*** akwasnie has quit IRC | 00:02 | |
*** jasonsb has joined #kolla | 00:14 | |
*** aginwala has quit IRC | 00:18 | |
*** aginwala has joined #kolla | 00:23 | |
*** shakamunyi has quit IRC | 00:35 | |
*** shakamunyi has joined #kolla | 00:35 | |
*** diogogmt has quit IRC | 00:37 | |
*** mcgregor has quit IRC | 00:37 | |
*** dims_ has joined #kolla | 00:39 | |
*** dims has quit IRC | 00:40 | |
*** asalkeld has quit IRC | 00:56 | |
*** barra204 has joined #kolla | 00:56 | |
*** shakamunyi has quit IRC | 00:57 | |
*** CheKoLyN has joined #kolla | 01:08 | |
*** jtriley has joined #kolla | 01:11 | |
*** barra204 has quit IRC | 01:14 | |
*** shakamunyi has joined #kolla | 01:15 | |
*** tfukushima has joined #kolla | 01:22 | |
*** Jeffrey4l has quit IRC | 01:25 | |
*** unicell1 has quit IRC | 01:27 | |
*** unicell has joined #kolla | 01:27 | |
*** aginwala has quit IRC | 01:30 | |
*** barra204 has joined #kolla | 01:31 | |
*** shakamunyi has quit IRC | 01:32 | |
*** Allen_Gao has joined #kolla | 01:34 | |
*** shakamunyi has joined #kolla | 01:36 | |
*** barra204 has quit IRC | 01:38 | |
*** barra204 has joined #kolla | 01:38 | |
*** tfukushima has quit IRC | 01:39 | |
*** shakamunyi has quit IRC | 01:39 | |
*** tfukushima has joined #kolla | 01:39 | |
*** aginwala has joined #kolla | 01:44 | |
*** shakamunyi has joined #kolla | 01:44 | |
*** barra204 has quit IRC | 01:45 | |
*** dims_ has quit IRC | 01:48 | |
*** shakamunyi has quit IRC | 01:48 | |
*** dims has joined #kolla | 01:48 | |
*** shakamunyi has joined #kolla | 01:49 | |
*** suro-patz has quit IRC | 01:54 | |
*** dims has quit IRC | 01:54 | |
*** shakamunyi has quit IRC | 01:54 | |
*** cemmason has quit IRC | 01:55 | |
*** aginwala has quit IRC | 01:57 | |
*** dims has joined #kolla | 02:06 | |
*** shakamunyi has joined #kolla | 02:08 | |
*** shakamunyi has quit IRC | 02:10 | |
*** shakamunyi has joined #kolla | 02:12 | |
*** shakamunyi has quit IRC | 02:22 | |
*** shakamunyi has joined #kolla | 02:23 | |
openstackgerrit | hardik proposed openstack/kolla: Added Mistral doc link in README.rst https://review.openstack.org/266105 | 02:26 |
---|---|---|
*** cemmason has joined #kolla | 02:30 | |
*** tpot has quit IRC | 02:31 | |
*** shakamunyi has quit IRC | 02:32 | |
*** diogogmt has joined #kolla | 02:33 | |
*** SourabhP has quit IRC | 02:33 | |
*** aginwala has joined #kolla | 02:36 | |
*** jtriley has quit IRC | 02:44 | |
*** salv-orlando has quit IRC | 02:47 | |
*** salv-orlando has joined #kolla | 02:58 | |
*** aginwala has quit IRC | 03:09 | |
*** mdnadeem has joined #kolla | 03:10 | |
*** aginwala has joined #kolla | 03:12 | |
*** SourabhP has joined #kolla | 03:12 | |
*** SourabhP_ has joined #kolla | 03:16 | |
*** aginwala has quit IRC | 03:16 | |
*** SourabhP has quit IRC | 03:17 | |
*** SourabhP_ is now known as SourabhP | 03:17 | |
*** dims has quit IRC | 03:24 | |
*** SourabhP has quit IRC | 03:25 | |
openstackgerrit | Kuo-tung Kao proposed openstack/kolla: add removing containers func for kolla-ansible https://review.openstack.org/264584 | 03:28 |
*** suro-patz has joined #kolla | 03:41 | |
*** tpot has joined #kolla | 03:41 | |
*** Allen_Gao has quit IRC | 03:47 | |
*** klint has joined #kolla | 03:51 | |
*** alisonh has quit IRC | 03:52 | |
*** tfukushima has quit IRC | 03:53 | |
*** tfukushima has joined #kolla | 03:55 | |
*** Allen_Gao has joined #kolla | 04:03 | |
*** salv-orl_ has joined #kolla | 04:10 | |
*** alisonh has joined #kolla | 04:12 | |
*** salv-orlando has quit IRC | 04:13 | |
*** rhallisey has quit IRC | 04:19 | |
*** dratushnyy has joined #kolla | 04:27 | |
*** RA_ has joined #kolla | 04:28 | |
*** coolsvap|away is now known as coolsvap | 04:29 | |
*** tfukushima has quit IRC | 04:30 | |
RA_ | kolla_docker.py looks good, is it going to be put in stable/liberty branch? | 04:30 |
*** alisonh has quit IRC | 04:31 | |
*** alisonh has joined #kolla | 04:34 | |
*** suro-patz has quit IRC | 04:48 | |
*** suro-patz has joined #kolla | 04:49 | |
*** suro-patz has quit IRC | 04:50 | |
nihilifer | RA_: peobably not | 05:04 |
nihilifer | or at least i'd be against backporting such a feature | 05:04 |
*** aginwala has joined #kolla | 05:09 | |
*** cemmason1 has joined #kolla | 05:14 | |
*** cemmason has quit IRC | 05:16 | |
RA_ | @nihilifer what are your thoughts on user kolla/master with openstack liberty? | 05:20 |
nihilifer | RA_: that's impossible. to deploy liberty, please use liberty branch of kolla | 05:22 |
nihilifer | the idea of kolla is to be stick to version of the other openstack project | 05:22 |
nihilifer | so kolla mitaka will deploy mitaka | 05:22 |
nihilifer | and our current master deploys mitaka from master branches of openstack projects :) | 05:22 |
nihilifer | unfortunately, that means that you have to pin version of docker to 1.8.2 when deploying liberty | 05:23 |
RA_ | yeah in our case we are using CentOS 7 and have had to stay with docker-1.7.1-115.el7.x86_64 | 05:26 |
nihilifer | RA_: well, you can always install docker from docker yum repository, not from centos | 05:27 |
nihilifer | and then use 1.8.2 | 05:27 |
nihilifer | to be honest, i'd recommend to go with 1.8.2 - this release was tested well | 05:29 |
RA_ | thanks, I'll try that | 05:29 |
RA_ | yeah I only became aware of the docker yum repo recently | 05:29 |
*** aginwala has quit IRC | 05:30 | |
*** CheKoLyN has quit IRC | 05:32 | |
*** aginwala has joined #kolla | 05:32 | |
*** tfukushima has joined #kolla | 05:34 | |
*** jtriley has joined #kolla | 05:40 | |
RA_ | so am I right to presume that kolla liberty will only get bug fixes and not new features? | 05:41 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Ceph repo has been moved to download.ceph.org https://review.openstack.org/265822 | 06:15 |
*** Jeffrey4l has joined #kolla | 06:16 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Add py27 py34 pypy unittest https://review.openstack.org/265193 | 06:20 |
*** aginwala_ has joined #kolla | 06:26 | |
*** aginwala has quit IRC | 06:30 | |
*** aginwala_ has quit IRC | 06:32 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Add more swift base container for less redundancy https://review.openstack.org/265677 | 06:42 |
*** tpot has quit IRC | 06:45 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Share the same image between swift services https://review.openstack.org/265677 | 06:46 |
*** salv-orl_ has quit IRC | 06:50 | |
*** salv-orlando has joined #kolla | 06:50 | |
*** chandankumar has joined #kolla | 06:54 | |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Vagrant multinode deployment https://review.openstack.org/262242 | 06:55 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Documenting Ansible deployment https://review.openstack.org/263909 | 06:55 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Fetching IP addresses for vagrant-hostmanager from libvirt https://review.openstack.org/263878 | 06:55 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: kolla-mesos-ansible script https://review.openstack.org/263907 | 06:55 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add Ansible playbooks for deploying Mesos https://review.openstack.org/263906 | 06:55 |
*** dratushn_ has joined #kolla | 07:07 | |
*** dratushn_ has quit IRC | 07:09 | |
*** suro-patz has joined #kolla | 07:26 | |
*** RA_ has quit IRC | 07:43 | |
*** egonzalez90 has joined #kolla | 07:44 | |
*** iceyao has joined #kolla | 07:48 | |
*** iceyao_ has joined #kolla | 07:49 | |
*** iceyao has quit IRC | 07:53 | |
*** egonzalez90 has quit IRC | 07:58 | |
*** salv-orl_ has joined #kolla | 08:01 | |
*** salv-orlando has quit IRC | 08:05 | |
*** salv-orl_ has quit IRC | 08:06 | |
*** suro-patz has quit IRC | 08:08 | |
*** macsz has joined #kolla | 08:08 | |
*** egonzalez90 has joined #kolla | 08:11 | |
*** athomas has joined #kolla | 08:15 | |
*** shardy has joined #kolla | 08:28 | |
*** kproskurin has joined #kolla | 08:32 | |
*** egonzalez90 has quit IRC | 08:38 | |
*** akwasnie has joined #kolla | 08:45 | |
*** macsz has left #kolla | 08:46 | |
*** macsz has joined #kolla | 08:46 | |
*** jmccarthy has quit IRC | 08:48 | |
*** jmccarthy has joined #kolla | 08:50 | |
*** mbound has joined #kolla | 09:01 | |
*** gfidente has joined #kolla | 09:11 | |
*** gfidente has joined #kolla | 09:11 | |
*** openstackgerrit has quit IRC | 09:17 | |
*** openstackgerrit has joined #kolla | 09:17 | |
*** salv-orlando has joined #kolla | 09:19 | |
*** iceyao_ has quit IRC | 09:19 | |
*** cloudnautique has quit IRC | 09:22 | |
openstackgerrit | Proskurin Kirill proposed openstack/kolla-mesos: Add multiple deployment support https://review.openstack.org/262247 | 09:38 |
openstackgerrit | Proskurin Kirill proposed openstack/kolla-mesos: Start script moved from mariadb ZK node to global node https://review.openstack.org/266260 | 09:38 |
*** athomas has quit IRC | 09:41 | |
*** Allen_Gao has quit IRC | 09:42 | |
*** athomas has joined #kolla | 09:46 | |
*** tzn has joined #kolla | 09:49 | |
*** clayton has quit IRC | 09:51 | |
*** salv-orlando has quit IRC | 09:51 | |
*** clayton has joined #kolla | 09:51 | |
*** macsz has quit IRC | 09:52 | |
*** macsz has joined #kolla | 09:52 | |
*** slagle has quit IRC | 09:52 | |
*** slagle has joined #kolla | 09:54 | |
*** salv-orlando has joined #kolla | 09:57 | |
*** Allen_Gao has joined #kolla | 10:00 | |
*** salv-orlando has quit IRC | 10:10 | |
*** RA has joined #kolla | 10:10 | |
*** RA is now known as Guest69175 | 10:10 | |
*** mdnadeem has quit IRC | 10:13 | |
*** mdnadeem has joined #kolla | 10:14 | |
*** Guest69175 has quit IRC | 10:25 | |
*** salv-orlando has joined #kolla | 10:28 | |
*** salv-orlando has quit IRC | 10:28 | |
*** tfukushima has quit IRC | 10:48 | |
*** coolsvap is now known as coolsvap|away | 10:56 | |
*** mickt has joined #kolla | 10:59 | |
*** dims has joined #kolla | 11:05 | |
*** tfukushima has joined #kolla | 11:08 | |
*** tfukushima has quit IRC | 11:10 | |
*** openstackgerrit has quit IRC | 11:17 | |
*** openstackgerrit has joined #kolla | 11:17 | |
*** mickt has quit IRC | 11:23 | |
*** mickt has joined #kolla | 11:32 | |
mickt | Hi guys, I've just deployed 2x control, compute, network nodes and have enabled ha during router creation; I can't seem to achive an active router, both show standby. I can't find anything to help in logs. Anyone awrae of any gotchas withthis setup? | 11:35 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla: Cinder ubuntu binary container https://review.openstack.org/266311 | 11:43 |
*** salv-orl_ has joined #kolla | 11:49 | |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Vagrant multinode deployment https://review.openstack.org/262242 | 11:50 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Documenting Ansible deployment https://review.openstack.org/263909 | 11:50 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Fetching IP addresses for vagrant-hostmanager from libvirt https://review.openstack.org/263878 | 11:50 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: kolla-mesos-ansible script https://review.openstack.org/263907 | 11:50 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add Ansible playbooks for deploying Mesos https://review.openstack.org/263906 | 11:50 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Documenting Ansible deployment https://review.openstack.org/263909 | 11:55 |
*** Windir has quit IRC | 12:05 | |
*** mdnadeem has quit IRC | 12:12 | |
*** akwasnie has quit IRC | 12:34 | |
*** mickt_ has joined #kolla | 12:42 | |
*** salv-orl_ has quit IRC | 12:43 | |
*** mickt has quit IRC | 12:43 | |
*** akwasnie has joined #kolla | 12:45 | |
*** shakamunyi has joined #kolla | 13:04 | |
*** akwasnie has quit IRC | 13:11 | |
*** akwasnie has joined #kolla | 13:12 | |
*** dwalsh has joined #kolla | 13:19 | |
*** akwasnie has quit IRC | 13:19 | |
*** akwasnie has joined #kolla | 13:21 | |
*** rhallisey has joined #kolla | 13:27 | |
*** diogogmt has quit IRC | 13:27 | |
*** diogogmt has joined #kolla | 13:27 | |
*** dratushnyy has quit IRC | 13:40 | |
*** diogogmt has quit IRC | 13:43 | |
SamYaple | mandre: not that it would pull each image, but the process of docker checking that it has the latest image and validating it takes a while. | 13:56 |
SamYaple | mandre: in the case of a local registry it takes about as long to checkin docker as it does to pull the images (raw data) | 13:57 |
*** klint has quit IRC | 13:57 | |
SamYaple | ansible 2.0 is released | 13:58 |
*** akwasnie has quit IRC | 14:04 | |
*** macsz has quit IRC | 14:04 | |
*** akwasnie has joined #kolla | 14:07 | |
*** chandankumar has quit IRC | 14:09 | |
*** jtriley has quit IRC | 14:10 | |
*** salv-orlando has joined #kolla | 14:11 | |
*** dwalsh has quit IRC | 14:12 | |
*** akwasnie has quit IRC | 14:13 | |
*** Allen_Gao has quit IRC | 14:20 | |
*** shardy is now known as shardy_mtg | 14:27 | |
*** dwalsh has joined #kolla | 14:29 | |
*** shardy_mtg has quit IRC | 14:32 | |
*** shardy_mtg has joined #kolla | 14:34 | |
*** jtriley has joined #kolla | 14:34 | |
*** stvnoyes has quit IRC | 14:39 | |
*** stvnoyes has joined #kolla | 14:40 | |
*** salv-orlando has quit IRC | 14:42 | |
*** chandankumar has joined #kolla | 14:45 | |
*** kbyrne has quit IRC | 14:46 | |
*** salv-orlando has joined #kolla | 14:54 | |
*** salv-orlando has quit IRC | 14:56 | |
*** salv-orlando has joined #kolla | 14:56 | |
*** salv-orl_ has joined #kolla | 15:05 | |
*** salv-orlando has quit IRC | 15:05 | |
*** salv-orl_ has quit IRC | 15:11 | |
*** dave-mccowan has joined #kolla | 15:17 | |
*** dave-mcc_ has joined #kolla | 15:19 | |
*** shardy_mtg is now known as shardy | 15:19 | |
*** inc0 has joined #kolla | 15:22 | |
*** CheKoLyN has joined #kolla | 15:22 | |
*** dave-mccowan has quit IRC | 15:22 | |
clayton | I was looking through the kolla ansible playbooks for nova-compute and noticed that it does a read-only volume mount of /lib/modules. Does anyone know what makes that required? | 15:23 |
clayton | originally I assumed maybe nova-compute ran modprobe or something, but that doesn't appear to be the case | 15:23 |
SamYaple | clayton: nova-compute does run modprobe of some things in some circustance | 15:24 |
rhallisey | clayton, I believe that was just to limit what it can look it | 15:24 |
rhallisey | we do it for lots of things | 15:24 |
clayton | is it running it indirectly? I can't find any references to modprobe in the nova code | 15:24 |
SamYaple | clayton: yea im pretty sure its inderct | 15:24 |
SamYaple | like brctl loads the module and things lke that | 15:25 |
clayton | that makes sense. I was mostly just curious. | 15:25 |
SamYaple | kinda like how qemu will load the kvm module if its not loaded but you run it with --kvm | 15:25 |
*** Windir has joined #kolla | 15:27 | |
*** kbyrne has joined #kolla | 15:29 | |
*** kbyrne has quit IRC | 15:29 | |
*** kbyrne has joined #kolla | 15:30 | |
*** cemmason1 has quit IRC | 15:34 | |
*** cemmason has joined #kolla | 15:35 | |
*** cemmason1 has joined #kolla | 15:39 | |
*** cemmason has quit IRC | 15:39 | |
*** cemmason has joined #kolla | 15:42 | |
*** cemmason1 has quit IRC | 15:43 | |
*** cemmason1 has joined #kolla | 15:46 | |
*** cemmason has quit IRC | 15:46 | |
*** cemmason has joined #kolla | 15:49 | |
*** kbyrne has quit IRC | 15:50 | |
*** cemmason1 has quit IRC | 15:50 | |
*** cemmason has quit IRC | 15:59 | |
*** cemmason1 has joined #kolla | 15:59 | |
*** thomasem-away is now known as thomasem | 16:03 | |
*** tzn has quit IRC | 16:04 | |
*** suro-patz has joined #kolla | 16:06 | |
*** tzn has joined #kolla | 16:07 | |
*** kbyrne has joined #kolla | 16:09 | |
*** cloudnautique has joined #kolla | 16:16 | |
*** cloudnautique has joined #kolla | 16:17 | |
*** cloudnautique has quit IRC | 16:18 | |
*** cloudnautique has joined #kolla | 16:20 | |
*** diogogmt has joined #kolla | 16:20 | |
*** egonzalez has joined #kolla | 16:22 | |
*** diogogmt has quit IRC | 16:27 | |
*** athomas has quit IRC | 16:28 | |
*** diogogmt has joined #kolla | 16:29 | |
*** cemmason has joined #kolla | 16:31 | |
*** cemmason1 has quit IRC | 16:31 | |
*** athomas has joined #kolla | 16:33 | |
*** cemmason1 has joined #kolla | 16:36 | |
*** cemmason has quit IRC | 16:38 | |
*** cemmason1 has quit IRC | 16:41 | |
*** cemmason has joined #kolla | 16:41 | |
*** cemmason1 has joined #kolla | 16:45 | |
*** egonzalez has quit IRC | 16:45 | |
*** cemmason has quit IRC | 16:45 | |
*** chandankumar has quit IRC | 16:49 | |
*** cemmason1 has quit IRC | 16:49 | |
*** cemmason has joined #kolla | 16:49 | |
*** asalkeld has joined #kolla | 16:51 | |
asalkeld | o/ | 16:51 |
openstackgerrit | Steve Noyes proposed openstack/kolla-mesos: Script to show status of kolla-mesos tasks https://review.openstack.org/266010 | 16:52 |
asalkeld | hi stvnoyes | 16:53 |
*** cemmason1 has joined #kolla | 16:53 | |
stvnoyes | hi angus | 16:53 |
openstackgerrit | Proskurin Kirill proposed openstack/kolla-mesos: Support deploy without configs https://review.openstack.org/266476 | 16:53 |
openstackgerrit | Proskurin Kirill proposed openstack/kolla-mesos: Add memcached support https://review.openstack.org/266477 | 16:53 |
*** cemmason has quit IRC | 16:53 | |
asalkeld | checking out your patch now.. | 16:53 |
stvnoyes | ok, thanks | 16:55 |
*** bradjones has quit IRC | 16:55 | |
*** suro-patz has quit IRC | 16:57 | |
*** bradjones has joined #kolla | 16:58 | |
*** bradjones has quit IRC | 16:58 | |
*** bradjones has joined #kolla | 16:58 | |
stvnoyes | angus, hold off for a minute just found a pep8 issue | 16:59 |
stvnoyes | (alphabetic ordering of imporrts) | 16:59 |
asalkeld | stvnoyes: also that should be in the cmd/ module | 16:59 |
stvnoyes | sure I can move it there. | 17:00 |
asalkeld | at some point soon we need a python API to this stuff | 17:00 |
asalkeld | so the cli and other upper layers can talk to kolla-mesos | 17:00 |
*** Jeffrey4l has quit IRC | 17:02 | |
*** salv-orlando has joined #kolla | 17:03 | |
asalkeld | kproskurin: reviewed https://review.openstack.org/#/c/266476/1 | 17:03 |
openstackgerrit | Steve Noyes proposed openstack/kolla-mesos: Script to show status of kolla-mesos tasks https://review.openstack.org/266010 | 17:05 |
kproskurin | asalkeld: So we have to gen configs at the time of cmd runs only? | 17:05 |
kproskurin | Not sure if it’s optimal | 17:05 |
asalkeld | kproskurin: it just won't work otherwise | 17:06 |
kproskurin | asalkeld: Thats sad. I’ll think about it. Ty | 17:07 |
asalkeld | the requires is there to so "i need something", that means a service running/something done. that influences the hostvars and thus how the rendering is done | 17:08 |
asalkeld | if we do the render too early it could well be incorrect | 17:08 |
kproskurin | Well right now I just cant see the exampe of any race. All hostvars what we using pretty static from the start | 17:13 |
asalkeld | kproskurin: you put one in: horizon which depends on memcached | 17:20 |
asalkeld | https://github.com/openstack/kolla-mesos/blob/master/config/horizon/templates/local_settings.j2#L129 | 17:20 |
kproskurin | But IP of memcached is knows at the start of the deploy, so we just put it in the configs | 17:20 |
kproskurin | known* | 17:21 |
asalkeld | but it's not | 17:21 |
asalkeld | everything is started together | 17:21 |
asalkeld | it's timing dependant | 17:21 |
kproskurin | oh I see | 17:21 |
kproskurin | it really can be anythere | 17:21 |
asalkeld | so it might well work sometimes | 17:21 |
asalkeld | and then just break | 17:21 |
* asalkeld heads to another meeting... | 17:22 | |
asalkeld | doh, i am apparently cloning myself and going to 2 meetings :-O | 17:23 |
*** sdake has joined #kolla | 17:25 | |
sdake | morning | 17:25 |
inc0 | hey, guys, please review https://review.openstack.org/#/c/257568/ | 17:25 |
inc0 | it's sitting there for few weeks now and that's one important patchset | 17:25 |
openstackgerrit | Merged openstack/kolla: Added Mistral doc link in README.rst https://review.openstack.org/266105 | 17:32 |
*** kproskurin has quit IRC | 17:34 | |
openstackgerrit | Merged openstack/kolla: Fix some mistal mistakes https://review.openstack.org/264846 | 17:35 |
openstackgerrit | Merged openstack/kolla: Fix missing jinja2 in source images for mesos https://review.openstack.org/265882 | 17:36 |
openstackgerrit | Merged openstack/kolla-mesos: Specify tests directory for oslo_debug_helper https://review.openstack.org/265172 | 17:36 |
*** mbound has quit IRC | 17:41 | |
*** cloudnautique has quit IRC | 17:42 | |
*** cloudnautique has joined #kolla | 17:43 | |
*** cloudnautique has quit IRC | 17:43 | |
*** cloudnautique has joined #kolla | 17:43 | |
*** cloudnautique has quit IRC | 17:44 | |
*** unicell has quit IRC | 17:44 | |
*** cloudnautique has joined #kolla | 17:45 | |
sdake | finally some email threads for kolla !! | 17:48 |
sdake | good job folks! | 17:48 |
sdake | it only took a year :) | 17:48 |
*** cemmason has joined #kolla | 17:48 | |
*** cemmason1 has quit IRC | 17:50 | |
*** athomas has quit IRC | 17:51 | |
*** cemmason has quit IRC | 17:53 | |
*** cemmason has joined #kolla | 17:53 | |
*** cemmason has quit IRC | 17:56 | |
*** cemmason has joined #kolla | 17:56 | |
*** jasonsb has quit IRC | 17:56 | |
sdake | inc0 did alicja move to the us or still in poland(?) | 17:59 |
*** athomas has joined #kolla | 17:59 | |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Vagrant multinode deployment https://review.openstack.org/262242 | 18:00 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Documenting Ansible deployment https://review.openstack.org/263909 | 18:00 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Fetching IP addresses for vagrant-hostmanager from libvirt https://review.openstack.org/263878 | 18:00 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: kolla-mesos-ansible script https://review.openstack.org/263907 | 18:00 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add Ansible playbooks for deploying Mesos https://review.openstack.org/263906 | 18:00 |
nihilifer | sdake: you say that we're becoming more agile by sending mails? ;) | 18:00 |
sdake | email slows things down | 18:01 |
sdake | which is good on big decision points | 18:01 |
*** cemmason1 has joined #kolla | 18:01 | |
*** cemmason has quit IRC | 18:01 | |
sdake | i love our irc culture | 18:01 |
sdake | its some of the best i've seen in any openstack project i've been involved with | 18:01 |
sdake | and i definately want to keep our irc culture rolling as is :) | 18:02 |
*** diogogmt has quit IRC | 18:02 | |
sdake | nihilifer is there docs for mesos | 18:02 |
*** macsz has joined #kolla | 18:02 | |
sdake | for the kolla mesos i mean | 18:02 |
nihilifer | sdake: so far we have this http://docs.openstack.org/developer/kolla-mesos/ | 18:03 |
nihilifer | the changes about vagrant which are on review bring more stuff about building environment | 18:03 |
sdake | nice let me read all of it | 18:03 |
sdake | can we just get all the docs in the kolla repo? | 18:03 |
sdake | or did that get rejected during reviews | 18:04 |
sdake | have to keep in mind i was out most of december with medical problems related to my tooth | 18:04 |
*** dave-mcc_ has quit IRC | 18:04 | |
nihilifer | umm... we thought that docs specific for mesos should go to kolla-mesos | 18:04 |
*** cemmason has joined #kolla | 18:04 | |
bmace | i tend to agree. odd having docs which describe doing something that you need to pull a separate repo to accomplish. | 18:05 |
*** cemmason1 has quit IRC | 18:05 | |
nihilifer | yep, i think that no one disagreed on this when asalkeld was introducing docs | 18:06 |
*** diogogmt has joined #kolla | 18:07 | |
SamYaple | docs for mesos in mesos +1 | 18:11 |
*** gfidente is now known as gfidente|afk | 18:11 | |
SamYaple | nihilifer: i got thinking about our little mesos/ansible code sharing conundrum | 18:12 |
SamYaple | we should break out ansible into kolla-ansible which deps on kolla (for kolla-ansible deploy tools) and you could make kolla-mesos dep on kolla-ansible in a way you only reuse some of the ansible parts (like the libs). the kolla-ansible repo would only be a few MB. | 18:13 |
SamYaple | idk thats what i got thinking about | 18:13 |
SamYaple | inc0: https://review.openstack.org/#/c/266105/ docs changes don't require TrivalFix tags or bugs under any circumstances | 18:17 |
*** unicell has joined #kolla | 18:17 | |
*** suro-patz has joined #kolla | 18:19 | |
sdake | samyaple no they dont but don't reject if they have them | 18:23 |
sdake | it causes no harm | 18:23 |
sdake | i usuallyl put trivialfix on doc fixes | 18:23 |
sdake | ok re mesos docs, i guess its fine to keep them separate, but a cross link between the docs would be nice | 18:24 |
sdake | its just that we hve a bunch of non-specific to the deployment method docs | 18:24 |
sdake | and ifsomeone looks at the mesos docs they will miss those common documents | 18:24 |
SamYaple | sdake: its about rejecting them if they _dont_ have them, not rejecting them if they do (look at link) | 18:26 |
sdake | oh i see | 18:27 |
SamYaple | sdake: on the subject of mesos docs, what it really sounds like you are saying is we need kolla docs, kolla-ansible docs, and kolla-mesos docs | 18:27 |
SamYaple | and we should start the process of breaking kolla-ansible out | 18:27 |
sdake | i want to make susre the timing is right on that | 18:28 |
rhallisey | SamYaple, might be next cycle we break out. Still need to tackle upgrades & big tent | 18:28 |
sdake | i dont want to do anything to impede upgrades | 18:28 |
sdake | if there is nothign else we do this cycle, we must absolutely have upgrades of atleast openstack services | 18:28 |
SamYaple | rhallisey: dont give him excuses he will latch on! | 18:28 |
rhallisey | lol | 18:28 |
sdake | i was already latched prior to rhallisey saying that | 18:28 |
SamYaple | rhallisey: but i agree on the upgrade bit, not the big tent | 18:29 |
sdake | listen guys, if we don't deliver upgrades, it will be a big blow to our project | 18:29 |
rhallisey | upgrade is def #1 | 18:29 |
SamYaple | i know sdake, but we do need to break out before it gets harder | 18:29 |
sdake | ya who cares about velocity of big tent implementation | 18:29 |
sdake | i mean i care sort of | 18:29 |
SamYaple | i dont htink it will impact upgrades, but hey no one seems to care bout that since the review has been sitting up there for months | 18:29 |
jpeeler | sdake: speaking of cycles, how about those midcycle hotel recommendations? | 18:29 |
sdake | but compared to upgrades | 18:29 |
rhallisey | I thought we did | 18:29 |
rhallisey | lol | 18:29 |
SamYaple | weeks* | 18:29 |
sdake | jpeeler had contractors over all day yesterday asking t o borrow tools and asking for this and that | 18:29 |
sdake | so basically i got shit done | 18:29 |
jpeeler | sdake: just a friendly reminder | 18:30 |
sdake | jpeeler i'll be doing that today | 18:30 |
rhallisey | SamYaple, what's there to do in Spartanburg | 18:30 |
SamYaple | rhallisey: drink | 18:30 |
rhallisey | perfect | 18:30 |
SamYaple | i think you want to ask about greenville though | 18:30 |
rhallisey | yes rather greenville | 18:31 |
SamYaple | rhallisey: drink | 18:31 |
rhallisey | yes | 18:31 |
SamYaple | :P there is actually a nice river through the center of town | 18:31 |
SamYaple | its quite nice | 18:31 |
rhallisey | ya I've been there a bunch | 18:31 |
rhallisey | been to the BWM factory | 18:31 |
sdake | sam is going to recommend 3 places to eat for dinner and we will vote first day on where we are going | 18:31 |
SamYaple | yea i live right near there | 18:31 |
rhallisey | actually that's spartanburg | 18:31 |
sdake | and i'll tie break ;) | 18:32 |
SamYaple | i think its technically spartanburg rhallisey, but its closer to greenvile | 18:32 |
SamYaple | basically its dead center between the two | 18:32 |
rhallisey | it's nice. I like SC | 18:32 |
SamYaple | Yea. i miss mytle beach though. thats where i hail from | 18:33 |
SamYaple | been in the upstate long enough though | 18:33 |
sdake | SamYaple you can have split repositories when I have upgrades of all openstack services (not infrastructure) - deal? :) | 18:33 |
*** tzn has quit IRC | 18:34 | |
rhallisey | wow myrtle beach is amazing | 18:34 |
rhallisey | I like Charleston & Myrtle beach in SC | 18:34 |
SamYaple | sdake: only a deal if you plan on reviewing them | 18:34 |
SamYaple | otherwise well never get anywhere | 18:35 |
sdake | just a note for folks booking hotels for greenville, fly to the greenville airport, not anywhere else | 18:35 |
SamYaple | (we could have upgrades done if we had reviews) | 18:35 |
sdake | yes i am going to start picking up my reviews slack | 18:35 |
sdake | folks I need to let everyone know as well | 18:35 |
sdake | I have a house remodel underway | 18:35 |
SamYaple | not just you i mean in general | 18:35 |
SamYaple | people are afraid of the upgrades i think | 18:35 |
sdake | I am putting tile down wherever there is carpet, and turning a bedroom into a walk in closet | 18:35 |
SamYaple | translation: moneymoneymoneymoneymoneymoneymoneymoney | 18:36 |
sdake | as a result, I may be a little laggy on some days for the enxt 2-3 weeks | 18:36 |
sdake | on other days when nothing is happening i'll be here from 8am-10pm | 18:36 |
sdake | so my kids download rate and upload rate is 950mbit/sec and 850mbit/sec | 18:37 |
sdake | after i had the contractors over yesterday | 18:37 |
sdake | to hardwire cat6 | 18:37 |
sdake | i think I can run 10gig on my cat6 as well but not sure | 18:38 |
sdake | speedtest.net results | 18:38 |
sdake | my wireless 5g on macbook 15" laptop is 350mb down 250mb up | 18:38 |
*** jasonsb has joined #kolla | 18:39 | |
inc0 | I think I'll visit mountains in SC | 18:41 |
sdake | anyone with a mac that uses outlook know how to turn on indented quoting? | 18:41 |
inc0 | mountains looks nice there | 18:41 |
openstackgerrit | Merged openstack/kolla-mesos: Add Ansible playbooks for deploying Mesos https://review.openstack.org/263906 | 18:41 |
sdake | i figured it out once but can't find the settings or web page | 18:42 |
sdake | for my newer laptop | 18:42 |
SamYaple | inc0: I will go with you! it will be cold | 18:42 |
inc0 | I like cold | 18:42 |
SamYaple | inc0: i love hiking, there are some decent places to hike here | 18:42 |
rhallisey | inc0, count me in :D | 18:42 |
inc0 | I don't have my winter clothes tho, I might have to buy something | 18:42 |
rhallisey | I'm an outdoor person too | 18:43 |
SamYaple | anyone been to table rock? | 18:43 |
SamYaple | its a decent trail with a nice view at the end | 18:43 |
rhallisey | nope | 18:43 |
inc0 | so, how about we stay there few more days and go hiking together? I really need mountains | 18:43 |
SamYaple | inc0: I would offer my house up, but I *may* not have it yet | 18:43 |
SamYaple | i should be closing 2 days before midcycle | 18:44 |
SamYaple | but who knows | 18:44 |
inc0 | airbnb is ok | 18:44 |
sdake | inc0 i can't afford the hotel stay | 18:44 |
sdake | and can't exepnse airbnb | 18:44 |
inc0 | I was thinking expansing hotel but after midcycle moving to airbnb | 18:44 |
SamYaple | should my house close on time, I will have 2 extra rooms for people to stay extra should they want too | 18:46 |
*** cemmason has quit IRC | 18:58 | |
*** cemmason1 has joined #kolla | 18:58 | |
*** tzn has joined #kolla | 18:59 | |
*** bmace has quit IRC | 19:01 | |
*** bmace has joined #kolla | 19:01 | |
*** chandankumar has joined #kolla | 19:04 | |
*** dwalsh has quit IRC | 19:04 | |
*** SourabhP has joined #kolla | 19:05 | |
*** cemmason1 has quit IRC | 19:05 | |
*** cemmason has joined #kolla | 19:05 | |
*** aginwala has joined #kolla | 19:06 | |
*** cemmason1 has joined #kolla | 19:09 | |
*** cemmason has quit IRC | 19:10 | |
*** athomas has quit IRC | 19:11 | |
*** cemmason1 has quit IRC | 19:12 | |
*** cemmason has joined #kolla | 19:12 | |
*** cemmason1 has joined #kolla | 19:16 | |
*** cemmason has quit IRC | 19:16 | |
*** aginwala has quit IRC | 19:17 | |
*** salv-orlando has quit IRC | 19:17 | |
*** salv-orlando has joined #kolla | 19:18 | |
*** athomas has joined #kolla | 19:19 | |
*** dwalsh has joined #kolla | 19:19 | |
*** cemmason1 has quit IRC | 19:20 | |
*** cemmason has joined #kolla | 19:20 | |
*** aginwala has joined #kolla | 19:23 | |
*** cemmason has quit IRC | 19:24 | |
*** cemmason1 has joined #kolla | 19:25 | |
sdake | nihilifer will eric make the kolla midcycle meeting? | 19:26 |
nihilifer | sdake: dunno | 19:28 |
nihilifer | hmm... he's not here. i'll ask him tommorow | 19:29 |
sdake | if we plan to do this heka conversion in mitaka, it really really would be helpful to have him at hte midcycle | 19:29 |
sdake | I know everyones budgets are tight | 19:29 |
sdake | if we wait to N then it probably doesn't matter | 19:30 |
sdake | here is what I don't want to happen | 19:30 |
sdake | I don't want rsyslog taken out and heka put in, and then nothing working | 19:30 |
sdake | understand? | 19:30 |
sdake | nihilifer ^^ | 19:30 |
*** suro-patz has quit IRC | 19:30 | |
sdake | SamYaple ping | 19:31 |
*** cemmason1 has quit IRC | 19:31 | |
*** cemmason has joined #kolla | 19:31 | |
sdake | a cisco QA engineer has approached me wanting to contribute to the gating work | 19:31 |
nihilifer | sdake: ok | 19:31 |
* nihilifer have to educate himself about all these monitoring/logging/whatever technologies | 19:32 | |
sdake | SamYaple can you give me a quick status of where we are today and where you recommend we go next? | 19:32 |
sdake | nihilifer your petty much our poitn man for our mirantis/fuel integration | 19:32 |
sdake | so unfortunately yo uwill have to learn all these fuel techs that want ot come into kolla | 19:33 |
sdake | because I trust you since you are aware of the code base and by your solid work | 19:33 |
sdake | but I don't know alot of these other fuel folks, and its not that I don't trust them, its just that I dont know if they understand enough of kolla to see how it would fit together and integrate | 19:34 |
sdake | nihilifer understand? | 19:34 |
nihilifer | yes sir ;) | 19:34 |
*** cemmason has quit IRC | 19:35 | |
*** cemmason has joined #kolla | 19:35 | |
SamYaple | sdake: as in, where is kolla nd where do we want to be? | 19:35 |
*** dave-mccowan has joined #kolla | 19:36 | |
SamYaple | hey our stacklytic stats finnaly updated! | 19:37 |
SamYaple | its been 2 weeks | 19:38 |
*** dwalsh has quit IRC | 19:39 | |
nihilifer | SamYaple: i also was thinking aboyt how we can share commons between kolla subprojects. i also see a problem with vagrant where we're duplicating stuff | 19:40 |
nihilifer | and it seems that i had to fix things that also occur in main kolla repo, but weren't discover | 19:41 |
nihilifer | i.e. dhcp leases | 19:41 |
*** chandankumar has quit IRC | 19:41 | |
*** jasonsb has quit IRC | 19:42 | |
nihilifer | the main question here is - do we can install ansible modules as librries somehow? | 19:42 |
SamYaple | nihilifer: kinda | 19:43 |
SamYaple | you can specify a library path with ansible at run time | 19:43 |
nihilifer | the only thing that comes to my mind is to have a dummy pip package storing ansible modules somwhere in /usr/share nad then all ansible runners (from kolla-ansible, kolla-mesos, kolla-anythingnewandmorehipster) pointing to them as their library | 19:43 |
SamYaple | or before hand | 19:43 |
SamYaple | so it will pck up those modules | 19:43 |
*** SourabhP has quit IRC | 19:43 | |
sdake | SamYaple sorroy was picking up rugrats from school | 19:43 |
sdake | so yes to your question | 19:43 |
nihilifer | SamYaple: ok, so seems that your idea is similar or even the same | 19:44 |
SamYaple | in this case if we had kolla-ansible installed then you could do (i forget the option exactly) ansble-playbook --path-to-library /usr/local/share/kolla/ansible/library/ | 19:45 |
nihilifer | maybe for vagrant there is some way to use common ruby library | 19:45 |
sdake | nihilifer fwiw I wont be entertaining splitting the repos until upgrades are done | 19:45 |
sdake | but this is still a discussion we can hav eat the midcycle | 19:45 |
SamYaple | sdake: this has been ongoing and will continue being ongoing | 19:45 |
nihilifer | sdake: ok | 19:45 |
sdake | SamYaple I mean yes to your question about the gating | 19:46 |
sdake | I can put this guy to work on gating where should he be working | 19:46 |
nihilifer | inc0: btw, sorry for giving always only +1. i'll try to play with your upgrade playbooks on multinode tommorow | 19:46 |
inc0 | nihilifer, no worries, I do want more eyes on it | 19:46 |
inc0 | it's just quite some time this review hangs there | 19:47 |
SamYaple | sdake: the gating is stalling due to no RAM in the gate. we need to reduce the usage of RAM by the different applications. thats config options more than anything | 19:47 |
inc0 | and upgrades is one of our priorities | 19:47 |
sdake | #1 priority | 19:47 |
SamYaple | sdake: with reduction of RAM I can finish out the multinode gate (which is almost done) | 19:47 |
inc0 | what sdake said;) | 19:47 |
sdake | SamYaple so should this guy sort out how to get the ram usage down on the single node gate? | 19:48 |
sdake | i want actual functional testing of the deployment | 19:49 |
sdake | pbourke was working on that, did that work stall? | 19:49 |
SamYaple | sdake: we are no where near functional testing | 19:49 |
SamYaple | we need to use a whole lot less ram | 19:49 |
*** jasonsb has joined #kolla | 19:49 | |
SamYaple | or drop services (like heat is going first) | 19:49 |
SamYaple | getting rid of heat, horizon, cinder was enough to get a samll cirros image going | 19:50 |
sdake | had to get rid of 3 services huh | 19:50 |
* sdake ughs | 19:50 | |
SamYaple | im not sure teh config options to start reducing ram usage | 19:50 |
*** SourabhP has joined #kolla | 19:50 | |
SamYaple | i can limit ram usage per docker container, but thats not super helpful | 19:50 |
SamYaple | right now our gate leaves us with about 100MB on a full deploy | 19:51 |
SamYaple | it times out alot of times because of actually OOM'ing | 19:51 |
SamYaple | additionally, new gates (bluebox ovh) dont have a swap disk | 19:51 |
sdake | and we can't use swap? | 19:51 |
* sdake groans | 19:52 | |
SamYaple | so we dont have a seperate disk for docker or swap, which means loopback and swapfile | 19:52 |
SamYaple | last time i tried the the gate legit timed out from being to slow | 19:52 |
sdake | lets say you had a qe engineer 100% available to work on kolla, what would you have them do? | 19:52 |
sdake | sort out how to reduce ram usage in the gates? | 19:53 |
SamYaple | reduce ram usage of serivces would be first thing | 19:53 |
SamYaple | so that may be simple, or complex | 19:53 |
*** dwalsh has joined #kolla | 19:53 | |
SamYaple | it may require dropping workers to 1 for each service or more complex, i dont know | 19:53 |
SamYaple | havent had a chance to work on it | 19:53 |
sdake | ya your not superman - we have a dude that wants to work on this stuff i think more or less full time | 19:54 |
SamYaple | once thats done I (or he) need to get the swap/nodisk sitaution figured out | 19:54 |
SamYaple | unify so we never fail on non-rax and hp gates | 19:54 |
sdake | so if I can get him integrated into the project it would be helpful :) | 19:54 |
SamYaple | finally the multinode is waiting on both of those things above | 19:54 |
SamYaple | i have the experimental gate in place and working, but i run out of RAM | 19:54 |
sdake | SamYaple fantastic information, I'll try to get him to work on it then | 19:55 |
sdake | I think one easy option is to reduce the number of processes per service | 19:55 |
sdake | for example the api servers run like 8 api services | 19:55 |
SamYaple | yes, and tahts easy, but it doesnt save _that_ much ram | 19:55 |
sdake | we could drop that down to 1 | 19:55 |
SamYaple | they default to one thread per core | 19:55 |
sdake | and I guess we are miles away from gating ceph | 19:56 |
SamYaple | with reach project-config/nodepool changes even further | 19:56 |
sdake | parsecs ;:) | 19:56 |
SamYaple | hp is going away and bluebox and the ovh gates dont have a seperate disk | 19:57 |
SamYaple | not everyone uses lvm | 19:57 |
SamYaple | the only common point is.... file based ugh | 19:57 |
*** aginwala has quit IRC | 19:57 | |
sdake | ceph file basad would wfm | 19:57 |
SamYaple | too slow. gate would time out | 19:57 |
sdake | I dont know if that would actualy work though | 19:57 |
SamYaple | stated above | 19:57 |
SamYaple | we may be able to do somethign fancy and prevent cloud-init from running and expanding to the full disk size. then we would be good | 19:58 |
sdake | i htought that timeout stuff ws about the gating using swap | 19:58 |
SamYaple | using a swapfile | 19:58 |
sdake | which cat wants to co-present with me at summit in austin on kolla | 19:59 |
sdake | SamYaple are you doing a lab presentation? | 19:59 |
SamYaple | yea | 20:00 |
sdake | if your willing pleaase add me as a copresenter to that, it might help your chances of acceptance | 20:01 |
SamYaple | we have already talked about this sdake | 20:01 |
sdake | i'll let you run the show | 20:01 |
sdake | right | 20:01 |
sdake | just ocnfirming | 20:01 |
SamYaple | you can be on that, i need help | 20:01 |
SamYaple | yup | 20:02 |
inc0 | sdake, what are you going to present? | 20:02 |
SamYaple | labs are hard for single person | 20:02 |
inc0 | I'll do upgrades talk with Sam | 20:02 |
sdake | inc0 here is the deal | 20:02 |
SamYaple | 2 or 3 is useful | 20:02 |
*** aginwala has joined #kolla | 20:02 | |
sdake | we get 1 talk for kolla at summit if we get an ytalk at all | 20:02 |
inc0 | you think it's that bad | 20:02 |
inc0 | ? | 20:02 |
sdake | we get one lab possibly as well | 20:02 |
sdake | i think what makes sense is sa a community we decide what talk we want to present | 20:03 |
sdake | and prepare as a team, but have 1-2 presenters | 20:04 |
sdake | upgrades seems like a fantastic topic | 20:04 |
inc0 | if you're ok with my heavy slavic accent | 20:04 |
sdake | typicallly the way it works is 1 person presents, and 1 person demos | 20:05 |
inc0 | so what we did in Vancouver is we splitted content between us | 20:05 |
inc0 | and it worked pretty well | 20:05 |
sdake | dont understand | 20:05 |
rhallisey | inc0, I thought you changed to a thick southern drawl | 20:05 |
inc0 | nah, I'm keeping slavic accent | 20:05 |
inc0 | people assume I'm russian and become afraid of mafia | 20:06 |
rhallisey | lol | 20:06 |
sdake | lol | 20:06 |
sdake | it also helps our chances of being accepted if the speaker is a PTL | 20:06 |
sdake | also heps our chances if the talk includes a demo recorded or live | 20:06 |
inc0 | I wanted live really | 20:07 |
inc0 | adrenaline is cool | 20:07 |
sdake | yes which means we need a pre-deployed liberty since upgrades take 20 minutes | 20:08 |
sdake | another possible topic is mesos, although it is unclear to me if that work will be in demoable shape | 20:08 |
sdake | nihilifer any thoughts ? | 20:08 |
sdake | I think we get the most bang for our respective buck with the upgrade talk | 20:08 |
sdake | SamYaple hae you found a 3rd victim for the lab? | 20:09 |
inc0 | so sdake SamYaple and me will do upgrades talk? | 20:10 |
inc0 | 3 of us? | 20:10 |
sdake | inc0 i think it makes more sense for me to give the presentation since sselection committtees prefer ptls over nonptls | 20:11 |
sdake | but we could do 3 | 20:11 |
*** gfidente|afk has quit IRC | 20:11 | |
sdake | i know we could make that work | 20:11 |
*** aginwala has quit IRC | 20:12 | |
inc0 | well I'd like to be there since it will be my and Sams work mostly (probably) | 20:12 |
sdake | the foundation doesn't like when people submit a bunch of speakers on a talk unless its a panel | 20:12 |
sdake | inc0 wfm | 20:12 |
sdake | ok how aobut this, i'll give the main preso, one of you can show the liberty environment as is, and one of you can show the upgrade process | 20:13 |
SamYaple | we are doing a 3 person upgrade talk? | 20:13 |
SamYaple | that seems like too much | 20:13 |
sdake | i think it would flow pretty well | 20:14 |
sdake | if you guys want to fly without me and take the risk, we could do that | 20:15 |
*** aginwala has joined #kolla | 20:15 | |
sdake | but trust me when i say having a ptl presesnt is hugely beneficial | 20:15 |
inc0 | I have nothing against you in presentation, since we'll get one shot at this we don't want to have competition in kolla presentations imho | 20:16 |
sdake | upgrades are the missing link in kolla, that is what we need to present | 20:16 |
SamYaple | i was saying ill just bow out. If my lab gets accepted tahts what I want most. me and inc0 have been doing upgrades is all so we have teh upgrade info | 20:19 |
sdake | SamYaple sounds good bro | 20:20 |
sdake | I relaly really really want a lab | 20:20 |
sdake | pelase dear god get a lab | 20:20 |
sdake | lets get good coverage too, atleast 3-4 people to run it | 20:20 |
sdake | rather assit you in running it | 20:21 |
sdake | a lab would be huge | 20:21 |
SamYaple | yes thats teh thing. lab with 3+ people == perfect. talk with 3 people is... well threes a crowd | 20:21 |
sdake | inc0 competittions in kolla presentations are ok, there is just limited spots inthe container track | 20:21 |
SamYaple | i think he was talking in-fighting over who is doign the presentation | 20:22 |
sdake | and kolla is sitll a hot topic in the foundation so we get a little bit of love from the selection committtee, but not 20% of the slot spots love :) | 20:22 |
sdake | SamYaple was that a misfire? | 20:22 |
inc0 | so, let's try to get both | 20:22 |
sdake | well lab is a different track from the container track | 20:23 |
sdake | so there is a completel ydifferent set of people that pick those talks | 20:23 |
sdake | so inc0, you be willing to copresent with me kolla and execute the role sam handled last time but with an upgrade bent? | 20:24 |
inc0 | so me+sdake will propose upgrades talk and Sam+sdake+someone else from core team for a lab? | 20:24 |
inc0 | yeah, why nopt | 20:24 |
sdake | inc0 your welcome to help with the lab too :) | 20:24 |
sdake | SamYaple will prepare it and hopefully get us all setup | 20:24 |
sdake | and we need 1 more person, possibly pbourke? | 20:25 |
SamYaple | well after last time im not waiting on you to prep it :P | 20:25 |
SamYaple | yea pbourke_ is on the lab with me | 20:25 |
sdake | dude our talk was a 9 out of 10 | 20:25 |
sdake | and i wrote it in 2 hoours :) | 20:25 |
SamYaple | thats a B! | 20:25 |
*** cemmason1 has joined #kolla | 20:25 | |
sdake | considering the deployment foxtrotted upand still worked, I was amazed :) | 20:26 |
*** aginwala has quit IRC | 20:26 | |
*** cemmason has quit IRC | 20:26 | |
*** aginwala has joined #kolla | 20:26 | |
sdake | inc0 one more advantage of co-present iwth me is I have 10gige hardware ready to rock as well all high end core i7 with 32gb ram | 20:27 |
sdake | 3 node HA | 20:27 |
*** cemmason has joined #kolla | 20:29 | |
SamYaple | (my servers are better) | 20:29 |
*** cemmason1 has quit IRC | 20:29 | |
sdake | possible | 20:30 |
sdake | so did folks see what ironic/neutron are doing to manage their release process? | 20:30 |
sdake | they are dropping blueprints and milestones | 20:31 |
sdake | and relying entirely on specs | 20:31 |
*** macsz has quit IRC | 20:31 | |
*** _tzn has joined #kolla | 20:31 | |
*** macsz has joined #kolla | 20:31 | |
*** cemmason has quit IRC | 20:32 | |
SamYaple | milestones i would like dropped to be honest | 20:32 |
SamYaple | but i like specs | 20:32 |
SamYaple | i like blueprints too | 20:32 |
sdake | i really llike milestones | 20:32 |
*** cemmason has joined #kolla | 20:32 | |
SamYaple | i know you do | 20:32 |
sdake | it gives a sense of timing | 20:32 |
SamYaple | maybe one day well hit one ;) | 20:32 |
sdake | haha | 20:33 |
SamYaple | :D | 20:33 |
sdake | we have hit a couple actually | 20:33 |
sdake | but this lat one was a fail | 20:33 |
sdake | the ones we hit were earlier in the lifecycle of kolla | 20:33 |
*** suro-patz has joined #kolla | 20:33 | |
sdake | as things get more complex its harder to hit our milestones because our gates are not ideal | 20:33 |
*** tzn has quit IRC | 20:34 | |
sdake | if our gates were 100% trustworthy our deadlines would be 100% on target | 20:37 |
SamYaple | oh come on thats a cop out. its not the gates its the you werent here for a month past the deadline | 20:37 |
*** salv-orlando has quit IRC | 20:39 | |
*** salv-orlando has joined #kolla | 20:40 | |
*** aginwala has quit IRC | 20:41 | |
*** macsz has quit IRC | 20:47 | |
*** macsz has joined #kolla | 20:49 | |
stvnoyes | hi asalkeld | 20:53 |
asalkeld | hi stvnoyes | 20:53 |
asalkeld | you ready for a review? | 20:53 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Add some additional debug to gate https://review.openstack.org/266568 | 20:53 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Test file-backend for gate https://review.openstack.org/266569 | 20:53 |
stvnoyes | i have a question about keystone. it's not getting thru setup | 20:53 |
asalkeld | ok | 20:53 |
SamYaple | stvnoyes: on mesos? | 20:53 |
stvnoyes | in extend_start it gets thru the kill httpd and then dies | 20:53 |
asalkeld | have a look on the mesos UI and look at stderr | 20:54 |
stvnoyes | when i comment out the kill it makes it thru extend_start | 20:54 |
stvnoyes | no error in stderr of keystone container | 20:54 |
asalkeld | stvnoyes: i submitted a patch for that | 20:55 |
asalkeld | https://github.com/openstack/kolla-mesos/blob/master/config/keystone/keystone_config.yml.j2#L33 | 20:55 |
asalkeld | it wasn't using bash | 20:55 |
*** shardy has quit IRC | 20:55 | |
asalkeld | can you make sure you have "bash kolla_mesos_extend_start" | 20:55 |
stvnoyes | hmmm ii thought i was up to date. let me check. thx | 20:56 |
asalkeld | (that's the most obvious thing i can think of) | 20:56 |
*** akwasnie has joined #kolla | 20:56 | |
stvnoyes | ok i have that change. i can see the echo messages in the logs all the way down to "status 0" after the kill. then it goes quiet. | 20:57 |
asalkeld | hold on, let me look | 20:59 |
stvnoyes | do you know why httpd is being killed at the end of extend_start? I couldn't see anything like that in the kolla code (though I could have missed it) | 20:59 |
asalkeld | stvnoyes: so the difference is kolla runs that container twice | 20:59 |
asalkeld | we run it once | 20:59 |
asalkeld | so it runs apache once just to configure stuff | 21:00 |
asalkeld | then we start it "properly" | 21:00 |
stvnoyes | ok that makes sense | 21:01 |
*** macsz has quit IRC | 21:01 | |
asalkeld | stvnoyes: so nothing interesting in the stderr/stdout for that container? | 21:02 |
stvnoyes | the logs i have right now are from the session where I commented out the httpd kill. I'll re-run. it'll take a few minutes. | 21:03 |
asalkeld | sure, thanks - can you then post to paste.openstack.org | 21:04 |
asalkeld | brb: getting coffee | 21:04 |
SamYaple | so asalkeld stvnoyes last month we moved account creation out of the keystone container like that. so you _should_ need to do that anymore. that said i recognize kolla-mesos-extend-start still does that | 21:05 |
SamYaple | if there are issue, i would recommend not doing it in the container anymore (we had issue). see https://review.openstack.org/#/c/257800/ | 21:06 |
*** dave-mccowan has quit IRC | 21:06 | |
*** dave-mccowan has joined #kolla | 21:07 | |
stvnoyes | SamYaple: thanks for the info. do you remember what the problem was doing that inside the container? | 21:07 |
SamYaple | jeffrey4l fixed it, but if i remember there was a hang of apache2 which was causing the user creates to fail | 21:11 |
SamYaple | but we didn't kill the process we just exited | 21:12 |
SamYaple | yea stvnoyes we didnt do this whole section https://github.com/openstack/kolla-mesos/blob/master/config/keystone/templates/extend_start.sh.j2#L43-47 | 21:13 |
SamYaple | so any hang you might be seeing, we never did | 21:13 |
*** macsz has joined #kolla | 21:14 | |
*** dratushnyy has joined #kolla | 21:15 | |
openstackgerrit | Hui Kang proposed openstack/kolla: Fix missing nova instance directory in nova_compute bootstrap https://review.openstack.org/266575 | 21:16 |
*** macsz1 has joined #kolla | 21:18 | |
*** rhallisey has quit IRC | 21:18 | |
*** macsz has quit IRC | 21:21 | |
inc0 | SamYaple, quick question, any good way to avoid "failed to parse json" during ceph deploy in Fetching keys task? | 21:21 |
asalkeld | stvnoyes: we should do that eqivelent change too | 21:21 |
*** cloudnautique has quit IRC | 21:21 | |
inc0 | I get error and it's not helpful at all | 21:21 |
*** cloudnautique has joined #kolla | 21:21 | |
*** mdnadeem has joined #kolla | 21:29 | |
*** aginwala has joined #kolla | 21:31 | |
sdake | btw ansible 2.0 is out folks | 21:32 |
sdake | they released today or yeesterday | 21:32 |
sdake | that is something we may need to tackle in n or later | 21:32 |
dmsimard | today | 21:33 |
dmsimard | I'm liking the playbook blocks, tried them and it works well but not in every scope | 21:33 |
stvnoyes | asalkeld: np, i'll take a look at the change for that | 21:34 |
asalkeld | thanks | 21:34 |
dmsimard | ansible 2 also has the new openstack modules backed by shade thanks to monty | 21:34 |
*** cemmason1 has joined #kolla | 21:37 | |
*** cemmason has quit IRC | 21:37 | |
openstackgerrit | Merged openstack/kolla-mesos: kolla-mesos-ansible script https://review.openstack.org/263907 | 21:38 |
openstackgerrit | Merged openstack/kolla-mesos: Vagrant multinode deployment https://review.openstack.org/262242 | 21:40 |
*** cemmason has joined #kolla | 21:40 | |
openstackgerrit | Merged openstack/kolla-mesos: Fetching IP addresses for vagrant-hostmanager from libvirt https://review.openstack.org/263878 | 21:40 |
*** cemmason1 has quit IRC | 21:41 | |
*** salv-orlando has quit IRC | 21:43 | |
*** salv-orlando has joined #kolla | 21:44 | |
*** mdnadeem has quit IRC | 21:46 | |
*** cloudnautique has quit IRC | 21:46 | |
*** cloudnautique has joined #kolla | 21:47 | |
*** cloudnautique has quit IRC | 21:47 | |
*** cloudnautique has joined #kolla | 21:48 | |
*** cemmason has quit IRC | 21:48 | |
*** cemmason has joined #kolla | 21:48 | |
*** pbourke_ has quit IRC | 21:52 | |
*** pbourke_ has joined #kolla | 21:53 | |
*** diogogmt has quit IRC | 21:56 | |
dmsimard | SamYaple: are you familiar with that issue at all ? https://groups.google.com/forum/#!topic/open-iscsi/GhKfxIix4ds | 21:57 |
dmsimard | I thought it was related to https://bugs.launchpad.net/kolla/+bug/1460129 but apparently it's another thing | 21:57 |
openstack | Launchpad bug 1460129 in kolla "Cinder needs tgtd" [Critical,Confirmed] - Assigned to Sam Yaple (s8m) | 21:57 |
*** tpot has joined #kolla | 21:57 | |
*** dwalsh has quit IRC | 21:59 | |
SamYaple | dmsimard: but no linting! | 22:02 |
SamYaple | dmsimard: ansible 2.0 ansible-lint is borked with blocks | 22:02 |
SamYaple | dmsimard: very familiar | 22:03 |
SamYaple | with the bug | 22:03 |
SamYaple | inc0: i mean i would run the command in the container. thats where i would start debugging | 22:04 |
*** aginwala has quit IRC | 22:06 | |
dmsimard | SamYaple: ansible-lint is not yet comaptible | 22:07 |
dmsimard | there's issues/pull requests opened | 22:07 |
dmsimard | SamYaple: back to the bug (google groups link) what's the actual impact ? None since default we use is Ceph ? | 22:07 |
dmsimard | i.e, we use krbd instead of iscsi | 22:08 |
SamYaple | dmsimard: none since we use --net=host for all of kolla | 22:08 |
SamYaple | and we dont use krbd, or rather qemu doesnt | 22:09 |
SamYaple | it uses the libs to talk directly to teh cluster | 22:09 |
*** salv-orl_ has joined #kolla | 22:09 | |
dmsimard | uh, something has to mount the volume somewhere ? | 22:09 |
SamYaple | nope | 22:09 |
SamYaple | qemu talks directly to the ceph cluster for reads and writes | 22:09 |
SamYaple | no mounting | 22:10 |
dmsimard | wow, I didn't even know it was able to do that | 22:10 |
*** jasonsb has quit IRC | 22:10 | |
dmsimard | is that new ? Got to read on that | 22:10 |
SamYaple | no its how its always been | 22:10 |
SamYaple | for openstack that is | 22:10 |
dmsimard | wth? lol | 22:10 |
*** cemmason has quit IRC | 22:11 | |
*** cemmason has joined #kolla | 22:11 | |
SamYaple | yea it was merged into qemu in like 2011 | 22:11 |
SamYaple | librados support that is | 22:11 |
*** salv-orlando has quit IRC | 22:12 | |
dmsimard | wow, I was always under the impression that there was a mount operation involved somehow | 22:12 |
dmsimard | I'll go to sleep less stupid tonight | 22:12 |
SamYaple | youre one of the lucky 10000 today | 22:13 |
dmsimard | there's always a relevant xkcd | 22:14 |
*** cemmason1 has joined #kolla | 22:14 | |
*** dims has quit IRC | 22:15 | |
*** cemmason has quit IRC | 22:15 | |
* dmsimard looks at it again | 22:15 | |
dmsimard | I'm not 30 yet so I'm safe | 22:15 |
dmsimard | whew. | 22:15 |
*** macsz1 has quit IRC | 22:16 | |
*** cemmason has joined #kolla | 22:17 | |
*** cemmason1 has quit IRC | 22:18 | |
*** diogogmt has joined #kolla | 22:20 | |
SamYaple | give it time | 22:21 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Test file-backend for gate https://review.openstack.org/266569 | 22:22 |
*** akwasnie has quit IRC | 22:28 | |
*** jpeeler has quit IRC | 22:28 | |
*** aginwala has joined #kolla | 22:28 | |
*** jtriley has quit IRC | 22:29 | |
*** _tzn has quit IRC | 22:31 | |
*** inc0 has quit IRC | 22:32 | |
openstackgerrit | Sam Yaple proposed openstack/kolla: Test file-backend for gate https://review.openstack.org/266569 | 22:36 |
*** akwasnie has joined #kolla | 22:39 | |
*** cemmason has quit IRC | 22:41 | |
*** cemmason has joined #kolla | 22:41 | |
*** cemmason1 has joined #kolla | 22:45 | |
*** cemmason has quit IRC | 22:46 | |
*** CheKoLyN has quit IRC | 22:59 | |
*** aginwala has quit IRC | 23:00 | |
*** aginwala has joined #kolla | 23:05 | |
*** cemmason has joined #kolla | 23:11 | |
*** cemmason1 has quit IRC | 23:11 | |
*** cemmason has quit IRC | 23:15 | |
*** dave-mccowan has quit IRC | 23:15 | |
*** cemmason has joined #kolla | 23:15 | |
*** tzn has joined #kolla | 23:17 | |
*** aginwala has quit IRC | 23:20 | |
*** aginwala has joined #kolla | 23:32 | |
*** cemmason1 has joined #kolla | 23:44 | |
*** cemmason has quit IRC | 23:44 | |
*** sdake has quit IRC | 23:45 | |
*** akwasnie has quit IRC | 23:50 | |
*** SourabhP has quit IRC | 23:56 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!