| *** achanda has joined #kolla | 00:06 | |
| *** achanda has quit IRC | 00:44 | |
| *** zhiwei has joined #kolla | 00:56 | |
| *** weiyu has joined #kolla | 01:06 | |
| *** erkules_ has joined #kolla | 01:23 | |
| *** erkules has quit IRC | 01:26 | |
| *** achanda has joined #kolla | 01:37 | |
| *** Haomeng has joined #kolla | 02:18 | |
| *** achanda has quit IRC | 02:22 | |
| *** achanda has joined #kolla | 02:24 | |
| *** weiyu has quit IRC | 02:27 | |
| *** weiyu has joined #kolla | 02:30 | |
| *** dayou has joined #kolla | 02:36 | |
| *** alisonh has quit IRC | 02:56 | |
| *** alisonh has joined #kolla | 03:00 | |
| *** klint has joined #kolla | 03:02 | |
| SamYaple | sdake_: you introduced that neutron regression | 03:13 |
|---|---|---|
| SamYaple | https://github.com/stackforge/kolla/commit/ce86de9915d24f3a3dc7688ff66a536b6adb780b | 03:13 |
| *** apmelton has quit IRC | 03:28 | |
| *** apmelton has joined #kolla | 03:30 | |
| *** jtriley has joined #kolla | 03:56 | |
| *** jtriley has quit IRC | 04:11 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Remove dest_filename from build.ini https://review.openstack.org/218718 | 04:11 |
| openstackgerrit | Kai Qiang Wu(Kennan) proposed stackforge/kolla: Improve scripts with some check before kick rm command https://review.openstack.org/218719 | 04:14 |
| *** jtriley has joined #kolla | 04:18 | |
| *** jtriley has quit IRC | 04:35 | |
| *** weiyu has quit IRC | 04:38 | |
| *** jasonsb has joined #kolla | 04:50 | |
| openstackgerrit | Kai Qiang Wu(Kennan) proposed stackforge/kolla: Improve scripts with some check before kick rm command https://review.openstack.org/218719 | 05:14 |
| *** weiyu has joined #kolla | 05:22 | |
| openstackgerrit | Merged stackforge/kolla: Remove dest_filename from build.ini https://review.openstack.org/218718 | 05:24 |
| openstackgerrit | Merged stackforge/kolla: Fix KOLLA_CONFIG_STRATEGY in neutron agents https://review.openstack.org/218682 | 05:25 |
| *** sdake_ has quit IRC | 05:25 | |
| *** sdake has joined #kolla | 05:26 | |
| *** sdake has quit IRC | 05:29 | |
| *** sdake has joined #kolla | 05:30 | |
| *** chandankumar has joined #kolla | 05:31 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: make nova-libvirt is child of base https://review.openstack.org/218635 | 05:35 |
| sdake | samyaple is cgorups /sys/cgroups on ubuntu? | 05:36 |
| sdake | its /sys/fs/cgroups on centos | 05:36 |
| SamYaple | its nothing on ubuntu, since ubuntu doesnt use systemd (yet) | 05:36 |
| SamYaple | its the same everywhere though /sys/fs/cgroup sounds right | 05:37 |
| sdake | i'll give that a go and see if it works | 05:38 |
| sdake | what is odd is this is recent | 05:38 |
| sdake | i could nova boot previously | 05:38 |
| SamYaple | doubtful or there are other issue at play. the cgropu bindings were never in ansible | 05:38 |
| SamYaple | if you could boot after an ansible deploy you should porbalby not add that mount and figure out what else is going on | 05:38 |
| sdake | dude look at the log | 05:39 |
| SamYaple | the less mounts we add in the better, and im frankly not positive the bind is needed | 05:39 |
| sdake | it tries to read a fil ewhich is not present | 05:39 |
| SamYaple | im aware | 05:39 |
| sdake | it is present outside the container but not inside | 05:39 |
| SamYaple | im aware, but you can make it present without a bind mount is my point | 05:40 |
| sdake | my suspicioon is recent change to libvirt | 05:40 |
| sdake | how can you make it present? | 05:40 |
| SamYaple | additionally maybe it isn't needed if you change some options | 05:40 |
| SamYaple | youd have to run systemd in the container I believe | 05:42 |
| SamYaple | i think we remove it | 05:42 |
| sdake | we are definately not doing that | 05:42 |
| SamYaple | i dont care what you do so long as ubuntu is unaffected is my point | 05:42 |
| sdake | ubuntu is unaffected by a bindmount to /sys/fs/cgroups | 05:43 |
| sdake | asuming that works | 05:43 |
| SamYaple | agreed thats why i suggested it | 05:43 |
| sdake | i'll test in a moment | 05:44 |
| SamYaple | /sys/fs/cgroup | 05:44 |
| SamYaple | btw | 05:44 |
| sdake | right | 05:44 |
| *** sdake_ has joined #kolla | 05:45 | |
| *** sdake has quit IRC | 05:49 | |
| sdake_ | samyaple how oten to you rebuild your images? | 05:56 |
| sdake_ | this /sys/run/cgroup thing is new in the last few days | 05:56 |
| SamYaple | 3-4 times a day with no cache | 05:56 |
| sdake_ | rathr sys/fs/cgroup | 05:56 |
| *** weiyu has quit IRC | 05:57 | |
| *** weiyu has joined #kolla | 06:06 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Nova requires a /sys/fs/cgroup bindmount https://review.openstack.org/218683 | 06:08 |
| *** jasonsb has quit IRC | 06:10 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: WIP: Ceph Ansible https://review.openstack.org/218570 | 06:18 |
| SamYaple | sdake_: im not sure nova_compute needs sys | 06:19 |
| SamYaple | nova_compute should just be a process talking to libvirt | 06:19 |
| Kennan | sdake_ there? | 06:36 |
| Kennan | or SamYaple: ? | 06:36 |
| sdake_ | yup | 06:36 |
| Kennan | I checked neutron-agents container | 06:36 |
| *** sdake_ is now known as sdake | 06:36 | |
| Kennan | all seems agents exited ? | 06:36 |
| sdake | there is a patch in the queue for that | 06:36 |
| Kennan | which one sdake? | 06:36 |
| *** mandre has joined #kolla | 06:37 | |
| sdake | https://review.openstack.org/#/c/218682/ | 06:38 |
| sdake | looks merged | 06:38 |
| sdake | try pull | 06:38 |
| sdake | and then rebuild your neutorn related containers | 06:38 |
| Kennan | ok sdake | 06:39 |
| Kennan | BTW: sdake | 06:40 |
| Kennan | I want to try the multi-node for kolla | 06:40 |
| sdake | it looks like centos+binry doesn't work for multinode | 06:41 |
| Kennan | is it nearly same as all-in-one ? just at the last step use playbook | 06:41 |
| sdake | but possibly ubuntu does, would lie confirmtion tho :) | 06:42 |
| sdake | you have to modify multinode inventory file | 06:42 |
| Kennan | ansible-playbook -i inventory/multinode -e @/etc/kolla/globals.yml -e @/etc/kolla/passwords.yml site.yml | 06:43 |
| Kennan | is it ? | 06:43 |
| Kennan | sdake | 06:43 |
| *** weiyu has quit IRC | 06:44 | |
| Kennan | from multinode inventory file, I find many nodes definiation | 06:44 |
| Kennan | I want to just try one control node and one comupte node, is it possibe ? | 06:44 |
| Kennan | the multinode, seems have control01 control02 control03 network01 compute01 storage01 | 06:45 |
| Kennan | so many nodes | 06:45 |
| *** weiyu_ has joined #kolla | 06:45 | |
| Kennan | :q | 06:46 |
| sdake | yes 1 control 1 compute should work | 06:47 |
| sdake | the rason we ru 3 control nodes is for ha | 06:47 |
| sdake | we run nwayactive ha for everything but haproxy which is n+2 ha | 06:48 |
| sdake | rather haproxy is n+m ha, 1+2 is the model | 06:49 |
| *** weiyu_ has quit IRC | 06:52 | |
| *** weiyu has joined #kolla | 06:53 | |
| openstackgerrit | Merged stackforge/kolla: make nova-libvirt is child of base https://review.openstack.org/218635 | 06:54 |
| Kennan | sdake: yes ha required 3 nodes | 06:57 |
| SamYaple | Kennan: you can specify the same host in different groups | 06:57 |
| SamYaple | so you can put 'control01' in the network and storage group | 06:57 |
| Kennan | ok SamYaple: so even ha still can run in one control node ? | 06:57 |
| SamYaple | well no... because one isnt HA | 06:58 |
| SamYaple | buy you can run al lsevices (including compute) on 2 nodes | 06:58 |
| SamYaple | that would be ha | 06:58 |
| SamYaple | granted the way we use quroums galera would lock up if you rebooted the node. you need at least 3 nodes to maintain quorum | 06:58 |
| SamYaple | but nevertheless for testing you can do 2 nodes no problme | 06:59 |
| Kennan | SamYaple: just want to confirm, I would run 2 nodes, HA proxy container can still boot up | 06:59 |
| Kennan | right? | 06:59 |
| SamYaple | yes | 06:59 |
| Kennan | ok. I thought HA must need 3 control nodes | 06:59 |
| Kennan | if one node can setup | 06:59 |
| Kennan | I would first try one | 07:00 |
| SamYaple | HA != haproxy | 07:00 |
| SamYaple | you need more than one node ot be highly availble | 07:00 |
| SamYaple | you can test with 2, but production should never be run without 3 | 07:00 |
| SamYaple | sdake: haproxy is available on all nodes to serve from all nodes. the vip is controlled by keepalived and thats what limits what gets connections | 07:00 |
| Kennan | OK SamYaple: so I used multinode inventory. it would still follow ha-proxy code logic.(although it is not HA mode), but ha related process, like keepalived and haprxoy | 07:02 |
| Kennan | would still run | 07:02 |
| SamYaple | even in all-in-one we run keepalived and haproxy Kennan | 07:02 |
| Kennan | SamYaple, what's that for those process? | 07:03 |
| Kennan | in all-in-one | 07:03 |
| SamYaple | all-in-one is no different than 'multinode', so the default deploy deploys keepalived and haproxy | 07:03 |
| SamYaple | you can actually grow your all-in-one install into multinode install | 07:03 |
| SamYaple | i do that all the time | 07:03 |
| Kennan | SamYaple: just change inventory file, it would grow to multinode | 07:04 |
| Kennan | right? | 07:04 |
| SamYaple | yea | 07:04 |
| Kennan | ok | 07:04 |
| SamYaple | ansible just follows the inventory file | 07:04 |
| SamYaple | so make that file right for your environment and youll be fine | 07:05 |
| Kennan | let me try | 07:05 |
| sdake | kennan make sure your hosts resolve in the inventory file | 07:05 |
| *** achanda has quit IRC | 07:05 | |
| sdake | via dns or via /etc/hosts | 07:05 |
| Kennan | ok sdake | 07:06 |
| SamYaple | sdake: would you mind reviewing https://review.openstack.org/#/c/218565/ | 07:09 |
| SamYaple | i think that syntax is right | 07:09 |
| sdake | in a moment | 07:09 |
| SamYaple | i am begining on some keystone v3 stuff and need ot make sure that is good | 07:09 |
| sdake | code looks good but i am not familiar in dtail with environment variables for openstack client | 07:13 |
| sdake | did you want me to test it? | 07:13 |
| SamYaple | sdake: no it works | 07:14 |
| SamYaple | just want to make sure there are no complaints | 07:14 |
| SamYaple | so i need to switch to v3 instead of v2 for keystone now | 07:14 |
| SamYaple | i was hoping for shade modules by now | 07:14 |
| SamYaple | but that seems... unlikely | 07:14 |
| *** weiyu has quit IRC | 07:15 | |
| *** weiyu has joined #kolla | 07:19 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Remove heat migration based upon heat patch https://review.openstack.org/218634 | 07:42 |
| *** akwasnie has joined #kolla | 07:44 | |
| sdake | folks stay focused on getting us a liberty-3 release we can actually release ;-) | 07:45 |
| sdake | https://launchpad.net/kolla/+milestone/liberty-3 | 07:45 |
| sdake | im off to bed | 07:45 |
| sdake | night | 07:45 |
| openstackgerrit | Hengqing Hu proposed stackforge/kolla: Expose exception objects via the as keyword https://review.openstack.org/218757 | 08:06 |
| *** jmccarthy has joined #kolla | 08:19 | |
| Kennan | hi SamYaple: | 08:34 |
| Kennan | how to use namespace for containers ? I executed on host OS, but seems command not work | 08:34 |
| Kennan | seting the network namespace "qrouter-be324886-d38a-421f-a138-8db49475026b" failed: Invalid argument | 08:34 |
| Kennan | Do I need to execute it on container? | 08:34 |
| SamYaple | Kennan: yes | 08:43 |
| SamYaple | a long conversation can be had on the subject, but basically we are waiting on a docker feature to allow us to have the namespaces outside of the container | 08:44 |
| Kennan | hi SamYaple: if I understand correctly, that neutron related namespace shared the OS itself | 08:45 |
| Kennan | container itself could not split other namespaces | 08:45 |
| *** mandre has quit IRC | 08:45 | |
| Kennan | ip stack namespace would be host OS | 08:45 |
| Kennan | for container it is just shared the host OS | 08:46 |
| Kennan | right? | 08:46 |
| SamYaple | Kennan: no different namespaces | 08:46 |
| SamYaple | its a long conversation | 08:46 |
| SamYaple | the network namespace (holding your physical interfaces) is shared with the host | 08:46 |
| SamYaple | _however_ the network namespaces that neutron creates are _not_ | 08:47 |
| SamYaple | those are different namespaces | 08:47 |
| Kennan | hi SamYaple: for ip netns. OK. you mean it is neutron namespace | 08:49 |
| SamYaple | yes, the network namespaces that neutron creates are _not_ accesible from the host at this time | 08:49 |
| SamYaple | we are waiting on a docker feature that was removed from docker in 1.7.0 | 08:50 |
| *** sdake has quit IRC | 08:50 | |
| *** mandre has joined #kolla | 08:51 | |
| Kennan | SamYaple: then. so right now, the limistaion would be: I need ip netns exec ** ping ** on neutron-agents container | 08:51 |
| Kennan | right ? | 08:51 |
| *** tfukushima has joined #kolla | 08:52 | |
| SamYaple | Kennan: yea. `docker exec neutron_agents ip netns exec q******* ping ***` | 08:52 |
| SamYaple | or enter the container `docker exec -it neutron_agents bash` | 08:52 |
| SamYaple | i dont know how familiar you are with dokcer? | 08:52 |
| Kennan | docker exec -it (sometimes eneter it, ctrl+p, ctril+q can not exit ) so not know if it is good | 08:53 |
| Kennan | maybe ctrl+c | 08:53 |
| SamYaple | Kennan: you can just type `exit` | 08:54 |
| SamYaple | docker exec is different than docker attach | 08:54 |
| SamYaple | exec actually creates a new process | 08:54 |
| Kennan | ok. let me try | 08:54 |
| *** dayou has quit IRC | 08:55 | |
| *** weiyu has quit IRC | 09:00 | |
| *** erkules_ is now known as erkules | 09:04 | |
| *** erkules has joined #kolla | 09:04 | |
| *** weiyu_ has joined #kolla | 09:13 | |
| *** mandre has quit IRC | 09:24 | |
| *** mandre has joined #kolla | 09:26 | |
| *** weiyu_ has quit IRC | 09:35 | |
| *** mandre_ has joined #kolla | 09:40 | |
| *** weiyu_ has joined #kolla | 09:41 | |
| *** mandre has quit IRC | 09:43 | |
| *** mandre_ is now known as mandre | 09:43 | |
| openstackgerrit | James McCarthy proposed stackforge/kolla: Swift WIP - add some missing services, also ansible configuration https://review.openstack.org/218795 | 10:24 |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Fix typo in task bootstrap task names https://review.openstack.org/218798 | 10:32 |
| SamYaple | pbourke coolsvap mandre if you guys have time, Heat could use some review love. Its be dragged through the mud long enough and is functional right now | 10:33 |
| SamYaple | https://review.openstack.org/#/c/205528/ | 10:33 |
| *** weiyu_ has quit IRC | 11:00 | |
| *** weiyu has joined #kolla | 11:02 | |
| *** zhiwei has quit IRC | 11:05 | |
| coolsvap | SamYaple, ack, just one question related to MySQL-python related updates are those related to heat-ansible change? | 11:05 |
| SamYaple | coolsvap: yes | 11:05 |
| SamYaple | the older version ofr MySQL-python has a bug | 11:06 |
| coolsvap | alright, i meant in the changes in ironic, designate n all as well | 11:06 |
| coolsvap | not blocking though, thanks for the patch :) | 11:06 |
| SamYaple | let me check what you are talking about | 11:07 |
| coolsvap | you have removed it from other services as well | 11:07 |
| SamYaple | coolsvap: oh yea that package gets installed in the base now since basically all other services use it | 11:07 |
| SamYaple | its a common dependacy | 11:07 |
| coolsvap | yes | 11:08 |
| * coolsvap nods | 11:08 | |
| vbel | good morning/evening. Would it be good for the project to have tempest container as well? With ansible config :) | 11:11 |
| vbel | just wonder | 11:11 |
| openstackgerrit | Merged stackforge/kolla: Add Ansible support for Heat https://review.openstack.org/205528 | 11:12 |
| openstackgerrit | Merged stackforge/kolla: Keystone container should use the openstackclient https://review.openstack.org/218565 | 11:12 |
| SamYaple | finalllllyy | 11:13 |
| SamYaple | stupid heat taking forever | 11:13 |
| SamYaple | we have heat in kolla now | 11:13 |
| SamYaple | wonderful | 11:13 |
| openstackgerrit | Merged stackforge/kolla: Fix typo in task bootstrap task names https://review.openstack.org/218798 | 11:15 |
| coolsvap | SamYaple, :) | 11:16 |
| SamYaple | pbourke: https://review.openstack.org/#/c/218531/ | 11:16 |
| SamYaple | hopefully that answers your questions | 11:16 |
| pbourke | SamYaple: thanks reading now | 11:18 |
| pbourke | btw are you seeing any permission issues with rabbitmq lately? | 11:18 |
| pbourke | seems to be lingering | 11:19 |
| SamYaple | no but i only use the ubuntu stuff | 11:19 |
| SamYaple | i make sure thats solid | 11:19 |
| SamYaple | if its centos related i havent tried that in a full stack in a month | 11:19 |
| *** weiyu has quit IRC | 11:20 | |
| *** 16WAAONOZ has joined #kolla | 11:21 | |
| *** 16WAAONOZ has quit IRC | 11:22 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: WIP: Ceph Ansible https://review.openstack.org/218570 | 11:26 |
| SamYaple | hey guys that ceph work is almost done | 11:27 |
| SamYaple | its still WIP, but please review it | 11:27 |
| SamYaple | it works mostly, but im testing out some things | 11:27 |
| *** pbourke has quit IRC | 11:30 | |
| *** pbourke has joined #kolla | 11:30 | |
| *** rhallisey has joined #kolla | 11:39 | |
| *** mandre_ has joined #kolla | 11:45 | |
| *** chandankumar has quit IRC | 11:46 | |
| *** openstackgerrit has quit IRC | 11:46 | |
| *** openstackgerrit has joined #kolla | 11:47 | |
| openstackgerrit | Hengqing Hu proposed stackforge/kolla: Expose exception objects via the as keyword https://review.openstack.org/218757 | 11:57 |
| *** dayou has joined #kolla | 12:06 | |
| *** inc0 has joined #kolla | 12:08 | |
| inc0 | good afternoon | 12:09 |
| dayou | :) | 12:09 |
| rhallisey | hey | 12:13 |
| rhallisey | inc0, when are you moving to texas? | 12:14 |
| inc0 | probably just after Tokyo | 12:14 |
| inc0 | all depends on visa process | 12:14 |
| rhallisey | oh cool | 12:14 |
| *** dayou has quit IRC | 12:16 | |
| inc0 | rhallisey, what's main difference between your ceph Dockerfiles and official? | 12:20 |
| SamYaple | lol | 12:20 |
| SamYaple | 'official' | 12:20 |
| SamYaple | those are horribad | 12:20 |
| rhallisey | inc0, the 'official' are bad | 12:20 |
| inc0 | yes, I know | 12:20 |
| inc0 | but why exactly? | 12:20 |
| rhallisey | they are horribly messy | 12:21 |
| rhallisey | let me get an example | 12:21 |
| * SamYaple points to any random 'official' script | 12:21 | |
| rhallisey | look at this -> https://github.com/ceph/ceph-docker/blob/master/osd/entrypoint.sh | 12:22 |
| SamYaple | hey look at my yaodu one https://github.com/SamYaple/yaodu/blob/master/ansible/roles/docker_build/files/ubuntu/ceph/entrypoint.sh | 12:23 |
| *** weiyu has joined #kolla | 12:23 | |
| SamYaple | :D | 12:23 |
| SamYaple | to be fair that is ceph mon and osd | 12:23 |
| rhallisey | their dockerfiles don't have much flow | 12:24 |
| rhallisey | it's just chaos | 12:24 |
| rhallisey | kolla is very defined per service, they need some organization | 12:25 |
| rhallisey | SamYaple, glad you're helping with the ansible ceph piece | 12:25 |
| rhallisey | there are a ton of moving parts I find the whole things to be confusing | 12:26 |
| rhallisey | it's also nice you've done it before | 12:26 |
| SamYaple | yes i have done it before.... its still a pita | 12:26 |
| rhallisey | SamYaple, got a question about the json rework. If a service only has 1 config file specified will that service default to running with that config file? | 12:28 |
| SamYaple | which service? i mean all the services take different options different ways | 12:29 |
| rhallisey | so specifically will neutron for example | 12:29 |
| SamYaple | in the case of memcache that isn't even a true service file | 12:29 |
| SamYaple | neutron will only read the config you give it on the cli as far as im aware | 12:29 |
| rhallisey | say that only 1 config file is found, is it the case neutron will always run with neutron.conf or something | 12:29 |
| rhallisey | oh ok | 12:29 |
| SamYaple | (unless you give it nothing in which case it has defaults to neutorn.conf i think) | 12:30 |
| SamYaple | but the whole l3_agent.ini thing is just a package convention | 12:30 |
| SamYaple | it didnt originate with neutron i dont htink (or it came as a result of what packagers wanted) | 12:30 |
| rhallisey | is there a case where you will only want say `--config-file ml2.ini` | 12:31 |
| SamYaple | you happen to specify all of your neutron options in ml2.ini | 12:31 |
| SamYaple | that would be the only case | 12:31 |
| *** weiyu has quit IRC | 12:31 | |
| SamYaple | i mean in the end neutron is just merging all the files you give it into a big command block | 12:31 |
| rhallisey | right. What I'm wondering is if there's a case where you won't use neutron.conf? | 12:32 |
| rhallisey | and only a single other config file | 12:33 |
| rhallisey | or will you always specify something in neutron.conf | 12:33 |
| *** mandre_ has quit IRC | 12:40 | |
| *** weiyu_ has joined #kolla | 12:42 | |
| SamYaple | i mean if you have all your options in a config you honestly only need to specify one config rhallisey | 12:44 |
| SamYaple | and it doesnt even need to be called neutron.conf | 12:44 |
| SamYaple | but youll always have to do neutron.conf | 12:44 |
| *** mandre_ has joined #kolla | 12:44 | |
| rhallisey | ok I can work with this | 12:45 |
| rhallisey | might be able to get rid of 'is_args' then | 12:45 |
| SamYaple | when i initially propsoed this json thing I was expecting us to place our CMD and ARGS in the json, so it doesn't really matter what options you need ,the deployment otol can generate tehm | 12:46 |
| rhallisey | SamYaple, ok cool | 12:49 |
| *** weiyu_ has quit IRC | 12:52 | |
| *** weiyu_ has joined #kolla | 12:56 | |
| *** mandre has quit IRC | 13:00 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: Ansible bits to rsyslog https://review.openstack.org/215062 | 13:00 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: rsyslog container https://review.openstack.org/214589 | 13:00 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: Ansible bits to rsyslog https://review.openstack.org/215062 | 13:02 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: rsyslog container https://review.openstack.org/214589 | 13:02 |
| *** vbel has quit IRC | 13:06 | |
| *** vbel has joined #kolla | 13:07 | |
| *** klint has quit IRC | 13:10 | |
| *** mandre_ has quit IRC | 13:14 | |
| *** mandre has joined #kolla | 13:15 | |
| *** weiyu_ has quit IRC | 13:17 | |
| *** chandankumar has joined #kolla | 13:18 | |
| *** mandre_ has joined #kolla | 13:20 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: WIP: Ceph Ansible - READY FOR REVIEWS https://review.openstack.org/218570 | 13:22 |
| SamYaple | alright im going home | 13:22 |
| inc0 | http://www.sebastien-han.fr/blog/2015/08/24/ceph-cluster-on-docker-for-testing/ | 13:23 |
| rhallisey | wow that's massive | 13:23 |
| inc0 | nooo SamYaple noo, don't go, stay | 13:23 |
| inc0 | review | 13:23 |
| SamYaple | yea inc0 thats leseb | 13:24 |
| SamYaple | he comes around here every once and a while | 13:24 |
| inc0 | leseb? | 13:24 |
| SamYaple | cool ugy | 13:24 |
| SamYaple | the guy who wrote teh article | 13:25 |
| inc0 | ah I didn't know his irc handle | 13:25 |
| SamYaple | i talked to him for a week or so and didn't realize it was the same guy ;) | 13:25 |
| SamYaple | that reminds me im supposed to be porting Kollas config-merge system to ansible-ceph for him | 13:25 |
| *** mandre has quit IRC | 13:30 | |
| *** mandre_ is now known as mandre | 13:30 | |
| inc0 | SamYaple, btw rsyslog is working, so review plz | 13:31 |
| inc0 | when it's polished up I'll start to port services to use it | 13:31 |
| inc0 | or start with elkstack'ish stuff | 13:31 |
| *** weiyu has joined #kolla | 13:33 | |
| inc0 | also, did you guys solve this haproxy ansible problem? | 13:34 |
| inc0 | about not being able to generate file due to lack of other servers addresses? | 13:34 |
| *** dims has joined #kolla | 13:36 | |
| *** weiyu has quit IRC | 13:43 | |
| *** inc0 has quit IRC | 13:44 | |
| *** Kennan has quit IRC | 14:03 | |
| *** Kennan2 has joined #kolla | 14:03 | |
| *** jasonamyers has joined #kolla | 14:09 | |
| *** ftcpops has joined #kolla | 14:34 | |
| *** mandre_ has joined #kolla | 14:45 | |
| *** jtriley has joined #kolla | 14:46 | |
| *** mandre_ has quit IRC | 14:47 | |
| *** jtriley has quit IRC | 14:48 | |
| *** achanda has joined #kolla | 14:49 | |
| *** banix has joined #kolla | 14:51 | |
| *** dims has quit IRC | 14:51 | |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Always ensure correct ownership of /var/lib/rabbitmq https://review.openstack.org/218890 | 14:52 |
| *** jtriley has joined #kolla | 14:54 | |
| *** banix has quit IRC | 14:57 | |
| *** banix has joined #kolla | 15:01 | |
| *** jruano has joined #kolla | 15:02 | |
| *** dims has joined #kolla | 15:03 | |
| *** achanda has quit IRC | 15:05 | |
| openstackgerrit | Mick Thomspon proposed stackforge/kolla: Add ansible support for Murano https://review.openstack.org/218894 | 15:17 |
| *** blahRus has joined #kolla | 15:22 | |
| vbel | how do you manage to use all-in-one inventory with localhost. For me it fails with kolla-ansible module trying to create database - looks like it tries to shortcut connection via socket file. Normal ip just works | 15:24 |
| vbel | ? | 15:24 |
| vbel | kolla-ansible container | 15:24 |
| *** inc0 has joined #kolla | 15:28 | |
| *** inc0 has quit IRC | 15:36 | |
| bmace | vbel, i had definitely gotten past some of the dbcreate stuff in an all-in-one.. at least the haproxy one, and i think the keystone one at least once and then got hung up on other things | 15:39 |
| bmace | not to say something couldn't have gotten busted since then | 15:39 |
| *** chandankumar has quit IRC | 15:39 | |
| vbel | bmace: thanks, checking it further. Yes, haproxy is the very first | 15:41 |
| vbel | user | 15:41 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: Ansible bits to rsyslog https://review.openstack.org/215062 | 15:43 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed stackforge/kolla: rsyslog container https://review.openstack.org/214589 | 15:43 |
| openstackgerrit | Mick Thomspon proposed stackforge/kolla: Add ansible support for Murano https://review.openstack.org/218894 | 15:46 |
| bmace | i'll try out my all-in-one again this morning based on our latest bits and see how it goes. | 15:47 |
| vbel | just delete users/databases by hand before testing | 15:50 |
| vbel | thanks! | 15:50 |
| *** jasonamyers has quit IRC | 15:51 | |
| bmace | i should need to? rather than just nuking the containers? | 15:51 |
| *** tfukushima has quit IRC | 15:53 | |
| vbel | or, yes, just delete all containers, right :) | 15:53 |
| *** vbel has quit IRC | 15:57 | |
| *** banix has quit IRC | 16:08 | |
| *** mfalatic has joined #kolla | 16:09 | |
| *** unicell has quit IRC | 16:25 | |
| *** jtriley has quit IRC | 16:51 | |
| *** sdake_ has joined #kolla | 16:52 | |
| *** jtriley has joined #kolla | 16:55 | |
| *** dims has quit IRC | 16:55 | |
| *** chandankumar has joined #kolla | 16:59 | |
| *** banix has joined #kolla | 17:00 | |
| sdake_ | morning | 17:07 |
| *** unicell has joined #kolla | 17:12 | |
| *** mandre has quit IRC | 17:14 | |
| *** mandre has joined #kolla | 17:14 | |
| *** mandre_ has joined #kolla | 17:15 | |
| *** mandre has quit IRC | 17:16 | |
| *** mandre_ is now known as mandre | 17:16 | |
| *** achanda has joined #kolla | 17:18 | |
| *** inc0 has joined #kolla | 17:18 | |
| inc0 | good evening guys, did we merge something gate-breaking? https://jenkins02.openstack.org/job/gate-kolla-functional-f21/244/console | 17:18 |
| *** banix has quit IRC | 17:18 | |
| *** jasonsb has joined #kolla | 17:20 | |
| *** banix has joined #kolla | 17:22 | |
| *** banix has quit IRC | 17:25 | |
| *** dims has joined #kolla | 17:29 | |
| *** akwasnie has quit IRC | 17:29 | |
| *** banix has joined #kolla | 17:29 | |
| *** inc0 has quit IRC | 17:29 | |
| *** dims_ has joined #kolla | 17:30 | |
| *** dims has quit IRC | 17:30 | |
| *** jasonsb has quit IRC | 17:33 | |
| *** jasonsb has joined #kolla | 17:33 | |
| *** sdake has joined #kolla | 17:34 | |
| *** jasonsb_ has joined #kolla | 17:37 | |
| *** sdake_ has quit IRC | 17:37 | |
| *** jasonsb has quit IRC | 17:38 | |
| *** loth has joined #kolla | 17:42 | |
| *** loth has left #kolla | 17:42 | |
| *** SourabhP has joined #kolla | 17:45 | |
| *** chandankumar has quit IRC | 17:45 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: nova-libvirt requires a /sys/fs/cgroup bindmount https://review.openstack.org/218683 | 17:45 |
| *** jtriley has quit IRC | 17:47 | |
| *** jtriley has joined #kolla | 17:50 | |
| *** chandankumar has joined #kolla | 17:52 | |
| SamYaple | sdake: it merged already but i figured out why the templates took 5 minutes longer to build | 18:00 |
| SamYaple | nova-libvirt was based on nova-base in the templates rather than 'base' | 18:00 |
| sdake | maybe that has someting to do with the bindmount equirement | 18:03 |
| SamYaple | doesn't sound like a requirement if that has something to do with it ;) | 18:06 |
| *** jtriley has quit IRC | 18:13 | |
| *** jtriley has joined #kolla | 18:17 | |
| *** chandankumar has quit IRC | 18:26 | |
| *** Slower has joined #kolla | 18:28 | |
| *** banix has quit IRC | 18:28 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Fix mysql permissions https://review.openstack.org/218982 | 18:39 |
| SamYaple | sdake: try that out and see if it fixes your multinode issue | 18:39 |
| SamYaple | I would bet money on the fact that it does fix the issue | 18:40 |
| *** jtriley_ has joined #kolla | 18:43 | |
| *** jtriley has quit IRC | 18:46 | |
| *** juggler has quit IRC | 18:47 | |
| *** daneyon_ has joined #kolla | 18:47 | |
| *** daneyon has quit IRC | 18:49 | |
| *** SourabhP has quit IRC | 18:50 | |
| *** chandankumar has joined #kolla | 18:50 | |
| *** SourabhP has joined #kolla | 19:01 | |
| *** sdake has quit IRC | 19:03 | |
| *** jtriley_ has quit IRC | 19:07 | |
| *** sdake has joined #kolla | 19:10 | |
| sdake | samyaple ther eis an open bug for that mysql problem | 19:12 |
| sdake | jpeeler rhallisey mind reviewing https://review.openstack.org/#/c/218683/ | 19:21 |
| rhallisey | ya I noticed that was removed | 19:21 |
| rhallisey | I wasn't sure if it was having any effect though | 19:22 |
| rhallisey | guess it was | 19:22 |
| sdake | https://review.openstack.org/#/c/218531/ that one too | 19:25 |
| *** banix has joined #kolla | 19:34 | |
| sdake | it was removed where? | 19:34 |
| openstackgerrit | Merged stackforge/kolla: Make Horizon work on centos+binary https://review.openstack.org/218531 | 19:36 |
| *** jtriley has joined #kolla | 19:36 | |
| sdake | rhallisey whre was it remoed? | 19:36 |
| rhallisey | sdake, it was apart of docker-compose | 19:36 |
| sdake | well it worked for a long time without it | 19:37 |
| sdake | and recenty stopped working | 19:37 |
| rhallisey | at one point | 19:37 |
| sdake | as of 1-2 days ago | 19:37 |
| sdake | jpeeler re ironic quick q | 19:37 |
| sdake | are you doing ansible deployment by deployign a whole new nova service | 19:38 |
| jpeeler | that was your recommendation and since it requires reconfiguring nova, that makes sense to me | 19:39 |
| sdake | cool you probably don't have to deploy all the service | 19:39 |
| sdake | s | 19:39 |
| sdake | probably conductor api sceduelr only | 19:40 |
| jpeeler | it's compute that gets reconfigured | 19:40 |
| jpeeler | so compute and scheduler | 19:41 |
| sdake | ok | 19:41 |
| *** chandankumar has quit IRC | 19:43 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Fix mysql permissions https://review.openstack.org/218982 | 19:45 |
| sdake | jpeeler can you take a look at https://review.openstack.org/#/c/218982/ | 19:47 |
| openstackgerrit | Merged stackforge/kolla: nova-libvirt requires a /sys/fs/cgroup bindmount https://review.openstack.org/218683 | 19:55 |
| *** jtriley has quit IRC | 19:57 | |
| *** dims has joined #kolla | 20:11 | |
| *** dims has quit IRC | 20:12 | |
| *** dims_ has quit IRC | 20:14 | |
| *** jtriley has joined #kolla | 20:24 | |
| *** diogogmt has joined #kolla | 20:36 | |
| *** achanda has quit IRC | 20:38 | |
| *** achanda has joined #kolla | 20:41 | |
| *** achanda has quit IRC | 20:41 | |
| *** achanda has joined #kolla | 20:41 | |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Replace config-external with a JSON file https://review.openstack.org/216874 | 20:45 |
| *** jtriley has quit IRC | 20:47 | |
| *** ftcpops has quit IRC | 20:48 | |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Replace config-external with a JSON file for Galera https://review.openstack.org/217880 | 20:51 |
| rhallisey | sdake, ^ those look for 'CMD' in the json file | 20:56 |
| rhallisey | all tested. Works with Galera. Just need reviews and more comments on the idea of implementing this way | 20:57 |
| *** dims has joined #kolla | 20:57 | |
| *** rhallisey has quit IRC | 20:58 | |
| *** jpeeler has quit IRC | 21:00 | |
| *** jtriley has joined #kolla | 21:01 | |
| *** jasonsb_ has quit IRC | 21:01 | |
| *** jpeeler has joined #kolla | 21:03 | |
| *** jpeeler has quit IRC | 21:03 | |
| *** jpeeler has joined #kolla | 21:03 | |
| *** ftcpops has joined #kolla | 21:07 | |
| openstackgerrit | Merged stackforge/kolla: Fix mysql permissions https://review.openstack.org/218982 | 21:08 |
| *** banix has quit IRC | 21:10 | |
| *** jasonsb has joined #kolla | 21:13 | |
| *** banix has joined #kolla | 21:14 | |
| sdake | 3 node baremetal deploy: | 21:22 |
| sdake | real8m39.090s | 21:22 |
| sdake | glad we fixed multinode before the deadline ;) | 21:30 |
| sdake | core rviewers - pleae dont ack changes until i tag or request acks | 21:30 |
| *** rhallisey has joined #kolla | 21:34 | |
| *** jtriley has quit IRC | 21:47 | |
| *** jtriley has joined #kolla | 22:08 | |
| *** banix has quit IRC | 22:12 | |
| *** jtriley has quit IRC | 22:13 | |
| *** jruano has quit IRC | 22:14 | |
| *** weiyu has joined #kolla | 22:37 | |
| *** Haomeng|2 has joined #kolla | 22:43 | |
| *** Haomeng has quit IRC | 22:46 | |
| *** ftcpops has quit IRC | 22:46 | |
| *** jruano has joined #kolla | 22:58 | |
| *** jasonsb has quit IRC | 23:13 | |
| *** daneyon has joined #kolla | 23:16 | |
| *** daneyon_ has quit IRC | 23:17 | |
| *** daneyon_ has joined #kolla | 23:18 | |
| *** daneyon has quit IRC | 23:22 | |
| *** jruano has quit IRC | 23:29 | |
| *** dayou has joined #kolla | 23:37 | |
| *** dayou has quit IRC | 23:48 | |
| sdake | samyaple please ping me when you get in | 23:52 |
| *** achanda_ has joined #kolla | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!