 | |
| *** tobias_fiberdata has quit IRC | 08:03 | |
| *** mcornea has joined #tripleo | 08:05 | |
| *** ohamada has joined #tripleo | 08:06 | |
| *** aufi has joined #tripleo | 08:10 | |
| *** liverpooler has quit IRC | 08:13 | |
| *** jpena|off is now known as jpena | 08:20 | |
| *** jlinkes has joined #tripleo | 08:20 | |
| *** derekjhyang has quit IRC | 08:21 | |
| apetrich | d0ugal, btw you were right both ways work for https://review.openstack.org/#/c/398289 like here http://paste.openstack.org/show/590169/ | 08:21 |
| apetrich | d0ugal, what I'm not sure is how to address the comment. Should I fix the docstring on the utils? | 08:22 |
| d0ugal | apetrich: good question. Let me read the commend again. | 08:23 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 08:24 |
| d0ugal | apetrich: I would just reply to Ben | 08:25 |
| d0ugal | apetrich: The Heat API docs in this case define it as: stack_identity: The UUID or the name of a stack. | 08:26 |
| d0ugal | apetrich: so stack_id can be either a uuid or a name. They are both forms of ID | 08:26 |
| apetrich | d0ugal, cheers | 08:26 |
| d0ugal | apetrich: so the docstring is correct, but a little confusing. | 08:26 |
| * apetrich copies and will paste | 08:27 | |
| d0ugal | apetrich: if you want, you could try and improve the docstring in heatclient: https://github.com/openstack/python-heatclient/blob/master/heatclient/v1/stacks.py#L267 | 08:27 |
| d0ugal | apetrich: but it is used in other places... so that might mean quite a few updates | 08:27 |
| d0ugal | apetrich: hah, I can reply if you want | 08:27 |
| *** dtantsur|afk is now known as dtantsur | 08:27 | |
| apetrich | it's fine. I'm doing that I need to rebase probably | 08:28 |
| *** leanderthal|afk is now known as leanderthal | 08:29 | |
| *** jaosorior has joined #tripleo | 08:29 | |
| *** shardy has joined #tripleo | 08:30 | |
| *** jpich has joined #tripleo | 08:30 | |
| *** tobias_fiberdata has joined #tripleo | 08:34 | |
| *** tobias-fiberdata has quit IRC | 08:37 | |
| jaosorior | anybody has a +A for this? https://review.openstack.org/#/c/400648/ :D | 08:38 |
| openstackgerrit | Govardhan Chintha proposed openstack/puppet-tripleo: HPELeftHandISCSIDriver support for cinder https://review.openstack.org/400873 | 08:40 |
| jaosorior | ccamacho: hey dude, if you have time could you review this https://review.openstack.org/#/c/398177/ ? | 08:43 |
| ccamacho | jaosorior, sure thing dude! | 08:43 |
| *** amoralej|off is now known as amoralej | 08:43 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 08:43 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 08:47 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Special case dib-python in dib-lint https://review.openstack.org/401127 | 08:47 |
| ccamacho | jaosorior, to reproduce it locally just need to deploy the overcloud with TLS && environments/services/barbican.yaml ? | 08:47 |
| flaper87 | shardy: hey, good morning :) | 08:55 |
| flaper87 | shardy: can haz +A? https://review.openstack.org/#/c/330659/ | 08:55 |
| shardy | flaper87: g'morning! Sure! | 08:56 |
| shardy | done | 08:56 |
| jaosorior | ccamacho: no, encrypted volumes can be tested without TLKS | 08:57 |
| jaosorior | *TLS | 08:57 |
| ccamacho | jaosorior, ACK | 08:57 |
| jaosorior | ccamacho: only thing is that you need to have barbican, cinder and nova deployed | 08:57 |
| *** oneswig has joined #tripleo | 08:57 | |
| jaosorior | ccamacho: http://docs.openstack.org/juno/config-reference/content/section_create-encrypted-volume-type.html | 08:57 |
| jaosorior | ccamacho: actually, this commit is relevant. It adds encrypted volumes to scenario004 https://review.openstack.org/#/c/398221/ basically what it does is that it sets the cinder volume that we use to boot the nova server as an encrypted volume | 08:58 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Special case dib-python in dib-lint https://review.openstack.org/401127 | 08:59 |
| jaosorior | ccamacho: and barbican is used to fetch the keys on the nova and cinder side | 08:59 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Trace package install in package-installs-v2 https://review.openstack.org/401048 | 08:59 |
| ccamacho | jaosorior, awesome, checking Ill give you feedback asap :) | 08:59 |
| *** arxcruz has joined #tripleo | 09:01 | |
| flaper87 | shardy: danke :D | 09:02 |
| shardy | be good to get the ci job running again soon so we can ensure it stays working | 09:02 |
| shardy | flaper87: also, I've been looking at composable upgrades for the puppet based deployments - have you spent any time looking at the upgrade workflow with containers? | 09:03 |
| shardy | flaper87: I think it's very important we consider that early this time, as it's been a massive pain point with the puppet based deployments | 09:03 |
| shardy | https://review.openstack.org/#/c/393448/ is the prototype I'm working on - I suspect almost all of it will be different for containers | 09:04 |
| flaper87 | shardy: yes and yes! I've looked into the composable upgrades (admittedly not enough) because I want to start doing this asap | 09:04 |
| shardy | but it'd be good to understand the differences | 09:04 |
| flaper87 | shardy: I fully agree with you | 09:04 |
| shardy | flaper87: Ok, sounds good | 09:05 |
| shardy | I was also wondering about how feasable converting to a container during upgrade would be | 09:05 |
| *** ohamada_ has joined #tripleo | 09:05 | |
| shardy | I suspect this will be much harder given our current arcitecture | 09:06 |
| shardy | but it'd be interesting to look at how hard it would be, e.g perhaps for just one service, e.g upgrade keystone | 09:06 |
| *** ohamada__ has joined #tripleo | 09:06 | |
| *** ohamada_ has quit IRC | 09:06 | |
| *** ohamada__ has quit IRC | 09:06 | |
| *** ohamada__ has joined #tripleo | 09:06 | |
| *** ohamada has quit IRC | 09:07 | |
| *** cylopez has joined #tripleo | 09:07 | |
| *** udesale has quit IRC | 09:09 | |
| *** udesale has joined #tripleo | 09:10 | |
| *** zoliXXL is now known as zoli | 09:13 | |
| *** zoli is now known as zoliXXL | 09:13 | |
| *** akrivoka has joined #tripleo | 09:13 | |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::agent class https://review.openstack.org/400618 | 09:14 |
| *** achadha_ has quit IRC | 09:14 | |
| *** anton has quit IRC | 09:15 | |
| *** zoliXXL is now known as zoli|brb | 09:15 | |
| *** achadha has joined #tripleo | 09:18 | |
| *** fragatina has quit IRC | 09:20 | |
| *** fragatina has joined #tripleo | 09:20 | |
| *** liverpooler has joined #tripleo | 09:23 | |
| *** paramite has joined #tripleo | 09:24 | |
| paramite | greetings, do we backport fixes to stable/mitaka or the branch has been already "closed"? | 09:24 |
| b00tcat | any idea why I'm getting these in the puppet 3.x unit tests? http://logs.openstack.org/18/400618/5/check/gate-puppet-tripleo-puppet-unit-3.6-centos-7/86c30f9/console.html the test that I committed is the following https://review.openstack.org/#/c/400618/4/spec/classes/tripleo_midonet_agent_spec.rb but it's pretty straightforward | 09:24 |
| mhenkel | bandini: shardy told me that vrrp is only used for neutron? | 09:24 |
| shardy | paramite: yes we do, but liberty is about to go EOL | 09:25 |
| b00tcat | paramite: only security and critical bug fixes afaik | 09:25 |
| mhenkel | I thought it is used for more than that: | 09:25 |
| openstackgerrit | Merged openstack/instack-undercloud: Remove deprecated network range 192.0.2.0 https://review.openstack.org/395702 | 09:25 |
| mhenkel | going through the code it provides the vip for internal_api, storage, storage_mgmt, redis and public/external | 09:26 |
| bandini | mhenkel: so in the non-HA deployments we do use keepalived + vrrp, but not in the HA deployments where the VIP is managed by pacemaker | 09:26 |
| jaosorior | sshnaidm|away: it merged!! | 09:26 |
| *** sshnaidm|away is now known as sshnaidm | 09:26 | |
| sshnaidm | jaosorior, \o/ | 09:27 |
| mhenkel | bandini: hmm, is that in newton-stable? | 09:27 |
| *** achadha has quit IRC | 09:27 | |
| mhenkel | ‘cause I have keepalived running after ha deployment | 09:27 |
| bandini | mhenkel: hohum. Can you PM me your full deployment command line? | 09:28 |
| mhenkel | sent | 09:30 |
| mhenkel | also on that topic we investigated a bit into anycast | 09:30 |
| mhenkel | thing is that more and more customers are deploying on l3 clos architectures where no L2 is available between racks | 09:31 |
| *** lucas-afk is now known as lucasagomes | 09:31 | |
| mhenkel | no L2 -> no pacemaker, no keepalived between controller nodes | 09:31 |
| *** achadha has joined #tripleo | 09:32 | |
| mhenkel | options are either dns anycast with a low ttl or ip anycast using bgp/ospf between controller nodes and the top of rack switches | 09:32 |
| paramite | shardy, b00tcat: ok thanks | 09:32 |
| mhenkel | we tried the second solution in our lab and it seems to work so far | 09:32 |
| mhenkel | we use quagga to advertise the same /32 ip address on all three controllers | 09:33 |
| *** zoli|brb is now known as zoli | 09:33 | |
| *** zoli is now known as zoliXXL | 09:33 | |
| mhenkel | the top of rack switches will use ECMP to distribute the traffic across the controllers | 09:33 |
| shardy | mhenkel: this is very interesting - it'd be good if we could loop dsneddon into this discussion later when he wakes up :) | 09:34 |
| dsneddon | mhenkel, That was exactly the HA solution that I designed for Cloudscaling, before I joined Red Hat. It worked great. | 09:35 |
| shardy | Oh hey dsneddon, I thought you'd be asleep ;) | 09:35 |
| dtantsur | shardy, dsneddon never sleeps apparently :) | 09:36 |
| mhenkel | dnseddon: you were with Cloudscaling. So you must now Randy | 09:36 |
| *** shinobu_ has joined #tripleo | 09:36 | |
| mhenkel | now -> know | 09:36 |
| dsneddon | mhenkel, I've tried in vain to convince some of the HA people I've talked to that it was a good idea, but they weren't convinced. | 09:36 |
| dsneddon | mhenkel, Frankly, I think that design makes a TON of sense along with kubernetes. | 09:36 |
| bandini | dsneddon: when was this? I wouldn't mind at least rediscussing/reevaluating this approach | 09:37 |
| dsneddon | mhenkel, Even if the BGP is totally internal to the k8s environment. | 09:37 |
| dsneddon | bandini, This was Havana and Icehouse. | 09:37 |
| dsneddon | bandini, I have some old OpenStack Summit presentations if you are interested. | 09:37 |
| bandini | dsneddon: I am, it'd be great if you can throw some of that stuff over the fence to me ;) | 09:38 |
| shardy | Yeah, it'd be good to perhaps capture this discussion in a spec if it's something folks want to revisit | 09:38 |
| bandini | shardy: yeah! | 09:38 |
| dsneddon | bandini, https://www.openstack.org/summit/san-diego-2012/openstack-summit-sessions/presentation/service-resiliency-doesn-t-always-mean-ha-or-cluster | 09:38 |
| mhenkel | dsneddon: as said, a lot of our customers are deploying hyper scale L3 clos architectures just to be forced to re-introduce L2 for OpenStack HA;) | 09:38 |
| dsneddon | bandini, I take over around 12:30 | 09:39 |
| *** nyechiel has joined #tripleo | 09:39 | |
| mhenkel | who needs to be convinced that this will be required? | 09:39 |
| bandini | dsneddon: thanks. I will look at it | 09:39 |
| *** gfidente has joined #tripleo | 09:40 | |
| *** achadha has quit IRC | 09:40 | |
| *** derekh has joined #tripleo | 09:41 | |
| dsneddon | mhenkel, Well, that's a good question. There was discussion around containers for TripleO where we talked about how to do HA, and we need to move away from Pacemaker for the sake of containers, that much is clear. | 09:42 |
| *** dbecker has joined #tripleo | 09:42 | |
| * bandini nods | 09:42 | |
| dsneddon | mhenkel, But as far as what to do, one of the options was to build a custom Ingress module. | 09:43 |
| dsneddon | mhenkel, For kubernetes. | 09:43 |
| dsneddon | mhenkel, But there is a bigger question about how to do the networking (since kubernetes doesn't yet have a full enough featureset, TripleO needs to deploy using host networking for the time being). | 09:44 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::agent class https://review.openstack.org/400618 | 09:44 |
| jaosorior | sshnaidm: right, back to the TLs everywhere job, this seems to be failing at the moment http://logs.openstack.org/38/391738/30/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-fakeha-caserver/b242695/console.html#_2016-11-23_09_40_55_762843 | 09:45 |
| dsneddon | mhenkel, If it makes sense to adopt kubernetes networking, then an Ingress processor makes sense, but if kubernetes is going to drag its feet on implementing IPv6, better support for routed networks, multiple networks per pod, etc., then who knows what the container networking will look like in a year? | 09:45 |
| dsneddon | mhenkel, But I think the question of how to do HA is very intertwined with that answer, so I think that needs to be a community question that looks at how HA is being done for various implementations of OpenStack on containers. | 09:46 |
| *** achadha has joined #tripleo | 09:46 | |
| mhenkel | dnseddon: with implementing CNI Google opened k8s to the world of SDN controllers | 09:46 |
| *** ccamacho is now known as ccamacho|afk | 09:47 | |
| dsneddon | mhenkel, Yeah, and depending on *which* container SDN solution, some of them would work better with anycast than others, but I'm all for the idea and bring it up often. | 09:47 |
| mhenkel | we’ll see per pod VNs/per app VNs/per Namespace VNs early next year | 09:47 |
| mhenkel | I know at least one which does pretty decent anycast ;) | 09:47 |
| dsneddon | mhenkel, I'm also thinking about how to move away from flat networks for the container hosts. Scale needs routed networks. | 09:48 |
| sshnaidm | jaosorior, seems like connectivity problem.. | 09:48 |
| sshnaidm | jaosorior, but let's see if CI works at all with this range as expected | 09:49 |
| jaosorior | sshnaidm: tried to recheck that job and see if it was a transient issue, or I'm messing up something in the network | 09:49 |
| sshnaidm | jaosorior, can you get some output about networks, dns, routes, etc from this machine? it would be helpful | 09:50 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 09:51 |
| shardy | Folks can I get reviews on https://review.openstack.org/#/q/topic:bug/1629187 please? | 09:53 |
| shardy | two patches both passing CI | 09:53 |
| *** achadha has quit IRC | 09:55 | |
| skramaja | I am working on DPDK nework config update scnario. if I modify network config of the compute node and then deploy (update), the modified network config (/etc/os-net-config/config.json) is not updated to the compute node. while investigating, figured out that heat's NetworkConfig has the updated value, but the swift container (ov-cs2n5ukli4-0-hddx3beb7ytj-NovaCompute-sywpq2tzoern) which serves os-collect-config's (request.json) is not | 09:55 |
| skramaja | updated. I am looking for pinters to debug on which component updates the swift container? | 09:55 |
| skramaja | shardy: ^ | 09:55 |
| *** liverpooler has quit IRC | 09:56 | |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-heat-templates: Include related parameters in environments https://review.openstack.org/401163 | 09:56 |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 09:57 |
| jaosorior | marios, shardy: Could you check this out? https://review.openstack.org/#/c/397350/ it's part of the work to get fernet tokens to TripleO. | 09:57 |
| jaosorior | marios, shardy: It doesn't set them as default yet, first we gotta get more testing into it. But at least it works on new deployments. As tested here https://review.openstack.org/#/c/398897/ | 09:58 |
| *** achadha has joined #tripleo | 10:00 | |
| *** dtantsur is now known as dtantsur|bbl | 10:00 | |
| marios | jaosorior: ack will try get to reviews later dude (noted though) have some ongoing emergencies to deal with :/ and they have to be done today | 10:00 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Deploy heat APIs over httpd https://review.openstack.org/394837 | 10:00 |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 10:02 |
| mhenkel | dsneddon: is there an option today in tripleo to distribute at least compute nodes across different subnets? | 10:02 |
| dsneddon | mhenkel, I'm working on it: https://blueprints.launchpad.net/tripleo/+spec/tripleo-routed-networks-deployment | 10:03 |
| dsneddon | mhenkel, I've done a few custom deployments for very large deployments | 10:04 |
| *** achadha has quit IRC | 10:04 | |
| dsneddon | mhenkel, But it isn't something that can be done in the general case yet. | 10:04 |
| shardy | skramaja: https://github.com/openstack/tripleo-heat-templates/blob/master/overcloud.j2.yaml#L90 | 10:04 |
| shardy | skramaja: you have to set NetworkDeploymentActions: ['CREATE', 'UPDATE'] in an environment file | 10:04 |
| mhenkel | dsneddon: yes, so did we for a joint customer in France using Director 8. But it was a major pain and very custom | 10:05 |
| shardy | skramaja: we only apply the config on deployment because there were some previous issues with bouncing nics on update causing nodes to get fenced | 10:05 |
| dsneddon | mhenkel, It's terrible. But the good news is that most of the functionality is there (Neutron supports DHCP relay, Ironic Inspector does if you manually create the dnsmasq config, and in Pike should do it automatically). | 10:06 |
| dsneddon | mhenkel, But it's a tricky thing to model in Heat, especially with the amount of refactoring going on for composable roles and containers. | 10:06 |
| mhenkel | I agree | 10:07 |
| shardy | dsneddon: I'd be interested to know more about the heat challenges - the refactoring for composable roles hasn't really changed the way we deploy groups of nodes at all | 10:07 |
| shardy | (and I know the way we group nodes is part of the problem, but that's not related to the refactoring) | 10:08 |
| *** achadha has joined #tripleo | 10:09 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Containerized Services for Composable Roles https://review.openstack.org/330659 | 10:09 |
| sshnaidm | jaosorior, can I get your review here please in your time? https://review.openstack.org/#/c/381094/ | 10:09 |
| dsneddon | shardy, Well, it's more that it's been a bit of a moving target, small changes here and there to the endpoint map, handling of VIPs. I was working on a review recently for adding a new network, and I kept having to manually rebase due to other changes. | 10:10 |
| jaosorior | sshnaidm: was just getting into reviews after brewing coffee; sure dude, I'll do it now | 10:11 |
| shardy | dsneddon: Ah, Ok, cool - yeah sorry about that - I just wanted to ensure there weren't more specific concerns re the architecture | 10:11 |
| *** zoliXXL is now known as zoli|lunch | 10:11 | |
| sshnaidm | jaosorior, thanks | 10:11 |
| dsneddon | shardy, I've been meaning to ask you a question about that, actually. | 10:11 |
| sshnaidm | I hope today it'll be merged.. | 10:11 |
| shardy | dsneddon: I actually think the availability of e.g jinja2 will make fully composable networks much easier | 10:11 |
| dsneddon | shardy, So, currently each role has it's own Nova::Server resource, and we pass it the ctlplane network. | 10:12 |
| dsneddon | shardy, For spine-and-leaf, I need to have some finer-grained control over that. I'd actually like to pass it either a fixed IP or even create a Neutron port and pass that. Is there any reason off the top of your head why that wouldn't work? | 10:12 |
| dsneddon | shardy, But eventually I'll probably implement a way to pass just the subnet, then we can create X number of subnets per network, and specify the correct subnets when we create the ports. | 10:15 |
| shardy | dsneddon: Yes I think that would probably work, similar to the approach discussed in https://review.openstack.org/#/c/256003/6/puppet/controller.yaml ? | 10:15 |
| shardy | dsneddon: however I wonder if we might be better off having a custom role per l3 segment, as then you could just have a different network setup per role, instead of mixing network configuration for e.g one giant "Compute" role | 10:16 |
| *** limao has quit IRC | 10:17 | |
| shardy | e.g you could have ComputeRack1 ComputeRack2 or whatever | 10:17 |
| shardy | specified via roles_data.yaml | 10:17 |
| shardy | and each can have a different network setup | 10:17 |
| dsneddon | shardy, Yes. I'm working on putting together a POC that does just that here: https://review.openstack.org/#/c/400423/1/roles_data.yaml | 10:17 |
| *** limao has joined #tripleo | 10:17 | |
| *** achadha has quit IRC | 10:18 | |
| dsneddon | shardy, But that would be a jinja loop eventually | 10:18 |
| shardy | dsneddon: Ah, nice, yeah exactly that | 10:18 |
| dsneddon | shardy, Thanks, I'll work on it some more, but I'm off to bed. | 10:21 |
| shardy | o/ | 10:22 |
| *** limao has quit IRC | 10:22 | |
| *** achadha has joined #tripleo | 10:23 | |
| skramaja | thanks shardy - the scenarios is, none of the other configuration is changed but only the dpdk nic (for provider network) is changed to another nic. since it is known to not work, i will take it as low priorty, have lot of other items pending :) | 10:23 |
| *** liverpooler has joined #tripleo | 10:24 | |
| jaosorior | sshnaidm: reviewed | 10:24 |
| jpich | shardy: Thank you for the comments at https://review.openstack.org/#/c/400966/ , I understand it's not the optimal approach. I'll keep in mind to remove these when they're no longer necessary | 10:25 |
| *** numans has quit IRC | 10:30 | |
| *** achadha has quit IRC | 10:32 | |
| *** numans has joined #tripleo | 10:33 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Fix ovs 2.4 to 2.5 upgrade - minor update non controllers https://review.openstack.org/400925 | 10:35 |
| *** isq has quit IRC | 10:35 | |
| shardy | jpich: thanks - I'm working through the other UI related bugs forwarded now | 10:36 |
| *** achadha has joined #tripleo | 10:36 | |
| *** newmember has quit IRC | 10:37 | |
| jpich | shardy: Thank you, I think the other patches have made good progress in merging, including backports. THT is the main project I was unsure of, I haven't worked on it before and am not familiar with the conventions | 10:38 |
| *** yolanda has quit IRC | 10:41 | |
| *** panda|Zz is now known as panda | 10:43 | |
| openstackgerrit | Marios Andreou proposed openstack/tripleo-heat-templates: Fix ovs 2.4 to 2.5 upgrade - minor update non controllers https://review.openstack.org/401195 | 10:44 |
| *** tosky has joined #tripleo | 10:44 | |
| *** achadha has quit IRC | 10:45 | |
| openstackgerrit | Saravanan KR proposed openstack/os-net-config: Handle deployment update for DPDK nic changes https://review.openstack.org/373680 | 10:49 |
| *** ccamacho|afk is now known as ccamacho | 10:50 | |
| *** udesale has quit IRC | 10:50 | |
| *** achadha has joined #tripleo | 10:50 | |
| *** athomas has quit IRC | 10:52 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 10:56 |
| d0ugal | gfidente: hey | 10:56 |
| gfidente | d0ugal hey | 10:59 |
| d0ugal | gfidente: Can you explain why we reverted this approach? https://github.com/openstack/tripleo-common/commit/ad64050485f79d2330c68b76a3f21b4c0d9a3e33 | 10:59 |
| *** misty-cloud has quit IRC | 10:59 | |
| *** achadha has quit IRC | 11:00 | |
| *** athomas has joined #tripleo | 11:00 | |
| gfidente | d0ugal because on upgrade the password is not there | 11:01 |
| gfidente | fsid actually in that case | 11:01 |
| gfidente | so it would generate a new one, changing it from the one in use | 11:01 |
| d0ugal | gfidente: I think that was an issue with all passwords and it was resolved | 11:02 |
| d0ugal | gfidente: otherwise I think that is a better place to do this IMP | 11:02 |
| d0ugal | IMO* | 11:02 |
| gfidente | d0ugal no the fsid was not there from previous deployments | 11:02 |
| gfidente | we never saved the fsid in there | 11:03 |
| d0ugal | gfidente: it was never passed to Heat? | 11:03 |
| gfidente | it was passed to heat, yes | 11:03 |
| *** liverpooler has quit IRC | 11:03 | |
| d0ugal | gfidente: as the same parameter name? | 11:03 |
| gfidente | same parameter name yes | 11:03 |
| d0ugal | gfidente: then it would be fine. on upgrades, when a plan is created for an existing stack it takes the passwords from heat | 11:03 |
| d0ugal | gfidente: https://github.com/openstack/tripleo-common/commit/ebe270c258d07b9a63d26a477814336188d57d53 | 11:03 |
| gfidente | but not saved in the passwords file | 11:03 |
| gfidente | we did it like this https://github.com/openstack/python-tripleoclient/blob/master/tripleoclient/v1/overcloud_deploy.py#L131 | 11:04 |
| d0ugal | gfidente: Yeah, so that should all work fine - we just removed it with the revert | 11:05 |
| *** achadha has joined #tripleo | 11:05 | |
| gfidente | d0ugal no I think we were not extracting it from heat | 11:05 |
| shardy | https://dashboards.rdoproject.org/rdo-dev | 11:05 |
| d0ugal | gfidente: we would be, if it wasted reverted and removed from the passwords. | 11:05 |
| d0ugal | gfidente: at the moment we wont be | 11:06 |
| shardy | current-tripleo promoted, great work to all who made that happen :) | 11:06 |
| d0ugal | \o/ | 11:06 |
| gfidente | d0ugal so maybe you can explain why this was happening differently | 11:06 |
| d0ugal | gfidente: because passwords was broken for upgrades too | 11:06 |
| d0ugal | gfidente: but I landed a fix for that, it would have fixed your case too AFAIK | 11:06 |
| d0ugal | gfidente: but your revert landed at the same time | 11:07 |
| d0ugal | or similar time | 11:07 |
| gfidente | d0ugal https://bugs.launchpad.net/bugs/1643701 | 11:07 |
| openstack | Launchpad bug 1643701 in tripleo-common ""openstack overcloud update" fails as CephClusterFSID changes" [Undecided,In progress] - Assigned to Giulio Fidente (gfidente) | 11:07 |
| d0ugal | gfidente: Yeah, it's exactly the same problem as with the passwords | 11:08 |
| d0ugal | because it was treated like a password | 11:08 |
| gfidente | so no in mitaka it wasn't | 11:08 |
| gfidente | generate_passwords never generated ut | 11:08 |
| d0ugal | Yeah, I know | 11:09 |
| d0ugal | That's fine. | 11:09 |
| d0ugal | and it wont be generated now, because the parameter is in heat | 11:09 |
| *** akrivoka has quit IRC | 11:09 | |
| openstackgerrit | Harald Jensås proposed openstack/tripleo-heat-templates: Pre Deploy ExtraConfig to install additiona puppet modules. https://review.openstack.org/401202 | 11:09 |
| *** achadha has quit IRC | 11:09 | |
| d0ugal | gfidente: https://github.com/openstack/tripleo-common/blob/master/tripleo_common/utils/passwords.py#L40-L43 | 11:11 |
| d0ugal | gfidente: that is the important bit | 11:11 |
| *** yolanda has joined #tripleo | 11:11 | |
| ccamacho | shardy, morning :) quick question related to the features to land for o-2 (composability related) Im trying to focus about the features we actually need | 11:12 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 11:12 |
| gfidente | d0ugal wait, isn't 45-46 regeneating the keys | 11:12 |
| gfidente | all the times as well? | 11:12 |
| d0ugal | gfidente: elif | 11:12 |
| *** numans has quit IRC | 11:13 | |
| d0ugal | gfidente: it only generates them if the parameter isn't in Heat | 11:13 |
| gfidente | and stack_env is the mistral env? | 11:13 |
| gfidente | or the heat env? | 11:13 |
| d0ugal | gfidente: no, it is the heat env | 11:13 |
| d0ugal | gfidente: yeah, sorry, that is confusing. | 11:13 |
| *** achadha has joined #tripleo | 11:14 | |
| gfidente | d0ugal and what is the fix you mentioned which would have fixed https://bugs.launchpad.net/bugs/1643701 | 11:14 |
| openstack | Launchpad bug 1643701 in tripleo-common ""openstack overcloud update" fails as CephClusterFSID changes" [Undecided,In progress] - Assigned to Giulio Fidente (gfidente) | 11:14 |
| d0ugal | gfidente: https://github.com/openstack/tripleo-common/commit/ebe270c258d07b9a63d26a477814336188d57d53 | 11:14 |
| d0ugal | gfidente: it is that if statement really, but a few other bits needed | 11:15 |
| shardy | ccamacho: the most urgent is composable upgrades, which I've started - I suspect it won't be finished for ocata-2 tho | 11:15 |
| ccamacho | so, when having some cycles, can you point me the blueprint to check? but what I have looked but im having a hard time to split those into items (https://blueprints.launchpad.net/tripleo/+spec/composable-ha and https://blueprints.launchpad.net/tripleo/+spec/overcloud-upgrades-per-service) | 11:15 |
| ccamacho | shardy ^ | 11:15 |
| ccamacho | ack | 11:15 |
| *** b00tcat has quit IRC | 11:15 | |
| gfidente | oh we didn't pull in stack_env | 11:15 |
| gfidente | ? | 11:15 |
| d0ugal | gfidente: I don't follow | 11:16 |
| shardy | ccamacho: https://blueprints.launchpad.net/tripleo/+spec/overcloud-upgrades-per-service is the one I'm referring to, prototype is at https://review.openstack.org/#/c/393448/ | 11:16 |
| gfidente | d0ugal the fix is passing stack_env, we were not | 11:16 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 11:16 |
| gfidente | so we were always regenerating those keys | 11:16 |
| shardy | ccamacho: that one is the highest priority as without it, folks deploying with composable roles on newton cannot upgrade | 11:16 |
| gfidente | and the passwords too | 11:17 |
| d0ugal | gfidente: Yeah, basically | 11:17 |
| jaosorior | gfidente, d0ugal: That's not too bad :P, password rotation :D | 11:17 |
| d0ugal | gfidente: but the generate function didn't have support for that arg before, so it was adding the check | 11:17 |
| shardy | ccamacho: feedback/help welcome! We can discuss ways to divide the work if you have bandwidth to get involved :) | 11:17 |
| d0ugal | jaosorior: lol, if only it worked :-D | 11:17 |
| gfidente | d0ugal so I guess we can re-apply the revert | 11:18 |
| d0ugal | gfidente: Yeah, I think so | 11:18 |
| gfidente | for ocata? | 11:18 |
| d0ugal | gfidente: and newton I would guess? | 11:18 |
| gfidente | and even remove the 'on stack_is_new' parts | 11:18 |
| d0ugal | gfidente: because I see it was reverted in Newton | 11:18 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Add dependencies for UI internationalisation https://review.openstack.org/401205 | 11:19 |
| gfidente | d0ugal yes we can backport it | 11:19 |
| gfidente | and remove the stack_is_new parts too | 11:19 |
| d0ugal | gfidente: Yeah, I think we can revert the revert and also revert https://github.com/openstack/tripleo-common/commit/20167e850a6bcb1fbb1bfa8cb030000600c3dc27 | 11:20 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 11:20 |
| d0ugal | gfidente: yo dawg, I heard you like reverts etc. | 11:20 |
| gfidente | d0ugal frankly | 11:20 |
| gfidente | if only it worked | 11:20 |
| gfidente | I wouldn't have | 11:20 |
| d0ugal | Indeed | 11:20 |
| d0ugal | It was a massive problem | 11:20 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 11:20 |
| d0ugal | for all the auto generated things | 11:21 |
| *** gchamoul is now known as gchamoul|afk | 11:21 | |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: DO NOT MERGE Transition to quickstart: simulate transition https://review.openstack.org/399257 | 11:21 |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: Transition to quickstart: add toci_gate_oooq.sh and jobs configurations https://review.openstack.org/399256 | 11:21 |
| gfidente | d0ugal fine you decide the timeframe for the patches | 11:22 |
| gfidente | I can help with votes | 11:22 |
| gfidente | and tests | 11:22 |
| d0ugal | gfidente: okay, I'll get a patch up shortly. | 11:22 |
| jpich | d0ugal: your backport was merged a couple of weeks ago though, while this bug was reported only a couple of days ago - it sounds like it should have been included in the packages the initial reporter used...? | 11:23 |
| d0ugal | jpich: Yeah, that bit does confuse me. | 11:24 |
| ccamacho | shardy, awesome, /me just checking the patch, yeahp that would be cool, to be able to split it into pieces/features to finish it asap | 11:24 |
| *** fzdarsky_ is now known as fzdarsky | 11:24 | |
| *** fzdarsky is now known as fzdarsky|lunch | 11:24 | |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::agent class https://review.openstack.org/400618 | 11:24 |
| d0ugal | jpich: I could be totally missing something | 11:24 |
| *** gchamoul|afk is now known as gchamoul | 11:24 | |
| gfidente | d0ugal jpich or it could be the build did not include that patch | 11:25 |
| shardy | ccamacho: sounds good | 11:25 |
| d0ugal | gfidente, jpich: Yeah, I'll try and test once I get this patch together. | 11:25 |
| shardy | ccamacho: I'm working on something else atm, but we can sync up this afternoon | 11:25 |
| shardy | ccamacho: the two big pieces to tackle next are how to handle services that switch from eventlet to httpd on upgrade | 11:25 |
| shardy | and how to refactor all the pacemaker stuff into the new model | 11:26 |
| jpich | d0ugal: Let me know when the patches are up, I'll help with testing as I can | 11:26 |
| shardy | I've got ideas on both, but other opinions on how to handle it will be welcome | 11:26 |
| jpich | d0ugal, gfidente: I'll ask on the initial bug about versions for tripleo-common | 11:26 |
| *** Guest83036 has quit IRC | 11:26 | |
| *** sudipto_ has quit IRC | 11:26 | |
| *** achadha has quit IRC | 11:27 | |
| ccamacho | shardy awesome, just let me know when you have some time to sync this afternoon and we can push that, Ill start taking notes and check code | 11:27 |
| *** achadha has joined #tripleo | 11:28 | |
| *** b00tcat has joined #tripleo | 11:29 | |
| *** b00tcat has quit IRC | 11:29 | |
| *** b00tcat has joined #tripleo | 11:29 | |
| openstackgerrit | Merged openstack/tripleo-quickstart-extras: collect-logs: internal rename and cleanup https://review.openstack.org/400929 | 11:30 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Add dependencies for UI internationalisation https://review.openstack.org/401205 | 11:30 |
| gfidente | d0ugal going to the 'user provided' value | 11:32 |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401212 | 11:32 |
| gfidente | on initial deployment I think this will always generate a password? | 11:32 |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401212 | 11:32 |
| d0ugal | gfidente: yes, but it can be overwritten - that is in a slightly different place. One sec | 11:33 |
| *** zoli|lunch is now known as zoli | 11:33 | |
| *** zoli is now known as zoliXXL | 11:33 | |
| gfidente | ack was looking at where/how it's overridden | 11:33 |
| d0ugal | gfidente: https://github.com/openstack/tripleo-common/blob/master/tripleo_common/actions/parameters.py#L164-L174 | 11:34 |
| *** ccamacho is now known as ccamacho|lunch | 11:34 | |
| d0ugal | gfidente: I'll double check that too. | 11:34 |
| shardy | gfidente: is there an existing launchpad bug for the merge_strategies problem? | 11:35 |
| gfidente | d0ugal yeah I wanted to test that, so that is checking against the mistral env though? | 11:35 |
| gfidente | shardy https://bugs.launchpad.net/bugs/1635409 | 11:35 |
| openstack | Launchpad bug 1635409 in tripleo "parameter_merge_strategies: merge behaves like overwrite" [High,Fix released] - Assigned to Giulio Fidente (gfidente) | 11:35 |
| *** milan has joined #tripleo | 11:36 | |
| gfidente | and don't know why it's assigned tome | 11:36 |
| *** dtantsur|bbl is now known as dtantsur | 11:36 | |
| gfidente | ah the backport | 11:36 |
| shardy | Ok so we need to reopen it as the TripleO fix was incomplete, or create a new one | 11:37 |
| shardy | thanks | 11:37 |
| shardy | I'm testing a fix now | 11:37 |
| d0ugal | gfidente, jpich: So, I think this should do it but I am just reviewing and testing it now too. https://review.openstack.org/#/c/401212/ | 11:37 |
| *** prateek has quit IRC | 11:37 | |
| openstackgerrit | Harald Jensås proposed openstack/tripleo-heat-templates: Pre Deploy ExtraConfig to install additiona puppet modules. https://review.openstack.org/401202 | 11:38 |
| *** cylopez has quit IRC | 11:40 | |
| *** achadha has quit IRC | 11:41 | |
| *** numans has joined #tripleo | 11:42 | |
| gfidente | d0ugal ack I am trying passing a custom value first | 11:45 |
| d0ugal | gfidente: I don't know how to test a full upgrade, but I am going to deploy and manually delete the plan | 11:46 |
| *** achadha has joined #tripleo | 11:46 | |
| d0ugal | gfidente: then create the plan again, to simulate getting the params back out of Heat | 11:46 |
| gfidente | d0ugal so I think the initial reproducer was deploy/update | 11:46 |
| jpich | d0ugal: Thanks, I'll pull it in my dev env and run a couple of tests from the UI too, though I guess we're back to the "way it used to work" so it should be fine :) | 11:47 |
| gfidente | which from what I can tell would only regenerate the pass if stack_env was not passed (aka, your patch was missing) | 11:47 |
| *** achadha has quit IRC | 11:51 | |
| *** liverpooler has joined #tripleo | 11:51 | |
| gfidente | mwhahaha if I wanted to add some params for qemu.conf in puppet-nova | 11:52 |
| openstackgerrit | Merged openstack/tripleo-docs: Fix minor doc errors https://review.openstack.org/363660 | 11:53 |
| gfidente | mwhahaha would you edit libvirt.pp or have a class in libvirt/ ? or anything else ? | 11:53 |
| openstackgerrit | Merged openstack/tripleo-docs: Replace mention of 'liberty' in directory path with '*' https://review.openstack.org/392597 | 11:53 |
| gfidente | EmilienM ^^ or any puppet guy around :) | 11:53 |
| *** ansmith has joined #tripleo | 11:54 | |
| yolanda | hi derekh , around? i'm having trouble again testing diskimage-builder branch | 11:55 |
| *** achadha has joined #tripleo | 11:55 | |
| *** yolanda has quit IRC | 11:55 | |
| gfidente | d0ugal, jpich it works for me with user-provided values | 11:55 |
| gfidente | d0ugal trying the stack-update | 11:55 |
| *** yolanda has joined #tripleo | 11:55 | |
| yolanda | i have this change https://review.openstack.org/#/c/400888/9 | 11:56 |
| yolanda | and it has a depends-on on a diskimage-builder change for feature/v2 branch | 11:56 |
| d0ugal | gfidente, gfidente: Yeah, it worked for me too with a custom value in an env file. I am waiting for my deploy to finish. Then I'll (manually) delete the plan and do a stack update (without the env file) - that'll be similar to how an upgrade will look. | 11:57 |
| yolanda | however, i holded a node, and i am looking at the content for delorean in diskimage-builder | 11:57 |
| yolanda | and actually looks as the mitaka content... | 11:57 |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: DO NOT MERGE Transition to quickstart: simulate transition https://review.openstack.org/399257 | 11:58 |
| *** akrivoka has joined #tripleo | 11:58 | |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: Transition to quickstart: add toci_gate_oooq.sh and jobs configurations https://review.openstack.org/399256 | 11:58 |
| gfidente | d0ugal manually means you delete the contents from swift only? | 11:58 |
| d0ugal | gfidente: swift and mistral | 11:58 |
| gfidente | oh but not the heat stack | 11:58 |
| d0ugal | gfidente: I say manual, because "openstack overcloud plan delete" refuses to work if there is a stack | 11:58 |
| d0ugal | gfidente: yeah | 11:58 |
| gfidente | ok got it | 11:58 |
| *** bkopilov has quit IRC | 11:58 | |
| *** achadha has quit IRC | 12:00 | |
| openstackgerrit | Michele Baldessari proposed openstack/tripleo-heat-templates: WIP DO NOT MERGE pacemaker remote profile support https://review.openstack.org/400972 | 12:00 |
| openstackgerrit | Tuan Luong-Anh proposed openstack/diskimage-builder: Replace six.iteritems() with .items() https://review.openstack.org/400514 | 12:03 |
| shardy | gfidente: ugh, this is harder than I expected, we do bad things related to user-environment.yaml in mistral | 12:03 |
| *** openstackgerrit has quit IRC | 12:03 | |
| * shardy tries to fix | 12:03 | |
| d0ugal | shardy: What are you working on? | 12:03 |
| *** openstackgerrit has joined #tripleo | 12:03 | |
| d0ugal | out of interest | 12:03 |
| shardy | https://bugs.launchpad.net/heat/+bug/1635409 | 12:03 |
| openstack | Launchpad bug 1635409 in tripleo "parameter_merge_strategies: merge behaves like overwrite" [High,In progress] - Assigned to Steven Hardy (shardy) | 12:03 |
| *** liverpooler has quit IRC | 12:04 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 12:04 |
| shardy | d0ugal: currently we merge the environment files in mistral, mangle part of the result into user-environment, and put the rest e.g parameter_defaults into the mistral environment | 12:04 |
| shardy | that doesn't work with heat parameter merging | 12:04 |
| d0ugal | ah | 12:04 |
| shardy | we need to maintain the list of user environments e.g -e one.yaml -e two.yaml | 12:04 |
| shardy | and pass that list to heat | 12:04 |
| d0ugal | Right, we break the order. | 12:04 |
| shardy | I commented about this when the workflows got implemented, but somehow we didn't quite get it sorted | 12:05 |
| *** prateek has joined #tripleo | 12:05 | |
| gfidente | shardy okay so the merging is happening in the client itself | 12:06 |
| shardy | gfidente: yeah, I need to remove the code that does that and pass the list to heat instead | 12:06 |
| *** prateek has quit IRC | 12:07 | |
| shardy | there is also a t-h-t change required to add merge strategies to the overcloud-resource-registry-puppet where the default *Services lists are defined | 12:07 |
| jpich | Is there a way to create an additional "baremetal" VM with instack-virt-setup or something like that later on? One of my nodes has become unusable, I'm hoping I could create a new one and delete the failing one without needing to reset my whole environment... | 12:07 |
| *** prateek has joined #tripleo | 12:08 | |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-heat-templates: Include related parameters in environments https://review.openstack.org/401163 | 12:08 |
| *** achadha has joined #tripleo | 12:09 | |
| *** jcoufal has joined #tripleo | 12:09 | |
| *** prateek has quit IRC | 12:10 | |
| openstackgerrit | Sven Anderson proposed openstack/tripleo-heat-templates: Add option for deploying ssh keys to nova users on compute nodes https://review.openstack.org/400942 | 12:11 |
| *** spredzy has quit IRC | 12:11 | |
| *** akrivoka has quit IRC | 12:12 | |
| *** sudipto_ has joined #tripleo | 12:13 | |
| *** achadha has quit IRC | 12:13 | |
| *** Guest83036 has joined #tripleo | 12:13 | |
| *** ansmith has quit IRC | 12:14 | |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Add dependencies for UI internationalisation https://review.openstack.org/401205 | 12:14 |
| *** akrivoka has joined #tripleo | 12:15 | |
| *** spredzy has joined #tripleo | 12:15 | |
| d0ugal | gfidente, jpich: all seems to be working. I added my testing notes to: | 12:16 |
| d0ugal | https://review.openstack.org/#/c/401212/ | 12:16 |
| *** prateek has joined #tripleo | 12:17 | |
| *** prateek has quit IRC | 12:17 | |
| d0ugal | bbiab, walking the dog | 12:17 |
| jpich | d0ugal: Looking good, no answer on package versions on the downstream bug yet | 12:17 |
| d0ugal | jpich: k, thanks. | 12:17 |
| *** achadha has joined #tripleo | 12:18 | |
| *** prateek has joined #tripleo | 12:18 | |
| openstackgerrit | Sagi Shnaidman proposed openstack-infra/tripleo-ci: Full quickstart gate run on OVB https://review.openstack.org/381094 | 12:19 |
| openstackgerrit | Sagi Shnaidman proposed openstack-infra/tripleo-ci: DONT MERGE: Test quickstart experimental job https://review.openstack.org/397609 | 12:19 |
| *** abregman is now known as abregman|mtg | 12:21 | |
| openstackgerrit | Sagi Shnaidman proposed openstack-infra/tripleo-ci: Full quickstart gate run on OVB https://review.openstack.org/381094 | 12:21 |
| openstackgerrit | Sagi Shnaidman proposed openstack-infra/tripleo-ci: DONT MERGE: Test quickstart experimental job https://review.openstack.org/397609 | 12:21 |
| *** lucasagomes is now known as lucas-hungry | 12:21 | |
| sshnaidm | derekh, hi | 12:23 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 12:23 |
| derekh | yolanda: what mitaka content is on it? | 12:23 |
| derekh | sshnaidm: hi | 12:23 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Adds basic internationalization support https://review.openstack.org/399626 | 12:24 |
| *** prateek has quit IRC | 12:24 | |
| sshnaidm | derekh, have you seen ever error like "AttributeError: 'HTTPClient' object has no attribute 'last_request_id'" in nova logs? I have it sometimes in dev env | 12:24 |
| derekh | sshnaidm: nope, never seen it | 12:25 |
| *** prateek has joined #tripleo | 12:25 | |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: DO NOT MERGE Transition to quickstart: simulate transition https://review.openstack.org/399257 | 12:25 |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: Transition to quickstart: add toci_gate_oooq.sh and jobs configurations https://review.openstack.org/399256 | 12:25 |
| *** pgadiya has quit IRC | 12:26 | |
| sshnaidm | derekh, it happens when I use last image from glance: template-centos-7-1479743400 , but if using template-centos-7-1478512463 for example - all works like a charm | 12:30 |
| EmilienM | hi | 12:31 |
| sshnaidm | derekh, any ideas how it could be related?? | 12:31 |
| sshnaidm | EmilienM, hi, thanks for reviews! | 12:31 |
| *** achadha has quit IRC | 12:32 | |
| derekh | sshnaidm: is anything installed from pip on the image? maybe something change version | 12:32 |
| *** hjensas has quit IRC | 12:32 | |
| *** thrash|g0ne is now known as thrash | 12:32 | |
| sshnaidm | derekh, but it fails on "nova boot --image template-centos...." step, can not create a server | 12:34 |
| derekh | sshnaidm: you mean you can't create the undercloud node at all | 12:35 |
| derekh | ? | 12:35 |
| derekh | sshnaidm: on rh1? | 12:35 |
| sshnaidm | derekh, yep | 12:36 |
| sshnaidm | derekh, something wrong with this image.. | 12:36 |
| *** achadha has joined #tripleo | 12:37 | |
| *** gchamoul is now known as gchamoul|afk | 12:38 | |
| *** gchamoul|afk is now known as gchamoul | 12:38 | |
| derekh | sshnaidm: the image has a queued status, looks like it didn't finish uploading | 12:38 |
| derekh | | 65b9fc04-1ab9-409c-bb66-79cd571050fe | template-centos-7-1479743400 | queued | | 12:38 |
| *** gchamoul is now known as gchamoul|afk | 12:38 | |
| *** shardy is now known as shardy_lunch | 12:39 | |
| *** weshay_ has quit IRC | 12:40 | |
| *** akrivoka has quit IRC | 12:40 | |
| sshnaidm | derekh, thanks, I see now.. I just try to boot sometimes from non ready image | 12:41 |
| sshnaidm | will take one before last then | 12:41 |
| derekh | sshnaidm: ok | 12:41 |
| *** hjensas has joined #tripleo | 12:46 | |
| *** fzdarsky|lunch is now known as fzdarsky | 12:47 | |
| *** gchamoul|afk is now known as gchamoul | 12:48 | |
| *** achadha has quit IRC | 12:51 | |
| openstackgerrit | yolanda.robla proposed openstack/instack-undercloud: Add new ELEMENTS_PATH to work with dib feature/v2 branch https://review.openstack.org/400888 | 12:52 |
| yolanda | derekh, can you look at conversation with jpena on rdo channel? just talking about it now | 12:53 |
| derekh | yolanda: ok | 12:53 |
| *** links has quit IRC | 12:54 | |
| *** dprince has joined #tripleo | 12:56 | |
| *** achadha has joined #tripleo | 12:59 | |
| *** egafford has joined #tripleo | 13:01 | |
| *** tzumainn has joined #tripleo | 13:02 | |
| *** jayg|g0n3 is now known as jayg | 13:04 | |
| *** cylopez has joined #tripleo | 13:05 | |
| openstackgerrit | Merged openstack/tripleo-ui: Refactor addWorkflowExecution https://review.openstack.org/394434 | 13:06 |
| *** ccamacho|lunch is now known as ccamacho | 13:08 | |
| rook | derekh: hey | 13:09 |
| *** morazi has joined #tripleo | 13:10 | |
| derekh | rook: howdy | 13:12 |
| rook | derekh: ever have a ovb stack create get stuck? | 13:12 |
| rook | heat stack-delete won't remove it? | 13:12 |
| *** adarazs has quit IRC | 13:12 | |
| derekh | rook: I have in the passed had to issue the heat stack-delete multiple times (after it failed), | 13:13 |
| *** noslzzp has joined #tripleo | 13:13 | |
| *** achadha has quit IRC | 13:13 | |
| rook | derekh: yeah, i am doing the same thing | 13:13 |
| rook | derekh: on the failed stack, i saw : https://gist.github.com/jtaleric/42eebb9d0891c3f2faeca1bdd09f5aab | 13:14 |
| derekh | rook: once that I remember I've had to manually delete all of the stacks resources and then heat stack-delete it afterwards | 13:14 |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Use kwargs to pass in data and error to Mistral Result https://review.openstack.org/375348 | 13:14 |
| openstackgerrit | Merged openstack/tripleo-specs: Composable Service Upgrades https://review.openstack.org/392116 | 13:14 |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Only start the deploy if the Heat stack isn't already in progress https://review.openstack.org/396567 | 13:14 |
| *** morazi has quit IRC | 13:14 | |
| derekh | rook: so nova delete, neutron port/subnet/net delete etc... , then finally stack delete | 13:15 |
| derekh | rook: never seen that db error | 13:15 |
| openstackgerrit | Merged openstack/tripleo-ui: Include the plan name on node assignment https://review.openstack.org/400949 | 13:15 |
| derekh | rook: brb | 13:15 |
| openstackgerrit | Merged openstack/tripleo-ui: Add pxe_drac to the node registration driver list https://review.openstack.org/400845 | 13:15 |
| *** noslzzp has quit IRC | 13:15 | |
| openstackgerrit | Merged openstack/tripleo-specs: Add a tag specific to documentation issues https://review.openstack.org/399005 | 13:15 |
| *** colonwq_ has joined #tripleo | 13:15 | |
| *** noslzzp has joined #tripleo | 13:16 | |
| *** colonwq has quit IRC | 13:16 | |
| *** colonwq_ has quit IRC | 13:16 | |
| *** fultonj has joined #tripleo | 13:16 | |
| *** colonwq has joined #tripleo | 13:16 | |
| *** adarazs has joined #tripleo | 13:18 | |
| *** akrivoka has joined #tripleo | 13:18 | |
| *** achadha has joined #tripleo | 13:18 | |
| rook | rlandy ovb should work with downstream right? | 13:19 |
| *** Guest59528 has quit IRC | 13:19 | |
| rlandy | rook: it does - we run a full set of gates with ovb | 13:19 |
| rook | rlandy: hm | 13:20 |
| rook | ok | 13:20 |
| rlandy | rook: sending you link in pvt | 13:20 |
| *** amoralej is now known as amoralej|lunch | 13:20 | |
| *** lucas-hungry is now known as lucasagomes | 13:23 | |
| openstackgerrit | Merged openstack/tripleo-docs: Add note with boot config image options https://review.openstack.org/336078 | 13:24 |
| openstackgerrit | Merged openstack/tripleo-validations: Updated from global requirements https://review.openstack.org/400916 | 13:25 |
| *** hoobaman has quit IRC | 13:26 | |
| derekh | rook: also, if your setting this up, I recommend applying the nova patch so baremetal nodes can be reused with being reimaged https://github.com/cybertron/openstack-virtual-baremetal/blob/master/patches/nova/nova-pxe-boot.patch | 13:26 |
| *** newmember has joined #tripleo | 13:26 | |
| *** prateek_ has joined #tripleo | 13:26 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-quickstart: Fix devmode for change to TripleO produced images https://review.openstack.org/400931 | 13:28 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-quickstart: Remove openstack-dashboard in convert_image.sh https://review.openstack.org/401259 | 13:28 |
| jaosorior | panda: ^^ | 13:28 |
| *** achadha has quit IRC | 13:28 | |
| *** morazi has joined #tripleo | 13:28 | |
| *** shardy_lunch is now known as shardy | 13:28 | |
| *** yolanda has quit IRC | 13:29 | |
| rook | derekh: that is on the baremetal nova compute nodes? | 13:29 |
| rook | or the undercloud ndoe | 13:29 |
| derekh | rook: on the ovb nova cloud that you are using | 13:30 |
| *** prateek has quit IRC | 13:30 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/tripleo-docs: Make it clear that removing ironic-inspector.sqlite is for Liberty only https://review.openstack.org/377917 | 13:31 |
| *** tzumainn has quit IRC | 13:31 | |
| derekh | rook: so e.g. we use RH1 to run ci tests on upstream, I've applied this patch to all of the compute nodes on RH1 | 13:31 |
| *** Guest83036 has quit IRC | 13:32 | |
| *** sudipto_ has quit IRC | 13:32 | |
| gfidente | EmilienM I'll W+1 d0ugal's change | 13:32 |
| gfidente | it needs to be backported | 13:32 |
| EmilienM | gfidente: ack, thx | 13:33 |
| *** tzumainn has joined #tripleo | 13:33 | |
| EmilienM | can someone review https://review.openstack.org/398584 please? | 13:34 |
| *** abregman|mtg is now known as abregman | 13:35 | |
| *** tzumainn has quit IRC | 13:35 | |
| *** tzumainn has joined #tripleo | 13:35 | |
| *** saneax-_-|AFK is now known as saneax | 13:36 | |
| *** achadha has joined #tripleo | 13:36 | |
| openstackgerrit | Merged openstack/tripleo-validations: validation_output.py: Provided output for ansible debug module. https://review.openstack.org/397177 | 13:37 |
| *** masco has quit IRC | 13:37 | |
| d0ugal | gfidente: shouldn't we wait for the ovb gates? | 13:38 |
| gfidente | d0ugal are you doing the backport? | 13:39 |
| d0ugal | gfidente: Yeah | 13:39 |
| *** ansmith has joined #tripleo | 13:40 | |
| panda | folks please, either vote on this https://review.openstack.org/382515 or comment on this http://lists.openstack.org/pipermail/openstack-dev/2016-October/105934.html so we can find solution suitable for everyone | 13:40 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::nsdb class https://review.openstack.org/400869 | 13:41 |
| EmilienM | jpich: do we have an update on removing postconfig things? i remember it was part of our actions to take during our design sessions | 13:41 |
| jaosorior | EmilienM: could you check this out? https://review.openstack.org/#/c/397350/ | 13:45 |
| *** jpena is now known as jpena|lunch | 13:45 | |
| EmilienM | jaosorior: oh nice | 13:45 |
| *** dmsimard|away is now known as dmsimard | 13:45 | |
| EmilienM | jaosorior: I remember my first implementation for keystone credentials had some issues and needed a second iteration | 13:46 |
| EmilienM | jaosorior: it was a problem with upgrades iirc | 13:46 |
| jaosorior | EmilienM: this doesn't set it up as a default yet, we hope to get some good testing of it before doing that | 13:46 |
| *** prateek_ has quit IRC | 13:46 | |
| openstackgerrit | Michele Baldessari proposed openstack/tripleo-heat-templates: WIP DO NOT MERGE pacemaker remote profile support https://review.openstack.org/400972 | 13:47 |
| *** prateek has joined #tripleo | 13:47 | |
| jaosorior | EmilienM: but yeah, at least on fresh deployments it seems fine https://review.openstack.org/#/c/398897/ | 13:48 |
| EmilienM | jaosorior: sec, let me review it | 13:49 |
| EmilienM | it's on my list | 13:49 |
| *** achadha has quit IRC | 13:50 | |
| EmilienM | jaosorior: this https://review.openstack.org/#/c/374892/ | 13:50 |
| *** achadha has joined #tripleo | 13:51 | |
| *** derekh is now known as derekh_afk | 13:51 | |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401267 | 13:51 |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart-extras: [WIP] add the containers prep role to the quickstart-extras https://review.openstack.org/400983 | 13:52 |
| EmilienM | jaosorior: I'm trying to figure if fernet & credentials are consistent in the code | 13:52 |
| *** b00tcat has quit IRC | 13:52 | |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart-extras: deployment updates for containerized compute https://review.openstack.org/400986 | 13:53 |
| jaosorior | ayoung: ^^ | 13:53 |
| EmilienM | jaosorior: but yes | 13:53 |
| EmilienM | it re-use what we did for credentials iiuc | 13:53 |
| EmilienM | in https://review.openstack.org/#/c/397381/2/tripleo_common/utils/passwords.py | 13:54 |
| EmilienM | and ayoung abandoned the patch for oooclient to have it in tripleo-common, it makes sense | 13:54 |
| openstackgerrit | Merged openstack/puppet-tripleo: Move calculation of neutron l3_ha into puppet profile https://review.openstack.org/398926 | 13:54 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Add dependencies for UI internationalisation https://review.openstack.org/401205 | 13:54 |
| panda | we're not using horizon at all on the undercloud, right ? | 13:54 |
| EmilienM | panda: no | 13:54 |
| EmilienM | jaosorior: +A https://review.openstack.org/#/c/397350/ | 13:55 |
| jaosorior | EmilienM: thanks for checking it out | 13:55 |
| *** achadha has quit IRC | 13:55 | |
| jpich | EmilienM: I wasn't at Summit, I think akrivoka was looking into it and from what I recall, it turned out everything had been migrated already? | 13:55 |
| *** abregman has quit IRC | 13:55 | |
| EmilienM | jaosorior: next step is to make the switch to use it by default? | 13:56 |
| EmilienM | jaosorior: this step is not required afik ^ | 13:56 |
| jaosorior | EmilienM: yep, that's the next step | 13:56 |
| EmilienM | jpich: right, I'll check in tripleoclient if we still have postconfig thing, and if yes maybe dropping it | 13:56 |
| panda | jaosorior: https://review.openstack.org/401259 is doing something that you already changed in https://review.openstack.org/400931 | 13:56 |
| akrivoka | EmilienM: my patch to remove the postconfig from the client was -2'd | 13:56 |
| jaosorior | EmilienM: I think bnemec wanted to leave a bit more time before we remove the postconfig stuff | 13:56 |
| *** prateek_ has joined #tripleo | 13:57 | |
| akrivoka | EmilienM: https://review.openstack.org/#/c/392302/ | 13:57 |
| jaosorior | panda: ?? | 13:57 |
| jaosorior | panda: one commit is over the other | 13:57 |
| akrivoka | EmilienM: see also bug https://bugs.launchpad.net/tripleo/+bug/1634992 | 13:57 |
| openstack | Launchpad bug 1634992 in tripleo "Remove post-deployment config from the tripleoclient" [Medium,Invalid] - Assigned to Ana Krivokapić (akrivoka) | 13:57 |
| EmilienM | oops ok I missed that | 13:57 |
| *** udesale has joined #tripleo | 13:58 | |
| panda | jaosorior: https://review.openstack.org/401259/1 is already included in https://review.openstack.org/400931/4 | 13:58 |
| jaosorior | EmilienM: see from Oct 20 https://review.openstack.org/#/c/389173/ | 13:58 |
| EmilienM | akrivoka, I guess we'll re-itere in a future cycle | 13:58 |
| *** udesale has quit IRC | 13:58 | |
| *** bkopilov has joined #tripleo | 13:59 | |
| EmilienM | jaosorior: makes sense | 13:59 |
| jaosorior | panda: yes, that's because https://review.openstack.org/400931/4 is on top of https://review.openstack.org/401259/1 | 13:59 |
| akrivoka | EmilienM: yeah I guess we just need to keep it for a while for backwards compatibility | 13:59 |
| *** udesale has joined #tripleo | 13:59 | |
| EmilienM | ++ | 13:59 |
| *** prateek has quit IRC | 13:59 | |
| *** ramishra has quit IRC | 13:59 | |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart: config for containerized-compute https://review.openstack.org/393348 | 13:59 |
| *** udesale has quit IRC | 13:59 | |
| *** udesale has joined #tripleo | 14:00 | |
| *** achadha has joined #tripleo | 14:00 | |
| panda | jaosorior: oh, you've rebased an older change on top of your new change. O_o ... ok | 14:00 |
| pradk | could i request reviews on https://review.openstack.org/#/c/396439/ please | 14:00 |
| jaosorior | panda: wait, was it an older commit?? | 14:01 |
| jaosorior | panda: I just cherry-picked it, what the hell | 14:01 |
| *** ramishra has joined #tripleo | 14:01 | |
| jaosorior | panda: what do you mean an older change? I see no major differences between patchset 3 and 4 from trown's patch :/ I'm getting pretty confused here. | 14:03 |
| jaosorior | panda: only difference is the fact that it's on top of my commit removing the dashboard | 14:03 |
| panda | EmilienM: uncompressed journal log in my system can reach 1G, that's why I was compressing it. Also my browser is usually able to open gz files ... is there some mechanism that is added to txt files to make them more readable ? | 14:03 |
| panda | jaosorior: I mean, trown patch existed before yours. When I usually add something new to an existing patch, I rebase mine on top of the existing. That's why I got confused, you're rebasing and existing patch on top of your new one | 14:05 |
| jaosorior | ah, right, yeah | 14:05 |
| *** abregman has joined #tripleo | 14:07 | |
| EmilienM | panda: mhh ok | 14:11 |
| openstackgerrit | Gabriele Cerami proposed openstack/tripleo-quickstart-extras: prep-network: add support to IPv6 topology https://review.openstack.org/400837 | 14:13 |
| openstackgerrit | Merged openstack/tripleo-ui: Missing comma in sample could be confusing https://review.openstack.org/399168 | 14:13 |
| panda | EmilienM: for example, that magic script that makes timestamps clickable, is working with gz files ? | 14:13 |
| jaosorior | EmilienM: also, could you check this? https://review.openstack.org/#/c/398177/ | 14:14 |
| *** pcaruana has quit IRC | 14:14 | |
| EmilienM | panda: I don't know why it didn't work for me | 14:15 |
| EmilienM | jaosorior: yes | 14:15 |
| *** [3]cdearborn has joined #tripleo | 14:15 | |
| *** b00tcat has joined #tripleo | 14:16 | |
| *** b00tcat has quit IRC | 14:16 | |
| *** b00tcat has joined #tripleo | 14:16 | |
| EmilienM | jaosorior: nice feature! | 14:17 |
| ayoung | EmilienM, there is one patch still outstanding for THT I think: .... and you just +2ed it | 14:17 |
| panda | jaosorior: not working for me either ... renaming to txt.gz | 14:17 |
| ayoung | excellent. We will have to deal with rotation at some point, but that can wait. | 14:17 |
| jaosorior | ayoung: that was exactly what we were talking about | 14:17 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::nsdb class https://review.openstack.org/400869 | 14:18 |
| pradk | can i get reviews on https://review.openstack.org/#/c/398579/ and https://review.openstack.org/#/c/398580/ please* | 14:18 |
| *** anton has joined #tripleo | 14:18 | |
| ayoung | Once that merges, we can talk about what we want for testing and for moving toward making it default | 14:18 |
| *** achadha has quit IRC | 14:18 | |
| *** parora has joined #tripleo | 14:18 | |
| openstackgerrit | Gabriele Cerami proposed openstack-infra/tripleo-ci: logging: extract journal to compressed text file, before gathering https://review.openstack.org/400817 | 14:19 |
| *** abregman is now known as abregman|mtg | 14:19 | |
| owalsh | bnemec: would you mind re-adding +2 on https://review.openstack.org/388162 | 14:20 |
| EmilienM | ayoung: for rotations do you have ideas? | 14:20 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::nsdb class https://review.openstack.org/400869 | 14:20 |
| ayoung | EmilienM, well, sort of, but I don't know heat well enough to know if they are reasonable | 14:20 |
| EmilienM | ayoung: I thought we could use barbican at some point, so store secrets | 14:21 |
| ayoung | EmilienM, what we do right now is store the keys in Heat. What should happen, then, is that Heat should automate changing the Keys at some schedule | 14:21 |
| *** yolanda has joined #tripleo | 14:21 | |
| jaosorior | EmilienM: we do need a secret store, for many reasons, but it's still unclear if it should be barbican | 14:21 |
| ayoung | I think Barbican would complicate matters for now. | 14:21 |
| *** prateek has joined #tripleo | 14:21 | |
| ayoung | there is a circular dependency between Barbican and Keystone that I am not ready to address | 14:22 |
| jaosorior | the overcloud nodes would need an identity and credentials to access an undercloud instance of barbican | 14:22 |
| EmilienM | ayoung: if you've seen what dprince is doing, the undercloud is going to become something ephemeral, so I don't think we should use the undercloud to rotate overcloud keystone fernet keys | 14:22 |
| *** prateek_ has quit IRC | 14:22 | |
| EmilienM | jaosorior: I would use the overcloud itself | 14:22 |
| jaosorior | EmilienM: aah, you're talking about that | 14:22 |
| *** amoralej|lunch is now known as amoralej | 14:22 | |
| jaosorior | it does become pretty complicated as ayoung points out | 14:22 |
| EmilienM | I'm still sad to see Keystone has not resolved this problem within their software | 14:22 |
| EmilienM | it's not providing good UX here | 14:23 |
| *** achadha has joined #tripleo | 14:23 | |
| jaosorior | barbican uses keystone... it's gonna be pretty complicated if keystone would in turn depend on barbican | 14:23 |
| jaosorior | ayoung: do you think custodia for the overcloud would address this? | 14:23 |
| *** parora has quit IRC | 14:24 | |
| *** jmiu has quit IRC | 14:24 | |
| ayoung | jaosorior, maybe, but we still need to seed it with something. I think it is a Heat problem to solve. | 14:24 |
| ayoung | Something like this | 14:24 |
| EmilienM | it's a bit the same problem as managing SSL certs here | 14:24 |
| EmilienM | and making sure they are synced | 14:24 |
| EmilienM | ayoung: no | 14:24 |
| *** prateek_ has joined #tripleo | 14:24 | |
| *** gchamoul has quit IRC | 14:24 | |
| EmilienM | I don't think we should solve that in Heat tbh | 14:24 |
| ayoung | Heat rotates the keys internally. Modifis the Hieradata. The OS-*config operations pick up the new values and update the key respons in /etc/keystone | 14:24 |
| EmilienM | first of all, it's a Keystone problem | 14:25 |
| ayoung | EmilienM, its kindof the only part of this that actually uses heat as designed | 14:25 |
| ayoung | it is an orchestration problem | 14:25 |
| EmilienM | Keystone itself should be working out of the box for operators | 14:25 |
| EmilienM | and it's not | 14:25 |
| EmilienM | it forces users to have orchestration to deal with fernet keys sync | 14:25 |
| ayoung | EmilienM, Keystone Fernet keys are only an issue if you need more than one Keystone server. That is what we are trying to solve here | 14:25 |
| EmilienM | (when doing multinode) | 14:26 |
| ayoung | Keystone itself does not know anything about other servers | 14:26 |
| EmilienM | ayoung: everyone in prod needs more than one keystone server | 14:26 |
| ayoung | and thus Keystone has no way to communicate the Keys with each other | 14:26 |
| EmilienM | like I said, it's a bad UX for operators here | 14:26 |
| EmilienM | ayoung: well, it should | 14:26 |
| jaosorior | EmilienM: why would the undercloud become ephemeral? | 14:27 |
| shardy | So, heat is a poor fit for scheduling periodic tasks, but we could use heat SoftwareDeployments to do this via mistral (which provides a "cron as a service" interface IIRC) | 14:27 |
| ayoung | EmilienM, we need to work within the defines of the OOO architecture, which is explicitly for large scale deployments. | 14:27 |
| ayoung | So, from a puppet perspective, the new key should be a new hiera value | 14:27 |
| EmilienM | ayoung: we'll fix the problem in tripleo but not for other operators | 14:27 |
| EmilienM | which is sad | 14:27 |
| shardy | http://docs.openstack.org/developer/heat/template_guide/openstack.html#OS::Mistral::CronTrigger | 14:27 |
| shardy | the task could be defined via the heat template tho, if you wish | 14:27 |
| *** prateek has quit IRC | 14:27 | |
| ayoung | EmilienM, puppet provider do not stand alone. They always need to get their data from somewhere | 14:28 |
| jaosorior | shardy: I would be fine with doing it via mistral, but you're talking about mistral int he undercloud, right? | 14:28 |
| EmilienM | shardy: yes, but iiuc during the Summit we said undercloud could become ephemeral in the future | 14:28 |
| EmilienM | shardy: with heat-undercloud blueprint | 14:28 |
| shardy | jaosorior: yes, so we'd have the single point of failure problem | 14:28 |
| jaosorior | ayoung: how often does rotation happen? | 14:28 |
| jaosorior | *should happen | 14:28 |
| shardy | EmilienM: No, that's the undercloud installer that's ephemeral, not the resulting undercloud | 14:28 |
| EmilienM | shardy: good | 14:29 |
| ayoung | EmilienM, so, I am not willing to make a complex solution based on that. Key rotation is a very seldom-run operation. If the undercloud goes away, you've lost MGMT of the overcloud til it comes back | 14:29 |
| ayoung | totally fits with key rotation | 14:29 |
| shardy | that said, we do avoid having it as a single point of failure wherever possible | 14:29 |
| ayoung | shardy, key rotation is run something like once a month | 14:29 |
| EmilienM | ayoung: do you manually scp the Swift ring when building a Swift cluster? | 14:29 |
| ayoung | EmilienM, I take the hobbit approach and toss rings into Volcanos | 14:29 |
| EmilienM | ayoung: Swift has this mechanism to balance and sync objects | 14:30 |
| shardy | ayoung: cool, so that should be fine, provided there's a means to catch up if the task trigger is missed, and they're rotated well before they expire | 14:30 |
| ayoung | I know nothing from Swift | 14:30 |
| mwhahaha | gfidente: it should probably be it's own class in like libvirt/qemu or something but the problem is that others might have defined a qemu.conf somewhere as well. | 14:30 |
| mwhahaha | gfidente: I recently went through this in the contrail plugin :/ | 14:30 |
| EmilienM | ayoung: telling to operators "do it yourself to balance fernet keys" is a terrible UX | 14:30 |
| jaosorior | shardy, EmilienM: I honestly think this task would be quite fine driven by mistral. | 14:30 |
| jaosorior | as shardy pointed out | 14:30 |
| EmilienM | because it works in devstack in Keystone gate doesn't mean it's good | 14:30 |
| ayoung | EmilienM, in non OOO systems, where does the puppet module gets it hiera data? | 14:30 |
| ayoung | a change to the hiera data should trigger a redeploy, and the keys are hiera data | 14:31 |
| EmilienM | ayoung: dunno, ask mfisch | 14:31 |
| *** gchamoul has joined #tripleo | 14:31 | |
| mwhahaha | ayoung: what do you mean? | 14:31 |
| *** jmiu has joined #tripleo | 14:31 | |
| EmilienM | jaosorior: we're just moving the problem imho | 14:31 |
| ayoung | mwhahaha, in a OOO system, puppet consumes hiera data published from Heat | 14:31 |
| mwhahaha | ayoung: usually the operator constructs their hieradata in their composition layer | 14:31 |
| mwhahaha | ayoung: so it could be via puppetmaster or by hand | 14:31 |
| *** marios_ has joined #tripleo | 14:32 | |
| ayoung | mwhahaha, so the rotation of fernet keys fall into that same problem set | 14:32 |
| EmilienM | my point is that keystone itself should be able to generate fernet keys and balance them itself | 14:32 |
| ayoung | EmilienM, that is actually a security hole | 14:32 |
| EmilienM | so we're moving this security hole to mistral or heat? | 14:32 |
| shardy | ayoung: so it's a feature that you can't do the key generation on the Keystone node (e.g via an actual cron job or something)? | 14:32 |
| *** tesseract has joined #tripleo | 14:32 | |
| shardy | I can understand that, just wanted to clarify | 14:32 |
| ayoung | Keystone consumes the keys, but the server itself does not generate them, even in the devstack case. You run a utility on the server itself to generate the new key, but there is nothing about transport | 14:33 |
| *** marios_ has quit IRC | 14:33 | |
| *** tesseract is now known as Guest48653 | 14:33 | |
| EmilienM | shardy: no you can't, and I think it's a design problem in Keystone | 14:33 |
| ayoung | shardy, it is a deliberate engineering decision | 14:33 |
| ayoung | EmilienM, actually, you *can* do it, but it does not provide for multi node support | 14:33 |
| EmilienM | ayoung: devstack gate doesn't test multiple keystone servers | 14:33 |
| shardy | ayoung: Ok, so the implication is all deployments must now have their own task scheduler and automation to deploy the keys, got it | 14:34 |
| ayoung | and it does not make sense for a OOO deploy because you would have to query the keys from one node and get them to the other nodes | 14:34 |
| *** jprovazn has quit IRC | 14:34 | |
| d0ugal | jaosorior: what are you considering doing with Mistral? :) | 14:34 |
| ayoung | shardy, yeah, and I think if you asked ggillies he would request that it be done via the undercloud. | 14:34 |
| jaosorior | d0ugal: fernet key rotation | 14:34 |
| ayoung | there is nothing magical about generating a new key. Its done on the undercloud now in THT | 14:35 |
| EmilienM | ayoung: can you tell me why you can't do it in Keystone itself? and you want installers to deal with that? | 14:35 |
| ayoung | EmilienM, I just said. Twice | 14:35 |
| ayoung | It is a TRANSPORT problem | 14:35 |
| ayoung | not a keystone problem | 14:35 |
| ayoung | Keystone servers do not know about each other | 14:35 |
| ayoung | the keys do not go in the database (we did discuss it) for security concerns | 14:36 |
| EmilienM | ayoung: couldn't keystone use swift ? | 14:36 |
| EmilienM | it's fast and secure | 14:36 |
| ayoung | They are written to disk and made as read only data for Keystone by a processs not running as the keystone user | 14:36 |
| EmilienM | (as an optional way to make rotations of course) | 14:36 |
| *** achadha has quit IRC | 14:37 | |
| ayoung | EmilienM, it *could* be done many ways. What we have here, in OOO is a way based on the Puppet modules, and keeping a consistent design approach. I can write the rotation as mistral, Ansible, shell, Assembly (well that would take some time) | 14:37 |
| *** kjw3 has joined #tripleo | 14:38 | |
| ayoung | but the current approach is to put the keys in hiera via THT, and have Puppet get them via the update process | 14:38 |
| EmilienM | ayoung: this approach could be changed, it's not something static | 14:38 |
| ayoung | Keystone does not know about Swift | 14:38 |
| EmilienM | well, it could know | 14:38 |
| ayoung | It would still be wrong | 14:39 |
| EmilienM | my point is that it sounds like it could be a feature in Keystone | 14:39 |
| ayoung | you are asking for Keystone to do things that the operators would explicitly find troublesome | 14:39 |
| EmilienM | I'm very surprised to see Keystone users happy with current Fernet implementation | 14:39 |
| *** Guest48653 has quit IRC | 14:39 | |
| *** tesseract- has joined #tripleo | 14:39 | |
| EmilienM | maybe can we take this on #openstack-keystone | 14:39 |
| ayoung | it absotutley could be, and it probably should not be | 14:39 |
| ayoung | EmilienM, no | 14:39 |
| ayoung | it is the wrong solution, for all the same reasons we use puppet instead | 14:40 |
| ayoung | We can make use of custodia, as jaosorior suggested, but that still really just says how to get the keys from the OS to the Keystone server | 14:40 |
| mwhahaha | are we trying to figure out how to do key rotation for the fernet keys? shouldn't it just be a tripleoclient action that calls a mistral template? | 14:40 |
| EmilienM | mwhahaha: it could be done like this but I'm still not convainced that it's a problem to be solved in TripleO itself | 14:41 |
| ayoung | we still need a way to sync the keys between keystone servers. If Keystone was going to manage that, we would do it via the database, but that have been discussed and ruled out | 14:41 |
| ayoung | this is one reason I really wanted it to be PKI. As Key sync is much easier there. You only need to distribute public keys | 14:41 |
| *** achadha has joined #tripleo | 14:41 | |
| mwhahaha | it should be solved in tripleo because that's the thing doing the deploying and management of the keystone configuration | 14:41 |
| mwhahaha | based on the way it's built today. | 14:42 |
| ayoung | mwhahaha, right. I didn't design OOO, I just live with the rules it sets | 14:42 |
| * mwhahaha throws rules at ayoung | 14:42 | |
| EmilienM | mwhahaha: based on the way it's done today, yes, but it doesn't mean it's the right way | 14:42 |
| EmilienM | we did like this in TripleO because Keystone doesn't provide a tool to make it out of the box | 14:43 |
| mwhahaha | people have been dealing with this since mitaka | 14:43 |
| EmilienM | but I seem to be alone in my thoughts :-) so... let's use mistral/heat whatever | 14:43 |
| mwhahaha | this is a larger issue with how do you do after deployment updates to an openstack environment | 14:43 |
| mwhahaha | there are issues with all of the services, but since tripleo is currently the thing that manages it for tripleo environments, it belongs in tripleo as a day2 op | 14:44 |
| mwhahaha | sure we can raise the issue with keystone, but then you get into keystone having to self cluster and it gets way complicated really fast | 14:44 |
| mwhahaha | for something a simple as a file update & process restart | 14:45 |
| EmilienM | afik it doesn't need restart | 14:45 |
| mwhahaha | well even easier then | 14:45 |
| mwhahaha | yea there probably should be an endpoint that accepts a new key and rotates it | 14:45 |
| mwhahaha | but the question is how do you distributed it across the X number of keystone servers | 14:46 |
| *** derekh_afk is now known as derekh | 14:47 | |
| *** tesseract- has quit IRC | 14:49 | |
| openstackgerrit | Dan Radez proposed openstack/tripleo-heat-templates: Adding Congress Support https://review.openstack.org/330050 | 14:50 |
| *** flepied1 has joined #tripleo | 14:50 | |
| *** jpena|lunch is now known as jpena | 14:51 | |
| *** jaosorior has quit IRC | 14:53 | |
| *** flepied has quit IRC | 14:53 | |
| *** jaosorior has joined #tripleo | 14:54 | |
| *** achadha has quit IRC | 14:54 | |
| *** kashyap has joined #tripleo | 14:55 | |
| *** jkilpatr_ has joined #tripleo | 14:55 | |
| kashyap | Hi folks, a random question -- how much is the size of an Undercloud installer disk image? | 14:56 |
| openstackgerrit | Merged openstack/instack-undercloud: Open firewall port for the TripleO UI https://review.openstack.org/396706 | 14:56 |
| openstackgerrit | Merged openstack/puppet-tripleo: Proxy manila in http mode https://review.openstack.org/400648 | 14:56 |
| openstackgerrit | Merged openstack/tripleo-puppet-elements: Element overcloud-agent replaces os-collect-config https://review.openstack.org/386303 | 14:56 |
| openstackgerrit | Merged openstack/tripleo-image-elements: Fix tr syntax in 51-hosts https://review.openstack.org/385560 | 14:56 |
| *** jkilpatr has quit IRC | 14:56 | |
| *** Goneri has joined #tripleo | 14:56 | |
| shardy | kashyap: The image consumed by tripleo-quickstart is 2.8G, but you can also run the undercloud installer on any RHEL/CentOS box - you don't have to start with an image | 14:58 |
| shardy | http://buildlogs.centos.org/centos/7/cloud/x86_64/tripleo_images/newton/delorean/ | 14:58 |
| * kashyap clicks | 14:58 | |
| openstackgerrit | Carlos Camacho proposed openstack-infra/tripleo-ci: Add Panko to scenario001 https://review.openstack.org/401299 | 14:59 |
| kashyap | shardy: I see, noted. I didn't explain the purpose of me asking for it. I was thinking whether this is possible: | 14:59 |
| *** dmarlin is now known as dmarlin_ | 14:59 | |
| *** achadha has joined #tripleo | 14:59 | |
| kashyap | shardy: "To provide an absolute minimal OpenStack setup in an xz-compressed disk image of no more than 300MB (that's the key)." | 15:00 |
| panda | kashyap: shardy when testing tripleo-ci created images, we just donwload the overcloud image, and convert it to the undercloud image, so we just have to download 1.2G | 15:00 |
| derekh | rook: a couple of other things worth mentioning (if your going for a larger then normal deploy on ovb) | 15:01 |
| kashyap | (NB: It's not really for real consumption; that disk image is more targetted for those who want to just download an image and see what it looks like. So to condense, could do hacks like: remove package manager metadata, log files, man pages, locales) | 15:01 |
| kashyap | panda: Noted. Thanks. | 15:01 |
| *** abregman_ has joined #tripleo | 15:01 | |
| shardy | kashyap: the undercloud isn't really a very minimal openstack setup and it's not particularly easy to customize it to remove pieces you don't need | 15:01 |
| *** cwolferh has joined #tripleo | 15:02 | |
| shardy | but I'm sure you could reduce the size I quoted quite a bit | 15:02 |
| rook | derekh: yeah, going with 60 now, 100 later | 15:02 |
| derekh | rook: double check that your bmc node didn't run out of RAM, it spins up a python process for each barmetal node you have | 15:02 |
| rook | yeah, i gave it 2048 | 15:02 |
| rook | so, hoping that should suffice | 15:02 |
| derekh | rook: ok | 15:02 |
| kashyap | shardy: Understood. (In the past, DEC-2014 to be precise, I was able to reduce the size to 618M xz-compressed.) | 15:03 |
| ccamacho | hey shardy if you have some time, we can sync about https://blueprints.launchpad.net/tripleo/+spec/overcloud-upgrades-per-service | 15:03 |
| kashyap | Thanks for all the pointers, folks. | 15:03 |
| derekh | rook: also your undercloud and overcloud controllers will need an appropriate amount of RAM | 15:03 |
| derekh | rook: the computes you | 15:03 |
| *** cdearborn has joined #tripleo | 15:04 | |
| rook | yeah, derekh i have the default 6G, my hope is to lower that to like 4GB | 15:04 |
| rook | so I can get higher density | 15:04 |
| *** achadha has quit IRC | 15:04 | |
| derekh | rook: when the deploy I did last week, I have a OVB env with 3 baremetal nodes with 16G and 50 nodes with 2G and used capabilites to the undercloud selected the big nodes for the controller | 15:04 |
| derekh | *so the undercloud | 15:05 |
| *** abregman|mtg has quit IRC | 15:05 | |
| *** oshvartz has quit IRC | 15:05 | |
| derekh | rook: bit the OVB templates don't support doing that out of the box, its was a bit of a manual tweaking involved | 15:05 |
| *** panda is now known as panda|brb | 15:05 | |
| derekh | bnemec: you were talking about supporting that at some stage were you? | 15:06 |
| derekh | bnemec: ^ | 15:06 |
| *** [3]cdearborn has quit IRC | 15:08 | |
| *** jaosorior has quit IRC | 15:09 | |
| beagles | bandini: what are your thoughts on the os-net-config on upgrade thing? should we error out if os-net-config fails? | 15:09 |
| *** jaosorior has joined #tripleo | 15:10 | |
| *** abregman_ has quit IRC | 15:11 | |
| *** abregman has joined #tripleo | 15:11 | |
| jaosorior | ayoung: hey dude, any idea why would the ipa-server installation fail with this error? http://logs.openstack.org/38/391738/33/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-fakeha-caserver/bf5de17/console.html#_2016-11-23_13_14_47_530845 | 15:12 |
| *** achadha has joined #tripleo | 15:13 | |
| EmilienM | panda|brb: I found a new blocker in nova/neutron | 15:14 |
| EmilienM | panda|brb: http://logs.openstack.org/38/401038/3/check/gate-puppet-openstack-integration-4-scenario004-tempest-centos-7/9948dfc/logs/nova/nova-compute.txt.gz#_2016-11-23_14_04_46_408 | 15:14 |
| EmilienM | we're hitting it in puppet CI but I'm sure we'll have it in tripleo CI too | 15:14 |
| EmilienM | panda|brb: on periodic jobs at least, where tempest is run | 15:15 |
| openstackgerrit | Merged openstack/tripleo-common: Install yum-plugin-priorities on centos images https://review.openstack.org/397330 | 15:16 |
| openstackgerrit | Merged openstack/tripleo-common: Drop *openstack/common* in flake8 exclude list https://review.openstack.org/373882 | 15:16 |
| openstackgerrit | Merged openstack/tripleo-common: Add port registration to IPMI power driver https://review.openstack.org/365953 | 15:16 |
| openstackgerrit | Merged openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401212 | 15:16 |
| *** limao has joined #tripleo | 15:16 | |
| openstackgerrit | Merged openstack/python-tripleoclient: Remove network-gateway element from image build https://review.openstack.org/398584 | 15:16 |
| *** gchamoul is now known as gchamoul|afk | 15:16 | |
| openstackgerrit | Dougal Matthews proposed openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401267 | 15:17 |
| bandini | beagles: added a comment | 15:17 |
| *** gchamoul|afk is now known as gchamoul | 15:17 | |
| bandini | beagles: I guess I can be convinced either way. bailing out with an error code and exit is probably a bit better | 15:17 |
| *** panda|brb is now known as panda | 15:17 | |
| panda | EmilienM: any clue what's the interface it's trying to create ? | 15:19 |
| EmilienM | panda: we're debugging on #puppet-openstack | 15:20 |
| openstackgerrit | Merged openstack/python-tripleoclient: Add unit tests for bracket_ipv6 https://review.openstack.org/396345 | 15:21 |
| openstackgerrit | Merged openstack/python-tripleoclient: Use environment fixture in tests https://review.openstack.org/394794 | 15:21 |
| openstackgerrit | Merged openstack/python-tripleoclient: Give better output on scale failures https://review.openstack.org/398226 | 15:21 |
| openstackgerrit | Merged openstack/python-tripleoclient: Refactor creation of no_proxy list for postconfig https://review.openstack.org/391197 | 15:21 |
| *** dtrainor has quit IRC | 15:22 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: WIP - Deploy TripleO with Puppet 4 https://review.openstack.org/371209 | 15:22 |
| *** achadha has quit IRC | 15:22 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: WIP - Deploy TripleO with Puppet 4 https://review.openstack.org/371209 | 15:23 |
| *** kashyap has left #tripleo | 15:24 | |
| *** tosky has quit IRC | 15:24 | |
| *** limao_ has joined #tripleo | 15:25 | |
| *** chlong has joined #tripleo | 15:25 | |
| *** tosky has joined #tripleo | 15:25 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Fix resource_registry path in enable-internal-tls https://review.openstack.org/400723 | 15:27 |
| *** achadha has joined #tripleo | 15:27 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Configure Keystone Fernet Keys https://review.openstack.org/397350 | 15:27 |
| *** limao has quit IRC | 15:28 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/tripleo-common: Do not ignore pm_port for Drac drivers https://review.openstack.org/400276 | 15:28 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/tripleo-common: Add port registration to IPMI power driver https://review.openstack.org/401316 | 15:29 |
| lucasagomes | derekh, ^ backported your patch cause we will need it in newton | 15:29 |
| *** achadha has quit IRC | 15:32 | |
| *** udesale has quit IRC | 15:32 | |
| *** chlong has quit IRC | 15:34 | |
| *** numans has quit IRC | 15:34 | |
| derekh | lucasagomes: ok, lgtm | 15:34 |
| lucasagomes | ty | 15:35 |
| *** achadha has joined #tripleo | 15:36 | |
| *** jlinkes has quit IRC | 15:38 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: Explicitly set rabbit hosts so its not overridden during upgrade https://review.openstack.org/401322 | 15:39 |
| openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Apply os-net-config with a script instead of element https://review.openstack.org/271450 | 15:39 |
| openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Cleanup some inline comments in network/config https://review.openstack.org/401324 | 15:39 |
| ansiwen | I wanna use the "condition" parameter for a resource, can I just bump the template version to Newton? | 15:39 |
| *** jprovazn has joined #tripleo | 15:39 | |
| ansiwen | or is there an alternative for conditional resources? | 15:40 |
| *** limao_ has quit IRC | 15:40 | |
| dprince | shardy: https://review.openstack.org/#/c/271450/ | 15:40 |
| *** limao has joined #tripleo | 15:41 | |
| dprince | shardy: so I wrote some rather horrible code to maintain the comments. I didn't use the new library... found it easier to just hack for now | 15:41 |
| dprince | shardy: it isn't perfect either. So if we'd have to live with this cleanup too https://review.openstack.org/#/c/401324/ | 15:42 |
| dprince | shardy: but, if we do that the rest of the (more meaningful) comments are now preserved | 15:42 |
| *** dmarlin has joined #tripleo | 15:42 | |
| dprince | shardy: I've kept the tools/yaml-nic-config-2-script.py in tree for now, but I wonder if we should just remove it | 15:44 |
| fultonj | off topic question: is anyone using tripleo-quickstart successfully on fedora? | 15:44 |
| *** achadha has quit IRC | 15:45 | |
| *** prateek_ has quit IRC | 15:46 | |
| *** prateek_ has joined #tripleo | 15:46 | |
| *** panda is now known as panda|brb | 15:48 | |
| *** chlong has joined #tripleo | 15:49 | |
| shardy | ansiwen: yes, you can just bump the version to newton or ocata HOT | 15:49 |
| ansiwen | shardy: ok, thanks | 15:49 |
| shardy | fultonj: it's not tested much at all, IIRC flaper87 was trying to get it working | 15:49 |
| *** gchamoul is now known as gchamoul|afk | 15:50 | |
| shardy | dprince: nice, thanks for the update | 15:50 |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: Explicitly set rabbit hosts so its not overridden during upgrade https://review.openstack.org/401322 | 15:50 |
| *** gchamoul|afk is now known as gchamoul | 15:50 | |
| openstackgerrit | Sven Anderson proposed openstack/tripleo-heat-templates: Add option for deploying ssh keys to nova users on compute nodes https://review.openstack.org/400942 | 15:50 |
| *** achadha has joined #tripleo | 15:50 | |
| openstackgerrit | Merged openstack/puppet-tripleo: Remove Combination alarms support https://review.openstack.org/398579 | 15:51 |
| shardy | dprince: wow, see what you mean - somewhat hacky but whatever works :) | 15:52 |
| shardy | dprince: FWIW I thought we should put any conversion script in-tree, as we do expect folks to have out-of-tree templates that will need conversion | 15:52 |
| dprince | shardy: right, I suppose the issue is with comments | 15:52 |
| dprince | shardy: the current approach doesn't handle them perfectly | 15:53 |
| shardy | dprince: did you check out ruamel? | 15:53 |
| shardy | I started looking at it, then got sidetracked by other tasks | 15:53 |
| dprince | shardy: yes, but it isn't a drop in | 15:53 |
| shardy | yeah, that's where I stopped, it's got a slightly different API from pyyaml | 15:53 |
| dprince | shardy: like we would need to re-write a good bit of the other stuff to use it I think | 15:53 |
| flaper87 | shardy: what did I do ? | 15:53 |
| dprince | shardy: so kind of a trade off | 15:54 |
| flaper87 | oh, qs on f24 ? | 15:54 |
| flaper87 | I manged to get it working | 15:54 |
| flaper87 | fultonj: ^ | 15:54 |
| fultonj | flaper87: cool | 15:54 |
| shardy | flaper87: yeah, fultonj was asking about qs w/fedora | 15:54 |
| dprince | shardy: also, there aren't any packages for ruamel or anything. But we could fix that | 15:54 |
| fultonj | flaper87: thanks, i guess i'll try a newish distro | 15:54 |
| flaper87 | fultonj: lemme remember what I did | 15:54 |
| fultonj | flaper87: i was thinking of giving up shiny new browsers to be able to work on a plane. | 15:54 |
| shardy | dprince: ack, I'm fine with this as a first pass - we can always improve the conversion script before the final release if a better approach can be found | 15:55 |
| flaper87 | fultonj: :D :D :D | 15:55 |
| flaper87 | fultonj: I think the only change I had to make landed already | 15:55 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::agent class https://review.openstack.org/400618 | 15:55 |
| flaper87 | fultonj: I did fight a little bit with it | 15:55 |
| flaper87 | fultonj: lemme know if it fails for you | 15:55 |
| flaper87 | and I'll try to help | 15:55 |
| shardy | dprince: we should be able to make it so that out of tree stuff still works despite the conversion in-tree, but it'd be nice to eventually remove all the oac support | 15:56 |
| fultonj | flaper87: cool, thanks. if i get it working and needed to use tricks then i'll write it up if it helps anyone | 15:56 |
| flaper87 | fultonj: I failed at that, tbh :( | 15:56 |
| flaper87 | fultonj: I know I symlinked some dirs for the virt images | 15:56 |
| fultonj | flaper87: you don't have an ansible script to install the ansible scripts? :) | 15:56 |
| * flaper87 pulls out the ansible script to install the ansible script that installs the ansible script | 15:57 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Remove Combination alarms support https://review.openstack.org/398580 | 15:57 |
| fultonj | flaper87: thanks, i'll see how it goes this weekend | 15:57 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Add new tripleo::network::midonet::nsdb class https://review.openstack.org/400869 | 15:57 |
| *** abregman is now known as abregman|mtg | 15:58 | |
| *** panda|brb is now known as panda | 15:58 | |
| openstackgerrit | Merged openstack/tripleo-docs: Make it clear that removing ironic-inspector.sqlite is for Liberty only https://review.openstack.org/377917 | 15:59 |
| *** alee has joined #tripleo | 15:59 | |
| alee | jaosorior, whats up? | 15:59 |
| *** achadha has quit IRC | 16:00 | |
| jaosorior | alee: hey dude, I'm trying to get freeipa for the overcloud, for the TLS everywhere job | 16:00 |
| jaosorior | alee: but I'm having some issues with the FreeIPA install | 16:00 |
| jaosorior | I'm seeing this http://logs.openstack.org/38/391738/33/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-fakeha-caserver/bf5de17/console.html#_2016-11-23_13_14_47_530845 | 16:00 |
| alee | jaosorior, ok - which version of freeipa? what platform? | 16:01 |
| alee | ah centos .. | 16:01 |
| jaosorior | alee: CentOS, yeah | 16:01 |
| jaosorior | alee: and this is the heat template that does the work https://review.openstack.org/#/c/391738/33/heat-templates/freeipa_server.yaml | 16:01 |
| *** rcrit has joined #tripleo | 16:02 | |
| alee | jaosorior, rcrit might know about this .. | 16:02 |
| alee | jaosorior, I think I saw something recently about this .. | 16:02 |
| alee | rcrit, http://logs.openstack.org/38/391738/33/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-fakeha-caserver/bf5de17/console.html#_2016-11-23_13_14_47_530845 | 16:03 |
| jaosorior | rcrit, right, so to bring context, I'm seeing this on a FreeIPA install http://logs.openstack.org/38/391738/33/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-fakeha-caserver/bf5de17/console.html#_2016-11-23_13_14_47_530845 | 16:03 |
| alee | rcrit, failed ipa install on centos | 16:03 |
| jaosorior | rcrit: trying to work on a TripleO job that includes FreeIPA so we can test TLS everywhere | 16:03 |
| alee | rcrit, I thought I saw something about this recently | 16:03 |
| rcrit | uh | 16:03 |
| alee | is this file /etc/httpd/conf.d/nss.conf provided by mod_nss? or created by the installer? | 16:04 |
| *** achadha has joined #tripleo | 16:04 | |
| rcrit | provided by the package | 16:04 |
| alee | rcrit, which package? | 16:05 |
| rcrit | mod_nss | 16:05 |
| alee | jaosorior, do you have mod_nss installed? | 16:05 |
| rcrit | it defaults to port 8443 to co-exist with mod_ssl, ipa switches the port to 443 | 16:05 |
| rcrit | I can see the pkg being installed in the logs | 16:05 |
| *** egafford has quit IRC | 16:05 | |
| rcrit | it's a dependency of freeipa | 16:05 |
| jaosorior | rcrit, alee: and this is the heat template that does the work https://review.openstack.org/#/c/391738/33/heat-templates/freeipa_server.yaml yes, I installed mod_nss | 16:05 |
| rcrit | gah, Adam's password will never go away | 16:06 |
| rcrit | jaosorior, got it, you're removing it, like 98/99 | 16:07 |
| jaosorior | sorry for the unreadability, been trying to make it work for 33 iterations now | 16:07 |
| *** egafford has joined #tripleo | 16:07 | |
| rcrit | er, line | 16:07 |
| jaosorior | rcrit: that makes sense | 16:07 |
| *** ealcaniz has quit IRC | 16:08 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP TLS everywhere job https://review.openstack.org/391738 | 16:09 |
| jaosorior | I wonder why that was there | 16:09 |
| jaosorior | oh well, I'll try it out | 16:09 |
| dprince | shardy: I added https://github.com/openstack/tripleo-image-elements/commit/cf94c5ebda7050569163f2536a46089280a45c9b before committing the patch I think | 16:09 |
| jaosorior | rcrit, alee thanks dude; sorry for the spam | 16:10 |
| jaosorior | *dudes | 16:10 |
| rcrit | yw | 16:10 |
| *** bana_k has joined #tripleo | 16:10 | |
| *** tremble has quit IRC | 16:10 | |
| shardy | dprince: Ah, yeah, just the commit message is still referencing the old commit | 16:11 |
| dprince | shardy: oh, right. I missed that part. DIdn't expect it to be in the commit message too I guess | 16:12 |
| shardy | dprince: Yeah sorry I looked at the commit message without fully diffing the code | 16:12 |
| bnemec | derekh: Yeah, that's literally the next thing on my todo list. :-) | 16:14 |
| *** cylopez has quit IRC | 16:14 | |
| jaosorior | alee: by the way, did you get time to check out the nameserver deal for the undercloud? | 16:15 |
| alee | jaosorior, I'm trying it out right now. the patch I posted has a typo and needs a couple of changes | 16:18 |
| alee | jaosorior, but right now it does not appear to be working | 16:18 |
| openstackgerrit | Sven Anderson proposed openstack/tripleo-heat-templates: Add ec2-api service https://review.openstack.org/398634 | 16:18 |
| jaosorior | alee: can you pass the link? | 16:18 |
| *** achadha has quit IRC | 16:18 | |
| shardy | jtomasek, d0ugal: Hey, can you provide any insight into what user-environment.yaml is used for in the UI? | 16:18 |
| alee | jaosorior, https://review.openstack.org/#/c/400937/ | 16:18 |
| d0ugal | shardy: I don't think it is used, yet | 16:18 |
| shardy | d0ugal: OK so we can potentially remove it then, phew :) | 16:19 |
| alee | jaosorior, I was hoping that just setting the option would all make it magically work - but I'm not seeing that | 16:19 |
| d0ugal | shardy: Yeah, I'd love to remove it. | 16:19 |
| d0ugal | shardy: https://github.com/openstack/tripleo-common/blob/6f3095bf1b7c8c286e8c4ddcb2c6c761a2700fd2/tripleo_common/actions/heat_capabilities.py#L94-L95 | 16:19 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Add option to set undercloud dns nameserver https://review.openstack.org/400937 | 16:19 |
| alee | jaosorior, it might work after the ipa client install, but I need it to happen before then | 16:19 |
| jaosorior | alee: rebased it, lets see | 16:20 |
| jtomasek | d0ugal, shardy: I don't remember the intention of its usage | 16:20 |
| d0ugal | shardy: that is the only place it is used in tripleo common IIRC, and the UI uses the response | 16:20 |
| alee | jaosorior, no -- there is a typo in it | 16:20 |
| shardy | d0ugal: OK so my approach is to instead add the user environments that aren't from tht/environments to user-files/blah | 16:20 |
| d0ugal | jtomasek: I don't think there is any intention for the GUI to ever use it | 16:20 |
| shardy | then add user-files/blah to the environments list in the mistral environment | 16:20 |
| jaosorior | alee: where? | 16:20 |
| d0ugal | shardy: cool, sounds good. | 16:20 |
| openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Apply os-net-config with a script instead of element https://review.openstack.org/271450 | 16:20 |
| jaosorior | ah | 16:21 |
| alee | jaosorior, cfgStrOpt instead of cfg.StrOpt | 16:21 |
| jaosorior | I see | 16:21 |
| jaosorior | yeah | 16:21 |
| d0ugal | shardy: I think that is what I wanted to do originally but failed | 16:21 |
| alee | jaosorior, and the validator test is invalid - should be removed | 16:21 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Add option to set undercloud dns nameserver https://review.openstack.org/400937 | 16:21 |
| alee | jaosorior, but even with those changes not working -- so trying to figure it out | 16:21 |
| shardy | d0ugal: Ok, well I think I'm close now, hopefully you can sanity check the result | 16:21 |
| jaosorior | alee: I can test it out in my local deployment as well | 16:22 |
| alee | jaosorior, cool | 16:22 |
| alee | jaosorior, I made a couple of changes to quickstart to get things going - but you've got the idea | 16:23 |
| *** achadha has joined #tripleo | 16:23 | |
| openstackgerrit | Merged openstack/instack-undercloud: Set heat.conf max_nested_stack_depth = 6 https://review.openstack.org/397508 | 16:23 |
| jaosorior | alee: I have the idea of what needs to happen in the end term. But I'm not entirely sure what's not working from that patch yet | 16:23 |
| d0ugal | shardy: great, looking forward to it | 16:23 |
| alee | jaosorior, basically you want undercloud_nameserver = foo in your undercloud.conf | 16:24 |
| d0ugal | shardy: I can't even remember what blocked me - it might just have been lack of time. | 16:24 |
| jaosorior | right | 16:24 |
| alee | jaosorior, well - there was something in instack-undercloud to support this, but I'm not sure it works | 16:24 |
| alee | jaosorior, in fact, I have a deep suspicion it doesn;t | 16:25 |
| *** leanderthal is now known as leanderthal|afk | 16:25 | |
| *** kjw3 has quit IRC | 16:25 | |
| *** bana_k has quit IRC | 16:25 | |
| *** cdearborn has quit IRC | 16:26 | |
| d0ugal | jpich: Can you +2 this now? https://review.openstack.org/#/c/375348/ | 16:26 |
| jaosorior | alee: it might actually be that it doesn't... it hasn't been use in a LONG time | 16:26 |
| alee | yup | 16:26 |
| openstackgerrit | Harry Rybacki proposed openstack/tripleo-quickstart-extras: Update overcloud-scale post extra roles merging https://review.openstack.org/400987 | 16:27 |
| *** achadha has quit IRC | 16:28 | |
| derekh | bnemec: cool, when I was setting up a env last week I had to create 2 separate envs and then ninja merge them together | 16:29 |
| openstackgerrit | Sven Anderson proposed openstack/puppet-tripleo: Implement Nova ec2api profile https://review.openstack.org/336562 | 16:29 |
| jpich | d0ugal: Sept 23 - +1, sure, what could possibly go wrong ;) | 16:29 |
| jaosorior | derekh, bnemec any idea where the setup-neutron command comes from? | 16:29 |
| d0ugal | jpich: yay | 16:29 |
| bnemec | derekh: Yeah, it can be done by hand, but it's a little clunky and requires some knowledge of the implementation details of stuff like build-nodes-json. | 16:29 |
| *** dkehn__ has joined #tripleo | 16:29 | |
| bnemec | jaosorior: I believe os-cloud-config. | 16:30 |
| derekh | bnemec: yup it was a bit tricky to get working | 16:30 |
| sshnaidm | EmilienM, should we recheck the oooq job or can merge it so? https://review.openstack.org/#/c/381094/ libery fail doesn't seem related anyhow (and anyway we disable liberty jobs) | 16:31 |
| EmilienM | sshnaidm: it's on my long list of things to review | 16:31 |
| EmilienM | sshnaidm: but indeed, liberty used to pass in previous patchset | 16:32 |
| EmilienM | so I guess it's fine | 16:32 |
| sshnaidm | EmilienM, cool, thanks | 16:32 |
| flaper87 | is there a way to extend qs ? I know it's a bunch of ansible roles, I guess what I'm asking is what the best way to do so is | 16:32 |
| * flaper87 should probably look into extras | 16:32 | |
| *** achadha has joined #tripleo | 16:32 | |
| jaosorior | alee: well, from what I can see, it seems that that variable's only purpose was to set the nameserver in neutron for a specific subnet | 16:32 |
| *** bogdando has quit IRC | 16:32 | |
| *** aufi has quit IRC | 16:33 | |
| alee | jaosorior, yeah - not exactly what we need | 16:33 |
| *** dprince has quit IRC | 16:33 | |
| openstackgerrit | Dimitri Savineau proposed openstack/tripleo-heat-templates: Remove duplicate neutron_mechanism_drivers https://review.openstack.org/314734 | 16:33 |
| EmilienM | flaper87: you might want to ask weshay or trown|outtypewww | 16:33 |
| jaosorior | alee: there are two ways to go now. Either changing adding functionality to that variable, or keep going with the quickstart approach | 16:33 |
| jpich | d0ugal, shardy: In case this might be helpful / can help avoid too much divergence between the UI and CLI, this is what a Mistral environment with a bunch of environments enabled from the UI looks like: http://paste.openstack.org/show/590233/ | 16:34 |
| jaosorior | flaper87: or ask in #oooq | 16:34 |
| openstackgerrit | Merged openstack/tripleo-docs: Add Newton repo details https://review.openstack.org/396736 | 16:34 |
| *** ramishra has quit IRC | 16:34 | |
| *** abehl has quit IRC | 16:34 | |
| alee | jaosorior, right | 16:34 |
| alee | jaosorior, any preference ? | 16:34 |
| weshay | flaper87, ya.. create a new role via the ansible-galaxy command | 16:34 |
| flaper87 | oh, there's oooq #feelingdumb | 16:34 |
| openstackgerrit | Alejandro Andreu proposed openstack/puppet-tripleo: Adds new tripleo::network::midonet::config class https://review.openstack.org/401342 | 16:34 |
| flaper87 | weshay: thought so, I'll do that :D | 16:35 |
| weshay | flaper87, then push it out to say your local github account or just keep it local | 16:35 |
| weshay | update the quickstart-extras-requirements w/ the uri | 16:35 |
| alee | jaosorior, I know the quickstart approach works -- but the question is whether it makes more sense to do this in more than just the quickstart case | 16:35 |
| weshay | flaper87, actually we have some doc on this | 16:35 |
| jaosorior | alee: lets ask | 16:35 |
| weshay | flaper87, and update the playbook | 16:35 |
| flaper87 | weshay: docs would work, thanks a bunch | 16:35 |
| alee | jaosorior, who? | 16:35 |
| weshay | flaper87, http://docs.openstack.org/developer/tripleo-quickstart/readme.html#working-with-quickstart-extras | 16:36 |
| weshay | http://docs.openstack.org/developer/tripleo-quickstart/working-with-extras.html#working-with-extras | 16:37 |
| flaper87 | weshay: thanks a bunch | 16:37 |
| EmilienM | sshnaidm: where can I see oooq logs? | 16:37 |
| jaosorior | bnemec, EmilienM: For FreeIPA integration, we want the undercloud to use FreeIPA as a nameserver. Now, there is a variable called UNDERCLOUD_NAMESERVER in instack-undercloud, but right now that can only set the neutron nameserver for the physical subnet (I guess ctlplane?); do you guys think it makes sense to add functionality to that variable to actually configure the undercloud's nameserver? I guess | 16:37 |
| jaosorior | modifying /etc/resolv.conf . Or should we add this kind of configuration elsewhere? | 16:37 |
| *** weshay is now known as weshay_lunch | 16:37 | |
| jaosorior | bnemec: EmilienM so, it would not be in instack-undercloud, but we could do this in quickstart | 16:37 |
| EmilienM | jaosorior: it should be in i-u I think | 16:38 |
| sshnaidm | EmilienM, in running patch: https://review.openstack.org/#/c/397609/ for example: http://logs.openstack.org/09/397609/10/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/b33cfc4/logs/undercloud/var/log/oooq/ | 16:38 |
| jaosorior | alee: ^^ | 16:38 |
| EmilienM | sshnaidm: thx | 16:38 |
| alee | EmilienM, jaosorior -- ok - I can add to instack-undercloud | 16:39 |
| EmilienM | sshnaidm: we need to verify we have feature testing parity | 16:39 |
| sshnaidm | EmilienM, how to verify it? | 16:40 |
| EmilienM | slagle: in your first review, have you looked at it already? | 16:40 |
| EmilienM | in https://review.openstack.org/#/c/381094/ I mean | 16:41 |
| *** achadha has quit IRC | 16:41 | |
| panda | EmilienM: what's your term of comparison ? | 16:41 |
| *** ccamacho has quit IRC | 16:41 | |
| *** yamahata has joined #tripleo | 16:42 | |
| EmilienM | panda: we need to make sure that switching to oooq doesn't change our coverage | 16:42 |
| *** cdearborn has joined #tripleo | 16:42 | |
| panda | EmilienM: this is part of the ongoing effort, there are some tasks to address exactly these problems | 16:43 |
| *** ramishra has joined #tripleo | 16:43 | |
| *** mcornea has quit IRC | 16:43 | |
| panda | EmilienM: we know we're missing some parts. For example idempotency, overcloud deletion, undercloud sanity check in undercloud install-only jobs | 16:43 |
| EmilienM | since we're running ovb first, I don't mind about undercloud things | 16:44 |
| EmilienM | idempotency and overcloud deletion is important though | 16:44 |
| panda | EmilienM: we're translating job by job all the things that are required. Another example is postci, all tripleo dev are looking for this on a failure, we have yet to add it to the log collection | 16:45 |
| *** achadha has joined #tripleo | 16:46 | |
| EmilienM | panda: good ! | 16:46 |
| sshnaidm | EmilienM, this could come in next revision, running this job on live patches will help to detect things that we didn't think about them | 16:47 |
| EmilienM | sshnaidm: ok | 16:47 |
| EmilienM | sshnaidm: I'm just making sure we're aware about that and our team agrees | 16:47 |
| sshnaidm | EmilienM, sure, it's not the finish at all :) | 16:48 |
| bnemec | Holy updates job pass, batman! https://review.openstack.org/#/c/386713 | 16:48 |
| sshnaidm | it's only start | 16:48 |
| alee | jaosorior, I'm trying to see what order the elements in i-u run .. how/where would I find that? | 16:49 |
| panda | EmilienM: I wasn't able to follow the nova error, is there a bug somewhere ? | 16:49 |
| EmilienM | sshnaidm, panda +2 | 16:50 |
| EmilienM | panda: not yet | 16:50 |
| jaosorior | alee: I'm not entirely sure to be honest | 16:50 |
| EmilienM | panda: beagles is investigating also | 16:50 |
| EmilienM | beagles: can you file a bug please? you know Neutron/Nova much better than I do | 16:50 |
| jaosorior | slagle, bnemec: How do we figure out the order in which the elements in instack-undercloud are run? | 16:50 |
| EmilienM | beagles: I can give you details if you miss some info but I think you have everything already | 16:50 |
| beagles | EmilienM: yup | 16:50 |
| EmilienM | beagles: I suggest to file it in neutron | 16:51 |
| beagles | EmilienM: k | 16:51 |
| sshnaidm | jaosorior, can you please re-review oooq patch in your time, I've addressed your issues: https://review.openstack.org/#/c/381094/ | 16:53 |
| jaosorior | sshnaidm: is it cool if I do that tomorrow? I'd like to give it another good read | 16:54 |
| openstackgerrit | Ronelle Landy proposed openstack/tripleo-quickstart: Modify OVB ci-script to tripleo-quickstart-extras https://review.openstack.org/398429 | 16:54 |
| sshnaidm | jaosorior, sure, thanks | 16:54 |
| slagle | jaosorior: the scripts are sorted numerically within each hook | 16:54 |
| slagle | a hook being pre-install/install/post-install | 16:54 |
| slagle | etc | 16:55 |
| jaosorior | alee: ^^ | 16:55 |
| jaosorior | slagle: thanks | 16:56 |
| alee | jaosorior, slagle - yeah - and I think I see which each hook is called in the install method of undercloud.py | 16:56 |
| alee | thanks | 16:56 |
| *** tzumainn has quit IRC | 16:56 | |
| jaosorior | alee: it seems that the hook where currently undercloud_nameserver is taken into account runs quite late in the deployment | 16:56 |
| alee | jaosorior, yeah - its not going to work there. | 16:57 |
| alee | jaosorior, I need to make changes before the puppet modules run | 16:57 |
| *** Guest83036 has joined #tripleo | 16:57 | |
| *** rhallisey_ has joined #tripleo | 16:57 | |
| *** sudipto_ has joined #tripleo | 16:57 | |
| openstackgerrit | Alex Schultz proposed openstack/instack-undercloud: Increase the default number of workers for heat engine https://review.openstack.org/399619 | 16:57 |
| slagle | alee: jaosorior : then the $UNDERCLOUD_NAMESERVER we were talking about yesterday must not do what you think it does | 16:58 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Fix roles listing for custom plan https://review.openstack.org/401351 | 16:58 |
| *** dprince has joined #tripleo | 16:58 | |
| slagle | alee: jaosorior : that was for seeing a nameserver in the undercloud's neutron, which means the value there gets used for overcloud nodes | 16:58 |
| *** ctayal has joined #tripleo | 16:58 | |
| slagle | *setting | 16:58 |
| alee | slagle, yeah - I want to actually set the nameserver for the undercloud node | 16:59 |
| jaosorior | slagle: indeed, we were noticing that a little bit ago. So alee was thinking about using that variable to set the nameserver for the undercloud, which is what we needed ni the first place | 16:59 |
| *** ccamacho has joined #tripleo | 16:59 | |
| alee | slagle, ie. setting it in /etc/resolv.conf | 16:59 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Fix roles listing for custom plan https://review.openstack.org/401351 | 16:59 |
| slagle | in that case, can that be configured via hieradata and including some puppet class in the manifest? | 17:00 |
| alee | slagle, jaosorior potentially .. | 17:01 |
| *** athomas has quit IRC | 17:04 | |
| *** achadha has quit IRC | 17:04 | |
| openstackgerrit | Ronelle Landy proposed openstack/tripleo-quickstart: Modify OVB ci-script to tripleo-quickstart-extras https://review.openstack.org/398429 | 17:04 |
| jaosorior | slagle: we could do that but I think it'll end up in us including another dependency just for this. | 17:04 |
| *** chandankumar has quit IRC | 17:05 | |
| slagle | jaosorior: what type of dependency? | 17:05 |
| jaosorior | slagle: another puppet module | 17:05 |
| *** rasca has quit IRC | 17:05 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Make the CloudDomain defaults match the doc strings https://review.openstack.org/400966 | 17:05 |
| slagle | if required, i think that would be fine given our choice to use puppet | 17:06 |
| jaosorior | alright | 17:06 |
| slagle | we've standardized on using puppet, if we can use it, we should | 17:06 |
| alee | slagle, jaosorior - if there isn't a module readily available, we could construct and do an exec within the manifest? | 17:08 |
| slagle | yes, i see no problem with that | 17:08 |
| *** beagles is now known as beagles-brb | 17:08 | |
| *** achadha has joined #tripleo | 17:09 | |
| slagle | an exec within the manifest would be better than a new bash script as part of an element | 17:09 |
| EmilienM | an exec or a provider | 17:09 |
| openstackgerrit | Merged openstack/tripleo-quickstart: Start using tripleo-quickstart-extras https://review.openstack.org/397700 | 17:10 |
| EmilienM | or find a module that edit resolv.conf | 17:10 |
| EmilienM | like https://github.com/saz/puppet-resolv_conf | 17:10 |
| EmilienM | jaosorior: ^ | 17:11 |
| alee | EmilienM, well - it has the right license | 17:11 |
| jaosorior | EmilienM: is that packaged already in RDO? | 17:11 |
| *** cdearborn has quit IRC | 17:11 | |
| EmilienM | jaosorior: no but we could add it, if we need it | 17:11 |
| EmilienM | like slagle said, that's the right way to do configuration in tripleo | 17:12 |
| jaosorior | well, nice :D | 17:12 |
| *** rhallisey_ has quit IRC | 17:12 | |
| jaosorior | alee: and if it has the right license, even better | 17:12 |
| *** rhallisey_ has joined #tripleo | 17:12 | |
| jaosorior | well, stuff is theoretically going pretty smooth :D | 17:13 |
| *** zoliXXL is now known as zoli|gone | 17:13 | |
| EmilienM | I never tried this module, we should before | 17:13 |
| EmilienM | maybe there is a better one somewhere but this is the one with the most downloads in puppetforge | 17:13 |
| EmilienM | and saz does good modules usually | 17:13 |
| *** panda is now known as panda|bbl | 17:13 | |
| EmilienM | mwhahaha: thoughts? ^ | 17:13 |
| * EmilienM brb lunch | 17:13 | |
| *** flepied1 has quit IRC | 17:14 | |
| jaosorior | alee: I gotta go | 17:14 |
| jaosorior | alee: but if you want me to continue with something you started, or test it out, send me a mail | 17:14 |
| alee | jaosorior, sure - thanls | 17:14 |
| *** ohamada__ has quit IRC | 17:15 | |
| jaosorior | have a good one everyone! | 17:15 |
| *** jaosorior has quit IRC | 17:15 | |
| alee | EmilienM, I can test it out and confirm that it works. | 17:15 |
| alee | EmilienM, what do we need to do to get it in RDO? | 17:16 |
| mwhahaha | might be able to manage it simply via augeas as well instead of pulling in a module | 17:16 |
| *** yamahata has quit IRC | 17:17 | |
| openstackgerrit | Ronelle Landy proposed openstack/tripleo-quickstart-extras: Add role to create and delete the OVB stack https://review.openstack.org/400997 | 17:17 |
| *** achadha has quit IRC | 17:18 | |
| sshnaidm | derekh, bnemec memory issues again :( http://paste.openstack.org/raw/590238/ | 17:18 |
| *** abehl has joined #tripleo | 17:19 | |
| sshnaidm | 58 stack atm | 17:20 |
| openstackgerrit | Julie Pichon proposed openstack/tripleo-heat-templates: Make the CloudDomain defaults match the doc strings https://review.openstack.org/401359 | 17:20 |
| alee | mwhahaha, EmilienM - I can certainly do that too. | 17:20 |
| mwhahaha | bcause doesn't os-net-config also manage resolv.conf? | 17:20 |
| bnemec | sshnaidm: Weird, we're not even close to the limit right now though: | RAM | 2665472 | 3145728 | | 17:21 |
| bnemec | There must have been a lot of job cancellations 5 hours ago. | 17:22 |
| *** achadha has joined #tripleo | 17:22 | |
| sshnaidm | bnemec, it was 3 hours ago if I understand right timestamps | 17:23 |
| bnemec | sshnaidm: I'm seeing 17:21:55.786 as the timestamp in the logs right now. | 17:23 |
| bnemec | Your log is 12:47:21.325 | 17:23 |
| bnemec | Not quite 5 hours, but close. | 17:24 |
| bnemec | That also corresponds to a spike in activity on http://grafana.openstack.org/dashboard/db/nodepool-tripleo-test-cloud | 17:24 |
| *** tzumainn has joined #tripleo | 17:25 | |
| alee | mwhahaha, I dont see anything in templates/net-config.json.template that mentions nameservers | 17:25 |
| mwhahaha | i'll look in a bit, in a meeting | 17:25 |
| alee | mwhahaha, sure - thanks | 17:25 |
| bnemec | alee: It is possible to set dns servers with os-net-config. See the test example here: https://github.com/openstack/os-net-config/blob/e479535b508b4072f64300809b824dfecc3a9182/os_net_config/tests/test_objects.py#L126 | 17:27 |
| sshnaidm | bnemec, yeah, it's huge peak at 12.00 on chart | 17:27 |
| *** achadha has quit IRC | 17:27 | |
| *** abregman|mtg has quit IRC | 17:28 | |
| *** hogepodge has joined #tripleo | 17:28 | |
| alee | bnemec, this modifies /etc/resolv.conf on the undercloud node? | 17:28 |
| bnemec | alee: Not exactly. It sets the dns servers in the ifcfg interface definition, and then that ends up in resolv.conf (I think). | 17:29 |
| bnemec | Interface-specific DNS servers has never entirely made sense to me. | 17:30 |
| EmilienM | alee: I like bnemec's proposal too! | 17:31 |
| EmilienM | alee: if we could investigate that way and see how it works, it would be great. Otherwise puppet/augeas would also do the job | 17:31 |
| alee | bnemec, EmilienM - ok -- let me try it out and see it works. | 17:31 |
| *** achadha has joined #tripleo | 17:32 | |
| *** lucasagomes is now known as lucas-hungry | 17:33 | |
| *** lucas-hungry is now known as lucas-afk | 17:33 | |
| *** hogepodge has quit IRC | 17:33 | |
| *** achadha has quit IRC | 17:37 | |
| *** florianf has quit IRC | 17:38 | |
| *** weshay_lunch is now known as weshay | 17:42 | |
| *** shardy has quit IRC | 17:43 | |
| *** hogepodge has joined #tripleo | 17:43 | |
| openstackgerrit | Dan Radez proposed openstack/tripleo-heat-templates: Adding Congress Support https://review.openstack.org/330050 | 17:44 |
| openstackgerrit | Merged openstack/tripleo-common: Use kwargs to pass in data and error to Mistral Result https://review.openstack.org/375348 | 17:45 |
| openstackgerrit | Merged openstack/tripleo-common: Revert changes to CephClusterFSID generation https://review.openstack.org/401267 | 17:46 |
| openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Deploy network envs appropriate for the job https://review.openstack.org/347503 | 17:47 |
| openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Test with bonded public interface https://review.openstack.org/385562 | 17:47 |
| *** chlong has quit IRC | 17:50 | |
| *** tzumainn has quit IRC | 17:50 | |
| *** cdearborn has joined #tripleo | 17:50 | |
| openstackgerrit | Marios Andreou proposed openstack/tripleo-heat-templates: Fix ovs 2.4 to 2.5 upgrade - minor update non controllers https://review.openstack.org/401365 | 17:50 |
| openstackgerrit | Miles Gould proposed openstack/diskimage-builder: Allow installing proliant tools from an OS package https://review.openstack.org/344214 | 17:52 |
| *** derekh has quit IRC | 17:52 | |
| *** oneswig has quit IRC | 17:53 | |
| *** achadha_ has joined #tripleo | 17:54 | |
| *** milan has quit IRC | 17:56 | |
| *** spredzy has quit IRC | 17:58 | |
| *** achadha has joined #tripleo | 17:58 | |
| *** achadha_ has quit IRC | 17:59 | |
| *** spredzy has joined #tripleo | 17:59 | |
| *** achadha has quit IRC | 18:00 | |
| *** lmiccini has quit IRC | 18:01 | |
| *** achadha has joined #tripleo | 18:01 | |
| *** beagles-brb is now known as beagles | 18:02 | |
| *** amoralej is now known as amoralej|off | 18:02 | |
| *** tzumainn has joined #tripleo | 18:03 | |
| openstackgerrit | Brent Eagles proposed openstack/tripleo-heat-templates: Run os-net-config before restarting cluster on update https://review.openstack.org/400960 | 18:08 |
| *** tosky has quit IRC | 18:09 | |
| *** chandankumar has joined #tripleo | 18:11 | |
| *** parora has joined #tripleo | 18:13 | |
| openstackgerrit | Brent Eagles proposed openstack/tripleo-heat-templates: Run os-net-config before restarting cluster on update https://review.openstack.org/400960 | 18:13 |
| openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Hiera optimization: use a new hiera hook https://review.openstack.org/384757 | 18:14 |
| *** jpena is now known as jpena|off | 18:15 | |
| *** prateek_ has quit IRC | 18:16 | |
| *** noslzzp has quit IRC | 18:16 | |
| *** noslzzp has joined #tripleo | 18:17 | |
| *** dtantsur is now known as dtantsur|afk | 18:24 | |
| *** prateek_ has joined #tripleo | 18:24 | |
| *** parora has quit IRC | 18:27 | |
| *** chandankumar has quit IRC | 18:29 | |
| openstackgerrit | Miles Gould proposed openstack/diskimage-builder: Allow installing proliant tools from an OS package https://review.openstack.org/344214 | 18:31 |
| alee | bnemec, EmilienM , sweet - looks like it worked :) | 18:33 |
| *** paramite has quit IRC | 18:33 | |
| alee | bnemec, EmilienM patch coming shortly .. | 18:33 |
| *** alee is now known as alee_lunch | 18:35 | |
| *** cylopez has joined #tripleo | 18:35 | |
| *** nyechiel has quit IRC | 18:35 | |
| *** cylopez has quit IRC | 18:37 | |
| *** Guest83036 has quit IRC | 18:38 | |
| *** sudipto_ has quit IRC | 18:38 | |
| *** prateek_ has quit IRC | 18:38 | |
| openstackgerrit | Miles Gould proposed openstack/diskimage-builder: Allow installing proliant tools from an OS package https://review.openstack.org/344214 | 18:42 |
| *** abehl has quit IRC | 18:47 | |
| *** rhallisey_ has quit IRC | 18:49 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Add necessary parameters for encrypted volumes support https://review.openstack.org/398177 | 18:50 |
| *** newmember has quit IRC | 18:50 | |
| *** yamahata has joined #tripleo | 18:54 | |
| *** ctayal has quit IRC | 18:56 | |
| jkilpatr_ | just started using the extras repo using rlandy's commit, easy transition if there ever was one, took me 5 min, hardest part was pulling the gerrit since it isn't merged yet | 19:02 |
| rlandy | jkilpatr_; I am going to copy your command above and paste it on my wall :) | 19:05 |
| rlandy | comment | 19:05 |
| jkilpatr_ | rlandy, I was prepared for a lot more pain on my end from this whole transition but its worked out to be pretty trivial on my end | 19:06 |
| jkilpatr_ | far better than I expected | 19:06 |
| *** bana_k has joined #tripleo | 19:10 | |
| *** cylopez has joined #tripleo | 19:11 | |
| sshnaidm | weshay, I think dprince manages the site (am I wrong?) | 19:14 |
| weshay | sshnaidm, I have no clue | 19:15 |
| sshnaidm | dprince, ^^ | 19:15 |
| weshay | sshnaidm, I have another pull request for you | 19:15 |
| dprince | sshnaidm: yep, I setup tripleo.org | 19:15 |
| sshnaidm | weshay, ^ | 19:16 |
| sshnaidm | weshay, got it | 19:17 |
| weshay | dprince, any interest in updating the status sight w/ Sagi's page? | 19:19 |
| weshay | http://status-tripleoci.rhcloud.com/ | 19:19 |
| weshay | site | 19:20 |
| weshay | rather | 19:20 |
| dprince | weshay: I'm interested. The current status page is static. Which is kind of nice in some regards | 19:20 |
| dprince | weshay: if I recall that version required a web application/database, etc. | 19:20 |
| weshay | flask | 19:21 |
| weshay | dprince, I suppose we could redirect and leave it on public openshift as well.. that may be less maint | 19:21 |
| weshay | don't know what is preferred | 19:21 |
| *** beagles is now known as beagles_brb | 19:21 | |
| *** fragatina has quit IRC | 19:22 | |
| dprince | weshay: perhaps architecting it to generate a static page would be possible, and then pull request it | 19:22 |
| *** pkovar has quit IRC | 19:22 | |
| dprince | weshay: or, update the generate_site.sh script to setup flask and update the site accordingly | 19:22 |
| weshay | ya | 19:23 |
| dprince | weshay: all of the current code is in the tripleo-ci repo | 19:23 |
| weshay | topic for next mtg I guess | 19:23 |
| sshnaidm | dprince, weshay but it'll be a webapp, with db, won't be this a problem? | 19:23 |
| sshnaidm | dprince, all site code in tripleo-ci repo?? | 19:23 |
| weshay | there are plans for a db, but none atm | 19:23 |
| dprince | sshnaidm: it isn't a problem outright. Just a change that would require care and an automatic update procedure so we could easily update it | 19:24 |
| dprince | sshnaidm: right now, any commits to the tripleo-ci repo go live and should be rendered within 10 minutes automatically | 19:24 |
| dprince | sshnaidm: ideally we would want a similar setup if we went full on web application, which is why I sort of prefer a cron script that generates a static page | 19:25 |
| dprince | sshnaidm: you could use a sqlite database I think and have it spit out a static page FWIW | 19:25 |
| sshnaidm | dprince, I think it's better to have a setup task for this site in tripleo-ci, but the site code itself to keep out | 19:26 |
| dprince | sshnaidm: fair enough. We'd also want to style the report so it fits into tripleo.org too | 19:26 |
| dprince | sshnaidm: we generate a template that is used by the CI report generator | 19:27 |
| sshnaidm | dprince, ok... css sorcerers are wanted :) | 19:27 |
| dprince | sshnaidm: everything is currently architected to be super light on maintenance, very little duplication. i.e. if we update the style I don't have to change it in multiple places | 19:28 |
| *** alee_lunch is now known as alee | 19:28 | |
| dprince | sshnaidm: I don't think what we have is beautiful or anything, but it mostly runs without maintenance | 19:28 |
| sshnaidm | dprince, I don't see a reason to maintain it special way.. | 19:30 |
| sshnaidm | dprince, either I don't follow you | 19:30 |
| sshnaidm | dprince, weshay I think it's worth to talk about 10 mins in bluejeans about it..? | 19:30 |
| *** abregman has joined #tripleo | 19:31 | |
| dprince | sshnaidm: if you change the webapp, we just need to fully automate it so it gets upgraded each time | 19:31 |
| dprince | sshnaidm: perhaps even fully re-installing it if needed if it errors out or something | 19:31 |
| *** ctayal has joined #tripleo | 19:32 | |
| dprince | sshnaidm: so long as we have those peices I'd be fine with running the new report on tripleo.org | 19:32 |
| weshay | I kind of like openshift a bit.. 2 cents because it is very much free of any maintenance | 19:32 |
| sshnaidm | dprince, it's if we host it on your server you mean | 19:32 |
| dprince | sshnaidm: right | 19:32 |
| sshnaidm | dprince, if it stays on openshift, you don't need to be worried about it maintanance at all | 19:33 |
| bnemec | Could we just embed the openshift page on tripleo.org? | 19:33 |
| dprince | sshnaidm: not entirely | 19:33 |
| sshnaidm | yep, that would be good start | 19:33 |
| dprince | sshnaidm: I would like any report we rely on to be easily maintainable by the team | 19:34 |
| dprince | sshnaidm: which is why the current code lives in tripleo-ci for the most part | 19:34 |
| weshay | the code can live in tripleo-ci | 19:34 |
| sshnaidm | it doesn't make sense to reinstall the whole webapp each commit, this is static page way | 19:35 |
| dprince | the goal here would be, if we (tripleo core) +2 a patch for this report/site it automatically gets deployed | 19:35 |
| weshay | aye.. agree | 19:35 |
| weshay | so the code just needs to move | 19:35 |
| dprince | sshnaidm: I wasn't suggesting re-installing on each commit was required. I just want an automated deployment mechanism for what we come up with here | 19:36 |
| openstackgerrit | John Fulton proposed openstack/tripleo-specs: Integrate TripleO with Tendrl for External Storage Deployment/Management https://review.openstack.org/387631 | 19:36 |
| *** tzumainn has quit IRC | 19:37 | |
| *** abregman is now known as abregman|afk | 19:38 | |
| sshnaidm | dprince, repeating bnemec question, could we as a first step to embed it into tripleo.org as is? | 19:39 |
| sshnaidm | I mean just show it there | 19:39 |
| bnemec | weshay: sshnaidm: On a related tangent, is that any sort of "official" openshift account? | 19:40 |
| bnemec | I'm wondering because I've been meaning to look for a more permanent home for http://ucw-bnemec.rhcloud.com/ | 19:41 |
| dprince | sshnaidm: we could, it would likely require work on both sides because ideally it would get styled with the owl, etc. | 19:41 |
| dprince | sshnaidm: so the webapp would need to use our template | 19:41 |
| openstackgerrit | Lukas Bezdicka proposed openstack/tripleo-heat-templates: Fix ovs 2.4 to 2.5 upgrade - minor update non controllers https://review.openstack.org/401365 | 19:41 |
| bnemec | I would love to add that to the docs, but it seems weird to link to my personal openshift account. | 19:41 |
| dprince | sshnaidm: and then we coul ddisplay it alongside of the other pages, sure | 19:41 |
| sshnaidm | dprince, could you please write/send me what is required exactly for this? | 19:42 |
| weshay | bnemec, pretty much the same thing, I was granted an upgrade but it's just a regular account | 19:42 |
| dprince | sshnaidm: this is how we generate a template now http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/scripts/website/generate_site.sh#n109 | 19:43 |
| dprince | sshnaidm: everything happens in that script pretty much | 19:43 |
| dprince | sshnaidm: if you could have openshift download/generate that template and then render the report in a similar fashion I think it'd work fine | 19:43 |
| weshay | I like that | 19:44 |
| bnemec | weshay: Okay, maybe I'll look into opening a tripleo account. ucw-tripleo wouldn't be a bad name for it. | 19:46 |
| sshnaidm | dprince, to render the static page you mean | 19:46 |
| *** cylopez has quit IRC | 19:47 | |
| weshay | bnemec, I think if RH would pony up for the expense, we could have the whole site w/ the proper domain name hosted on openshift | 19:48 |
| weshay | maybe we would require that openshift was running on openstack first | 19:48 |
| sshnaidm | weshay, and new openshift, with docker! | 19:48 |
| *** panda|bbl is now known as panda | 19:49 | |
| openstackgerrit | Ade Lee proposed openstack/instack-undercloud: Add option to set undercloud dns nameserver https://review.openstack.org/400937 | 19:49 |
| openstackgerrit | Ade Lee proposed openstack/instack-undercloud: Add code to support novajoin in the undercloud https://review.openstack.org/399220 | 19:49 |
| bnemec | weshay: That would be nice. OpenShift is pretty slick for this sort of thing - just push a commit and it does the magic to re-deploy your site with the changes. | 19:51 |
| weshay | bnemec, ya man.. :) | 19:51 |
| bnemec | The only thing I recall being a potential issue was that the git repo is single user, so you might need to set up a sync job from the gerrit-fronted git repo to the openshift one. | 19:51 |
| dprince | bnemec: why not run the entire site on openshift then? | 19:52 |
| bnemec | dprince: We probably could. | 19:54 |
| *** cylopez has joined #tripleo | 19:54 | |
| bnemec | Although I think the web apps would still need to be separate from the static html. | 19:54 |
| openstackgerrit | Pradeep Kilambi proposed openstack/python-tripleoclient: Add Panko password to deployment https://review.openstack.org/401402 | 19:55 |
| *** beagles_brb is now known as beagles | 19:55 | |
| *** fragatina has joined #tripleo | 19:56 | |
| *** jpich has quit IRC | 20:03 | |
| *** ctayal has quit IRC | 20:03 | |
| *** achadha_ has joined #tripleo | 20:04 | |
| *** achadha has quit IRC | 20:07 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: Explicitly set rabbit hosts so its not overridden during upgrade https://review.openstack.org/401404 | 20:08 |
| openstackgerrit | Alejandro Andreu proposed openstack/tripleo-puppet-elements: Separate Datastax repository from the Midonet one https://review.openstack.org/394978 | 20:08 |
| *** achadha_ has quit IRC | 20:09 | |
| *** dbecker has quit IRC | 20:11 | |
| openstackgerrit | John Fulton proposed openstack/tripleo-specs: Integrate TripleO with Tendrl for External Storage Deployment/Management https://review.openstack.org/387631 | 20:11 |
| openstackgerrit | John Fulton proposed openstack/tripleo-specs: Integrate TripleO with Tendrl for External Storage Deployment/Management https://review.openstack.org/387631 | 20:14 |
| openstackgerrit | Alejandro Andreu proposed openstack/tripleo-puppet-elements: Separate Datastax repository from the Midonet one https://review.openstack.org/394978 | 20:18 |
| openstackgerrit | Alejandro Andreu proposed openstack/tripleo-puppet-elements: Enable MidoNet element to install MEM https://review.openstack.org/401408 | 20:18 |
| *** abregman|afk has quit IRC | 20:21 | |
| openstackgerrit | Carlos Camacho proposed openstack-infra/tripleo-ci: Add Panko to scenario001 https://review.openstack.org/401299 | 20:22 |
| *** nyechiel has joined #tripleo | 20:24 | |
| *** cylopez has left #tripleo | 20:29 | |
| *** fragatina has quit IRC | 20:30 | |
| *** paramite has joined #tripleo | 20:34 | |
| *** shinobu_ has quit IRC | 20:38 | |
| *** paramite has quit IRC | 20:41 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Fix ovs 2.4 to 2.5 upgrade - minor update non controllers https://review.openstack.org/401195 | 20:44 |
| *** dprince has quit IRC | 20:46 | |
| *** achadha has joined #tripleo | 20:47 | |
| *** newmember has joined #tripleo | 20:47 | |
| *** flaper87 has quit IRC | 20:52 | |
| *** ansmith has quit IRC | 20:59 | |
| *** fragatina has joined #tripleo | 21:01 | |
| *** jprovazn has quit IRC | 21:03 | |
| *** TSCHAK has quit IRC | 21:08 | |
| -openstackstatus- NOTICE: Due to a problem with the cinder volume backing the log server, jobs are failing with POST_FAILURE. Please avoid issuing 'recheck' commands until the issue is resolved. | 21:11 | |
| *** ChanServ changes topic to "Due to a problem with the cinder volume backing the log server, jobs are failing with POST_FAILURE. Please avoid issuing 'recheck' commands until the issue is resolved." | 21:11 | |
| *** dhill_ has quit IRC | 21:11 | |
| panda | ouch | 21:11 |
| *** morazi has quit IRC | 21:18 | |
| *** jayg is now known as jayg|g0n3 | 21:20 | |
| *** jkilpatr_ has quit IRC | 21:22 | |
| *** nyechiel has quit IRC | 21:26 | |
| *** jkilpatr_ has joined #tripleo | 21:37 | |
| openstackgerrit | Ade Lee proposed openstack/tripleo-quickstart: changes to add novajoin to undercloud https://review.openstack.org/398772 | 21:41 |
| openstackgerrit | Ade Lee proposed openstack/tripleo-quickstart: Add option to set DNS name server on the undercloud https://review.openstack.org/401452 | 21:41 |
| *** rbowen has quit IRC | 21:42 | |
| *** pradk has quit IRC | 21:43 | |
| *** dhill_ has joined #tripleo | 21:44 | |
| *** ctayal has joined #tripleo | 21:50 | |
| *** yolanda has quit IRC | 21:50 | |
| *** jcoufal has quit IRC | 21:55 | |
| *** oneswig has joined #tripleo | 22:02 | |
| *** ccamacho has quit IRC | 22:11 | |
| *** bana_k has quit IRC | 22:16 | |
| *** dmarlin has quit IRC | 22:16 | |
| *** oneswig has quit IRC | 22:17 | |
| *** slagle has quit IRC | 22:18 | |
| *** chlong has joined #tripleo | 22:19 | |
| *** saibarspeis has joined #tripleo | 22:21 | |
| *** saibarspeis has quit IRC | 22:21 | |
| dsneddon | beagles, Should I merge now that it passed CI? https://review.openstack.org/#/c/400960/ | 22:24 |
| beagles | dsneddon, not yet thanks... I haven't managed to give it a run through a proper update yet. Hopefully later this evening | 22:25 |
| EmilienM | beagles: I'm back, what's up about nneutron? | 22:26 |
| dsneddon | beagles, Yeah, I was wondering if that did the trick. | 22:26 |
| beagles | EmilienM, looking into the test suite at the moment (among other things), there was some weird crap with security groups which kind of ties in with some weird OVS logging around that port | 22:27 |
| beagles | EmilienM, juggling a few blockers at the moment :( | 22:28 |
| EmilienM | beagles: don't ask me why but I've lost the launchpa dlink | 22:28 |
| EmilienM | can I have it again please? | 22:28 |
| beagles | EmilienM, yup 1s | 22:28 |
| EmilienM | oh I found it https://bugs.launchpad.net/neutron/+bug/1644312 | 22:29 |
| openstack | Launchpad bug 1644312 in neutron "VIF plug notifications from neutron->nova are not always produced when running tempest" [Undecided,New] | 22:29 |
| beagles | EmilienM, yup that's it | 22:29 |
| EmilienM | beagles: ok so no progress, right? | 22:29 |
| EmilienM | beagles: i'll try to bump our CI to a previous commit, from yesterday | 22:29 |
| beagles | EmilienM, basically yeah | 22:29 |
| *** chlong has quit IRC | 22:30 | |
| openstackgerrit | Brent Eagles proposed openstack/puppet-tripleo: Allow firewall rules that aren't state matching https://review.openstack.org/401461 | 22:35 |
| *** panda is now known as panda|zZ | 22:35 | |
| *** limao has quit IRC | 22:36 | |
| *** beagles is now known as beagles_brb | 22:43 | |
| *** mhenkel has quit IRC | 22:44 | |
| *** mhenkel_ has joined #tripleo | 22:44 | |
| *** ChanServ changes topic to "TripleO : http://tripleo.org/ | https://wiki.openstack.org/wiki/TripleO | Meetings On Tuesdays at 14:00 UTC in #openstack-meeting-alt" | 22:55 | |
| -openstackstatus- NOTICE: The affected filesystems on the log server are repaired. Please leave 'recheck' comments on any changes which failed with POST_FAILURE. | 22:55 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: yum-minimal: add systemd to initial install https://review.openstack.org/401504 | 23:00 |
| *** iranzo has quit IRC | 23:01 | |
| openstackgerrit | Ronelle Landy proposed openstack/tripleo-quickstart-extras: WIP: Use generic cloud image for undercloud https://review.openstack.org/401548 | 23:04 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Catch errors in DIB_INIT_SYSTEM export https://review.openstack.org/400542 | 23:04 |
| *** rlandy is now known as rlandy|bbl | 23:06 | |
| *** thrash is now known as thrash|g0bble | 23:07 | |
| *** beagles_brb is now known as beagles | 23:14 | |
| openstackgerrit | Alex Schultz proposed openstack/python-tripleoclient: Implement overcloud delete command https://review.openstack.org/401024 | 23:14 |
| openstackgerrit | Brent Eagles proposed openstack/puppet-tripleo: Do not configure state matching when using GRE https://review.openstack.org/401461 | 23:16 |
| *** dkehn__ has quit IRC | 23:29 | |
| *** cdearborn has quit IRC | 23:31 | |
| *** jbadiapa has quit IRC | 23:39 | |
| openstackgerrit | Alex Schultz proposed openstack/puppet-tripleo: Standardize httpchk for http services https://review.openstack.org/394686 | 23:39 |
| *** nyechiel has joined #tripleo | 23:42 | |
| *** achadha has quit IRC | 23:43 | |
| *** achadha has joined #tripleo | 23:44 | |
| *** achadha_ has joined #tripleo | 23:48 | |
| *** achadha has quit IRC | 23:52 | |
| *** achadha_ has quit IRC | 23:53 | |
| *** jbadiapa has joined #tripleo | 23:53 | |
| *** achadha has joined #tripleo | 23:54 | |
| *** newmember has quit IRC | 23:57 | |
| *** achadha has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!