Wednesday, 2017-07-05

*** charcol has quit IRC00:00
*** acormier has quit IRC00:07
*** japestinho has joined #openstack-ansible00:13
*** chyka has joined #openstack-ansible00:14
*** fergus has joined #openstack-ansible00:16
*** fyu has joined #openstack-ansible00:27
*** fergus has quit IRC00:27
*** fyu is now known as fergus00:27
*** thorst has joined #openstack-ansible00:27
*** thorst has quit IRC00:27
*** cuongnv has joined #openstack-ansible00:48
*** qcrivera has joined #openstack-ansible01:18
*** charcol has joined #openstack-ansible01:31
*** markvoelker has joined #openstack-ansible01:36
*** cuongnv_ has joined #openstack-ansible01:37
*** cuongnv has quit IRC01:39
*** qcrivera has quit IRC01:54
*** qcrivera has joined #openstack-ansible01:59
*** chyka has quit IRC02:08
*** markvoelker has quit IRC02:08
*** cuongnv_ is now known as cuongnv02:12
*** lbragstad has joined #openstack-ansible02:17
*** qcrivera has quit IRC02:20
*** thorst has joined #openstack-ansible02:28
*** thorst has quit IRC02:33
*** japestinho has quit IRC02:43
*** charcol has quit IRC02:44
*** dixiaoli has joined #openstack-ansible02:52
*** markvoelker has joined #openstack-ansible03:06
*** yifei has quit IRC03:23
*** yifei has joined #openstack-ansible03:23
*** charcol has joined #openstack-ansible03:23
*** yifei has quit IRC03:24
*** gkadam has joined #openstack-ansible03:24
*** yifei has joined #openstack-ansible03:25
*** gkadam has quit IRC03:26
*** gkadam has joined #openstack-ansible03:27
*** markvoelker has quit IRC03:39
*** lbragstad has quit IRC03:55
*** thorst has joined #openstack-ansible03:58
*** thorst has quit IRC04:03
*** chhavi has joined #openstack-ansible04:04
*** cNilesh has joined #openstack-ansible04:06
*** hybridpollo has quit IRC04:07
*** evrardjp[m] has quit IRC04:11
*** evrardjp[m] has joined #openstack-ansible04:12
*** evrardjp[m] has joined #openstack-ansible04:12
*** charcol has quit IRC04:16
*** charcol has joined #openstack-ansible04:32
*** markvoelker has joined #openstack-ansible04:36
*** chhavi has quit IRC04:38
*** chhavi has joined #openstack-ansible04:41
*** chhavi has quit IRC04:44
*** dixiaoli has quit IRC05:00
*** markvoelker has quit IRC05:08
*** dixiaoli has joined #openstack-ansible05:27
*** tobberydberg has joined #openstack-ansible05:34
*** udesale has joined #openstack-ansible05:46
*** Oku_OS-away is now known as Oku_OS05:56
*** tobberyd_ has joined #openstack-ansible05:56
*** thorst has joined #openstack-ansible05:59
*** tobberydberg has quit IRC05:59
*** thorst has quit IRC06:05
*** markvoelker has joined #openstack-ansible06:05
*** tobberyd_ is now known as tobberydberg06:17
*** aszeszo has joined #openstack-ansible06:23
*** markvoelker has quit IRC06:39
*** markus_z has joined #openstack-ansible06:42
*** cshen_ has joined #openstack-ansible06:43
*** armaan has quit IRC06:45
*** armaan has joined #openstack-ansible06:45
*** udesale__ has joined #openstack-ansible06:49
*** udesale has quit IRC06:49
*** charcol has quit IRC06:52
*** tobberydberg has quit IRC06:53
*** tobberydberg has joined #openstack-ansible06:53
*** tobberydberg has quit IRC06:54
*** tobberydberg has joined #openstack-ansible06:55
*** tobberydberg has quit IRC06:59
*** tobberydberg has joined #openstack-ansible06:59
*** yolanda_ has joined #openstack-ansible07:01
*** mbuil has joined #openstack-ansible07:01
*** tobberydberg has quit IRC07:07
*** tobberydberg has joined #openstack-ansible07:07
*** tobberydberg has quit IRC07:08
*** tobberydberg has joined #openstack-ansible07:09
*** rodolof has quit IRC07:14
*** rodolof has joined #openstack-ansible07:14
*** gouthamr has joined #openstack-ansible07:15
*** shardy has joined #openstack-ansible07:16
*** rodolof has quit IRC07:23
*** cNilesh has quit IRC07:29
*** rodolof has joined #openstack-ansible07:35
*** markvoelker has joined #openstack-ansible07:36
*** oanson has quit IRC07:38
*** oanson has joined #openstack-ansible07:38
openstackgerritMarkos Chandras (hwoarang) proposed openstack/openstack-ansible-tests master: test-install-{nova,neutron}: Add support for openSUSE  https://review.openstack.org/48039307:51
*** tobberydberg has quit IRC07:57
*** tobberydberg has joined #openstack-ansible07:57
*** thorst has joined #openstack-ansible08:01
*** thorst has quit IRC08:05
*** tobberydberg has quit IRC08:06
*** tobberydberg has joined #openstack-ansible08:07
*** tobberydberg has quit IRC08:08
*** markvoelker has quit IRC08:08
*** pbandark has joined #openstack-ansible08:08
*** tobberydberg has joined #openstack-ansible08:09
*** tobberydberg has quit IRC08:09
*** tobberydberg has joined #openstack-ansible08:09
*** shardy is now known as shtab08:27
*** shtab is now known as shardy08:27
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Update nova test playbook  https://review.openstack.org/48040908:29
*** tobberydberg has quit IRC08:31
*** tobberydberg has joined #openstack-ansible08:31
*** tobberydberg has quit IRC08:33
*** tobberydberg has joined #openstack-ansible08:33
*** tobberydberg has quit IRC08:33
*** tobberydberg has joined #openstack-ansible08:34
hwoarangodyssey4me: could you add opensuse support in this^^^^  and then I can drop my patchset on https://review.openstack.org/#/c/480393/08:37
*** tobberydberg has quit IRC08:38
*** tobberydberg has joined #openstack-ansible08:39
odyssey4mehwoarang sure08:39
hwoarangthank you08:39
odyssey4meyou can perhaps adjust it to just do neutron on top of https://review.openstack.org/48021708:39
hwoarangok good point08:40
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Update nova test playbook  https://review.openstack.org/48040908:40
openstackgerritMarkos Chandras (hwoarang) proposed openstack/openstack-ansible-tests master: test-install-neutron: Add openSUSE support  https://review.openstack.org/48039308:43
odyssey4mehwoarang posed a question in https://review.openstack.org/#/c/480292/108:49
hwoarangodyssey4me: but this thing does something different on centos right?08:51
odyssey4mehwoarang I have no idea.08:51
odyssey4meAll that task in question is doing is the same as the one above - enabling a module.08:51
odyssey4meIf the ansible module understands how to do that on the underlying OS, then it should work.08:51
hwoarangyeah me neither... The comment from cloudnull seems to suggest that a2enmod only works on Debian (and now suse)08:52
odyssey4mehmm, it requires http://docs.ansible.com/ansible/apache2_module_module.html#requirements-on-host-that-executes-module08:52
hwoarangyeah08:53
hwoarangno clue if that's something centos can do08:53
hwoarangso i didn't touch it08:53
odyssey4meyep, after a little looking around it appears that's correct08:54
odyssey4me:(08:54
odyssey4meoh well +308:54
odyssey4me*208:54
hwoarang:)08:58
*** electrofelix has joined #openstack-ansible08:59
*** cNilesh has joined #openstack-ansible08:59
mbuilcan somebody point me to where the user_variables.yml is created in the aio flavor? That file in my deployment is very different to the "example" file under OSA/etc/openstack_deploy/user_variables.yml and I can't find anything in the bootstrap-host role :(09:00
odyssey4membuil the user_variables.yml file is in the bootstrap-host templates09:01
odyssey4mehttps://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/templates/user_variables.aio.yml.j209:01
*** tobberydberg has quit IRC09:03
*** tobberydberg has joined #openstack-ansible09:04
mbuilodyssey4me: aaah great! It makes sense :)09:04
*** markvoelker has joined #openstack-ansible09:06
*** tobberydberg has quit IRC09:08
odyssey4measettle so with the comments in https://review.openstack.org/#/c/480258/8/doc/source/index.rst I was under the impression that after this merges the structure would change? is that not the case?09:08
*** tobberydberg has joined #openstack-ansible09:08
odyssey4meare we stuck with /ocata/ocata forever?09:08
odyssey4meAnd, what will the right link be?09:08
mbuilodyssey4me: where do you think is the best place to define the neutron plugin that we want to use in our deployment? Do you think user_variables.yml is ok or is that file reserved for other kind of stuff?09:09
asettleodyssey4me: it's a good question. I can't imagine why it would be (ocata,ocata) forever09:09
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: March to the beat of the new docs drum  https://review.openstack.org/48025809:09
asettleIt is currently that because we put that in there (if you'll recall my amazing patch)09:10
odyssey4membuil that file's for overrides09:10
odyssey4meI don't think we want to add any more services to the default aio deployment at this time.09:10
mbuilodyssey4me: sorry, I meant in general, not just for aio09:11
odyssey4membuil it'll come down to how you want to make it easy for a deployer to consume the service - ideally they should implement a single setting, and the group_vars should enable things to work across all the integrations for them.09:11
odyssey4meSo, the question is, what integrations are needed for tacker?09:11
odyssey4meYou are the guy working on tacker, right?09:12
openstackgerritJuan Vidal Allende proposed openstack/openstack-ansible-os_neutron master: Increase Vagrant VM memory from 2048 to 4096 MB  https://review.openstack.org/48045909:12
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: March to the beat of the new docs drum  https://review.openstack.org/48025809:13
odyssey4meandymccr how much of a problem is it do you think that the mitaka/liberty docs are all gone09:13
odyssey4membuil all user-space changes to configuration should be done in user_variables, or by using the vars plugin's /etc/openstack_deploy/{group_vars,host_vars} directories09:14
mbuilodyssey4me: I am thinking on a general problem we will have in OPNFV. We will need to work with different deployments. A clear example is OVS because we need to test two scenarios, one with plain OVS and another with OVS+DPDK. We will need to define a condition for that in os_neutron so that it know what OVS to deploy. However, I am not sure where we should write the variable (e.g. ovs_dpdk: true/false) that will steer that condition. I was09:16
mbuilthinking about writing it in user_variables.yml, but I am not sure if this is following good practices09:16
*** admin0 has joined #openstack-ansible09:16
odyssey4meah yes09:16
asettleodyssey4me: he's just popped out for errands. I can't overly speak for him but LIberty and Mitaka are solidly EOL09:16
asettleYou can still access via tags in the repo, yes?09:17
odyssey4measettle yeah, they're in git of course - but no longer published09:17
odyssey4meand while they're EOL, there are still a lot of people running them09:17
asettleodyssey4me: As long as they are still accessible, I don't overly see the issue. Instead of linking to the published source, link to the github source.09:17
asettleThey can build from there using sphinx09:17
asettleCheckout the tag, build09:17
asettleVoila09:17
odyssey4methe old docs in openstack are meant to remain as a reference, so I'm personally not pleased at them being gone09:18
asettleAt least, that's what we do for docs.09:18
odyssey4meI remember when the essex docs were yanked when it went EOL, yet I was still a cloud operator running essex.09:18
asettleodyssey4me: to be honest, we've never published EOL docs. We do have to keep consistency. As long as they are still available.09:18
odyssey4meoperators do not understand how to build docs or use the tooling09:18
asettleodyssey4me: instructions are available?09:19
asettleWe are working on a proper archiving plan, but it's going to take ages to roll out because there's nobody to implement it, and we have to deal with this initial migration first and foremost09:19
odyssey4mehmm, yeah - understood09:19
odyssey4meoptions are there, of course, I just don't think it's very nice to operators to do this sort of thing09:19
odyssey4mebut I understand that we need progress09:20
openstackgerritJuan Vidal Allende proposed openstack/openstack-ansible-tests master: Increase Vagrant VM memory from 2048 to 4096 MB  https://review.openstack.org/48046309:20
odyssey4mehmm, we still have a mitaka branch, so I could push a patch for the purpose of ensuring a last publish is done :)09:20
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/mitaka: Dummy patch to republish docs  https://review.openstack.org/48046609:23
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/mitaka: Dummy patch to republish docs  https://review.openstack.org/48046609:23
*** dixiaoli has quit IRC09:38
*** pbandark has quit IRC09:38
*** markvoelker has quit IRC09:39
openstackgerritMerged openstack/openstack-ansible-tests master: Split cinder API and the rest  https://review.openstack.org/48022109:39
*** pbandark has joined #openstack-ansible09:46
andymccrodyssey4me: yeah maybe we should do that - i'd say we would be better off with mitaka docs still (at least)09:47
andymccrmaybe we should plan something and talk to infra at the PTG or before09:47
asettleodyssey4me: yeah, so, while we work on the migration, we can have pointers to the github pages for the next 2/3 months while we figure it out09:49
asettleodyssey4me: I just left a message for Andreas and Doug to start thinking about this change too09:51
*** thorst has joined #openstack-ansible10:02
*** thorst has quit IRC10:06
*** cuongnv has quit IRC10:14
odyssey4meandymccr asettle https://review.openstack.org/480466 will hopefully republish the last mitaka docs10:15
odyssey4meeven if they do go to the wrong place10:15
odyssey4meanywhere is better than nowhere10:15
*** schwicht has quit IRC10:21
openstackgerritMerged openstack/openstack-ansible-tests master: Update neutron test playbook  https://review.openstack.org/48021710:25
*** schwicht has joined #openstack-ansible10:25
*** udesale__ has quit IRC10:31
*** markvoelker has joined #openstack-ansible10:36
asettleodyssey4me: where are they being linked from?10:40
asettleOr are you just reviving them to the old link?10:40
odyssey4measettle I'd need to revise https://review.openstack.org/480258 to re-include them after we see where they live.10:40
asettleodyssey4me: okay cool10:40
odyssey4meor patch on top of https://review.openstack.org/480258 - depending on what happens first10:41
*** tobberydberg has quit IRC10:43
*** tobberydberg has joined #openstack-ansible10:44
*** smatzek has joined #openstack-ansible10:46
*** thorst has joined #openstack-ansible10:47
*** thorst has quit IRC10:49
*** stuartgr has joined #openstack-ansible10:53
andymccrodyssey4me: we're processing EOL'ing so roles have been EOL'd for mitaka - i'll hold off on EOL'ing stable/mitaka - i need to do a sha bump to point to mitaka-eol tag for upstream/role repos10:58
andymccranyway10:58
andymccrso we should be ok to get that docs stuff fixed10:58
odyssey4meyup10:59
*** thorst has joined #openstack-ansible11:00
andymccrbbiab getting some lunch!11:00
*** schwicht has quit IRC11:01
*** tobberyd_ has joined #openstack-ansible11:08
*** shardy is now known as shardy_lunch11:08
*** markvoelker has quit IRC11:10
*** tobberydberg has quit IRC11:11
*** armaan has quit IRC11:23
*** armaan has joined #openstack-ansible11:23
*** hw_wutianwei has quit IRC11:36
*** thorst has quit IRC11:36
*** cshen_ has quit IRC11:38
*** schwicht has joined #openstack-ansible11:39
*** cshen_ has joined #openstack-ansible11:39
*** acormier has joined #openstack-ansible11:45
*** admin0 has quit IRC11:48
*** admin0 has joined #openstack-ansible11:51
*** acormier has quit IRC11:53
*** acormier has joined #openstack-ansible11:54
*** acormier has quit IRC11:58
*** vishwanathj has joined #openstack-ansible12:03
*** vishwanathj has quit IRC12:04
logan-odyssey4me jmccrory regarding the connection plugin issue, https://review.openstack.org/#/c/466743/ hacks together some tests for a lot of the cases fixed recently. the test case for https://review.openstack.org/#/c/472061/ is the one still failing there12:06
*** markvoelker has joined #openstack-ansible12:07
odyssey4melogan- ja, been watching that one - hoping for it to be resolved at some point12:07
*** markvoelker has quit IRC12:16
*** markvoelker has joined #openstack-ansible12:17
*** schwicht has quit IRC12:20
openstackgerritMerged openstack/openstack-ansible stable/mitaka: Dummy patch to republish docs  https://review.openstack.org/48046612:21
*** tobberyd_ is now known as tobberydberg12:24
mgariepymorning everyone12:24
odyssey4mebother, andymccr this is popping up on our docs build jobs: http://logs.openstack.org/63/480463/2/gate/gate-openstack-ansible-tests-docs-ubuntu-xenial/528c148/console.html#_2017-07-05_12_19_23_94137812:24
odyssey4menot sure if it's a general problem just yet12:25
odyssey4mecloudnull answered your Q's in https://review.openstack.org/479395 & https://review.openstack.org/48018612:25
andymccrhmm12:36
andymccrthats not great12:36
hwoarangi wonder what changed since this morning12:39
*** shardy_lunch is now known as shardy12:41
*** thorst has joined #openstack-ansible12:44
*** thorst_ has joined #openstack-ansible12:45
*** cshen_ has quit IRC12:47
*** thorst has quit IRC12:48
odyssey4meandymccr hwoarang AJAeger says that dhellman is already looking into it12:56
hwoarangaha12:57
odyssey4methey suggest we disable shinxmark for now, which for us is a >40 repo task... I'd rather just live with docs fails for a few hours or a day12:57
hwoaranghmm12:58
odyssey4meI'd rather we cap using u-c, than disable sphinxmark across all repositories.12:58
odyssey4mehmm, we can work around this12:58
odyssey4mehang a sec, let me try something12:58
openstackgerritMerged openstack/openstack-ansible-os_neutron stable/newton: Simplify database migrations  https://review.openstack.org/47888812:59
*** cathrichardson has joined #openstack-ansible12:59
mhaydenodyssey4me: the ansible-hardening docs are still throwing a 404 -- is there a step i've missed somewhere?13:00
odyssey4meoh bother, the docs tox envs don't use our tests repo requirements files13:00
*** lucasxu has joined #openstack-ansible13:00
odyssey4memhayden yeah, you need to push a patch and merge it to publish13:00
odyssey4meall old docs are gone13:00
mhaydenoh hah13:00
mhaydenjust any ol' patch?13:00
odyssey4meyup, merge anything and it should trigger a docs build13:00
mhaydenokay, thanks13:00
*** esberglu has joined #openstack-ansible13:01
openstackgerritMajor Hayden proposed openstack/ansible-hardening master: [Docs] Adjust hardening domains page  https://review.openstack.org/48057313:02
mgariepyandymccr, mind if I fix your patch https://review.openstack.org/#/c/480236/ ?13:03
openstackgerritMajor Hayden proposed openstack/openstack-ansible-security master: [Docs] Adjust deprecation warning  https://review.openstack.org/48057413:03
odyssey4mehmm, this is weird - u-c has sphinxmark 0.1.18 and it's 0.1.19 that's breaking things13:03
*** schwicht has joined #openstack-ansible13:04
odyssey4meoh, it's not installing the new one13:04
*** qcrivera has joined #openstack-ansible13:04
andymccrmgariepy: sure!13:10
asettleodyssey4me: passing docs patch. We can continue to iterate if anything is up13:10
asettleThanks for doing that13:10
andymccrahh just gotta create the directory13:11
andymccractually i made the same mistake in the patch i want to submit upstream13:11
andymccrand fixed that13:11
openstackgerritMarc Gariépy proposed openstack/openstack-ansible master: Set PrivateDevices=false for CentOS7 Ceph deploys  https://review.openstack.org/48023613:13
mgariepyhmm not sure what the perms should be for the directory.13:13
*** admin0 has quit IRC13:14
andymccrmgariepy: i just set it to defaults and didnt set them13:14
openstackgerritMarc Gariépy proposed openstack/openstack-ansible master: Set PrivateDevices=false for CentOS7 Ceph deploys  https://review.openstack.org/48023613:15
mgariepyroot:root 0755 (from memcached config file.13:15
mgariepyfrom : https://github.com/openstack/openstack-ansible-memcached_server/blob/master/tasks/memcached_config.yml#L53-L6113:16
mhaydenodyssey4me: could you help me get this over the hump? https://review.openstack.org/480573 :)13:16
mhaydenmgariepy: started getting some more virtlogd selinux denials over the weekend13:17
mhaydenhaven't dug into them yet13:17
mgariepyselinux is painful !13:17
mgariepyhaha :D13:17
mhaydenhah, well this isn't a red hat project, so there's not much effort going into policies here outside of what we're doing13:18
mhaydenred hat still tells people to disable selinux for clustering the last time i looked :(13:18
mgariepyaren't we doing cluster ?13:18
mgariepyshouldn't we disable it ?13:18
mgariepy;p13:18
*** hw_wutianwei has joined #openstack-ansible13:18
odyssey4memhayden unfortunately the docs jobs are broken right now13:20
odyssey4meso best to hold off13:20
odyssey4mesee https://review.openstack.org/48057713:20
*** jwitko has joined #openstack-ansible13:21
mhaydenhooray13:21
*** acormier has joined #openstack-ansible13:21
mgariepymhayden, is your cluster still on ocata ?13:22
mhaydeni dropped a worthless onesie on it13:22
mhaydenmgariepy: yeah13:22
mhaydeni'm still wrestling with some interesting bridge issues after containers are online for a long time13:22
mgariepywhat is causing the selinux issue ? juste upgrading libvirt ?13:22
mhaydeni'll get into a situation where the container can reach other containers on the same subnet13:22
mhaydenbut it can't get traffic forwarded off the bridge and out to other networks13:23
mhaydenit looks like it might be a conntrack issue because the actual connection tracking flow is missing for the container when it happens13:23
mhaydenif i tcpdump the bridge with promisc enabled, everything is back to normal13:23
mhaydeni even tried arping from inside the container -- no dice13:24
mhaydenodyssey4me: i don't think i use sphinxmark for the security hardening stuff13:24
mhaydeni can check again13:24
mgariepyhmm. that's pretty weird.13:25
mhaydenthe traffic isn't making it onto the wire13:25
mhaydeni see the outbound, but nothing coming back13:25
mhaydenwell i don't see the outbound trafifc on the physical interface13:25
mhaydenit's confusing13:25
*** askb has quit IRC13:25
mgariepyyou have 4 servers ? 3 controllers/compute + 1 compute ?13:26
mhaydenmgariepy: i'm using the new hotness HYPERCONVERGED13:27
odyssey4memhayden indeed you do not, so https://review.openstack.org/479894 is incoming13:27
mhayden4 servers -- all four are hypervisors13:27
mhaydenshared infra on nodes 1,2,313:27
mhaydenopenstack on nodes 2,313:27
mhaydenrsyslog on node 013:27
mgariepyI'll poke my friends in victoria to see if they see something similar13:28
mgariepyit's the same for all the containers or just those that runs openstack ?13:29
mgariepyor maybe the neutron agent on the compute is messing things up ?13:29
mhaydencould be, haven't narrowed it down quite yet13:30
mhaydenthe instances are ok13:30
mhaydenjust the containers running management stuff that break13:30
mhaydenand it's not a specific container -- seen it on galera, rabbit, nova, glance, etc13:31
mgariepyyou have instance on all 4 server ?13:31
mhaydenmgariepy: yeah13:32
mhaydenrunning 7 instances right now13:32
mhaydenicanhazip is running on there, so it gives the ethernet cards a good workout13:32
mhaydenand it has a BOATLOAD of conntrack flows13:32
*** gkadam has quit IRC13:33
mhaydenmy linux router out front shows about 252K conntrack flows right now13:33
mhaydenthat creeps up over 425K at certain times of day13:33
*** cNilesh has quit IRC13:33
openstackgerritNolan Brubaker proposed openstack/openstack-ansible-os_swift master: Don't run sync in upgrades  https://review.openstack.org/47904113:34
*** klamath has joined #openstack-ansible13:37
*** klamath has quit IRC13:38
*** klamath has joined #openstack-ansible13:38
openstackgerritMerged openstack/ansible-hardening master: Doc migration fixes  https://review.openstack.org/47989413:39
asettleodyssey4me: are you in the openstack-doc channel? Doug is asking about the docs publishing13:41
mhaydenevrardjp[m]: thanks for the comments on the ansible-keepalived patch13:46
mhaydeni'll try to work on that today13:46
andymccrmgariepy: https://github.com/ceph/ceph-ansible/pull/1654 for reference13:50
mgariepyandymccr, nice13:52
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Fix regression with container_name var  https://review.openstack.org/48059113:53
*** acormier has quit IRC13:58
*** udesale has joined #openstack-ansible14:01
*** armaan has quit IRC14:02
*** armaan has joined #openstack-ansible14:02
*** schwicht has quit IRC14:02
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Fix the physical_host_addr vars lookup  https://review.openstack.org/48059414:03
*** galstrom_zzz is now known as galstrom14:06
*** schwicht has joined #openstack-ansible14:07
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Fix regression with container_name var  https://review.openstack.org/48059114:10
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Add container-aware connection plugin tests  https://review.openstack.org/46674314:10
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Fix the physical_host_addr vars lookup  https://review.openstack.org/48059414:10
*** pbandark1 has joined #openstack-ansible14:10
*** pbandark has quit IRC14:12
*** pbandark1 is now known as pbandark14:12
*** schwicht has quit IRC14:24
mhaydenevrardjp[m]: regarding that keepalived selinux PR, where do you think it belongs in OSA?14:24
mhaydeni guess it could go into the hosts role14:25
*** aszeszo has quit IRC14:27
*** woodard has joined #openstack-ansible14:29
*** schwicht has joined #openstack-ansible14:30
mgariepymhayden, I guess the config should go in https://github.com/openstack/openstack-ansible/blob/master/group_vars/all/keepalived.yml14:30
mhaydenseems like a good idea -- where should the .te files go?14:31
mgariepythe haproxy selinux rule for keepalived is very specific to our configuration14:31
mhaydentrue14:31
odyssey4memhayden an option would be the haproxy role, activated by a var in group_vars14:33
odyssey4meanother option would be to use the playbook to do the necessary things14:33
mhaydenwell what i'm thinking is that we're doing SELinux adjustments in a few different places14:33
mhaydendidn't we centralize apparmor stuff in one place?14:33
odyssey4meI know we setup an apparmor profile in lxc_hosts, but I'm not aware of any other config being done.14:34
odyssey4meThat apparmor stuff is specifically for LXC.14:34
mhaydentrue14:35
mhaydeni'd like to centralize our SELinux stuff into one place14:35
mgariepywe could have a selinux role !14:36
mgariepy:P14:36
mhaydenERMAGERD14:36
odyssey4mewell, is there an selinux role that could be used in ansible-galaxy?14:36
odyssey4meif not, perhaps there should be one14:36
mhaydenone could be made it not14:36
* mhayden looks14:36
mgariepywould be a lot better to have selinux stuff working as a module from ansible.14:36
odyssey4mebut I'd suggest then that the role be prefixed with 'ansible-' and not 'openstack-ansible-' as it's a general role which could be used across projects and outside of openstack14:37
odyssey4meor that ^14:37
mgariepybut that's another issue.14:37
odyssey4methere may already be an selinux role?14:37
odyssey4me*module14:37
mhaydeni don't mind building/maintaining one in github14:37
mhaydenmost of the ones i'm finding actually disable selinux14:37
mhaydensigh14:37
odyssey4mehahahaha14:37
mhaydenodyssey4me: don't laugh14:38
mgariepythey follow rhel docs !14:38
mhayden:P14:38
mgariepylol14:38
*** cpuga has joined #openstack-ansible14:39
mhaydenone problem i often find with ansible-galaxy is that the roles are low quality, or won't pass linting tests :/14:39
mhaydeni found one that will do it, but it needs an overhaul14:39
mgariepylink ?14:39
odyssey4meyup14:39
odyssey4meI typically check to see whether it has regular maintenance, releases and CI testing. If not, I don't bother with it.14:40
mhaydenhttps://github.com/MWojtowicz/ansible-semodule14:40
mhaydenno ci14:40
odyssey4meIt's often more work to overhaul it than it's worth,14:40
mhaydenlast commit on Feb 2114:40
mgariepy1 commit14:40
*** cpuga_ has joined #openstack-ansible14:41
mgariepynice :)14:41
mhaydenoh hah14:41
mhayden:P14:41
odyssey4mewell, you could reach out to the owner14:41
mhaydeni'd also like to find the most optimized way to get this done14:41
mhaydencould easily copy all of the .te's at once and async build the modules14:41
*** schwicht has quit IRC14:42
mhaydenbut even if we did it this way (bring in a role from galaxy), where would the .te files live?14:44
mgariepywhat else are we configuring for selinux ?14:44
mhaydenthe ansible-hardening role enables it14:44
*** cpuga has quit IRC14:44
mhaydenwe have selinux stuff in swift, nova, keepalived, haproxy right now14:44
mgariepythe te file should live in the role that needs it.14:44
mgariepyor depends on it (for haproxy/keepalived )14:44
*** acormier has joined #openstack-ansible14:45
mgariepyso the role can be consumed outside of OSA ?14:45
mhaydenright14:45
mhaydenthe names of the policies could easily be tossed in via variable14:45
mhaydeni'm just wondering if we ought to put some selinux playbook goodness into OSA that does all of the SELinux work up front14:46
mhaydenbut then again, that wouldn't help IRR testing :/14:46
mhaydenso i think your idea is best, mgariepy14:46
mgariepyto enable selinux there is an ansible module14:46
mhaydenwe could write a "secompile" module :P14:47
mhaydenbut that seems like overkill14:47
mgariepyhaha :)14:47
mhaydenit would be a bunch of python calling out to shell14:47
openstackgerritMerged openstack/openstack-ansible stable/newton: Implement rolling upgrades for cinder  https://review.openstack.org/47625614:50
*** perniciouscaffei has quit IRC14:51
*** schwicht has joined #openstack-ansible14:54
*** markus_z has quit IRC14:56
*** fergus has quit IRC14:56
*** fergus has joined #openstack-ansible14:58
*** lucasxu has quit IRC15:00
*** Oku_OS is now known as Oku_OS-away15:01
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova master: Implement data migrations for rolling upgrades  https://review.openstack.org/47939515:05
*** lbragstad has joined #openstack-ansible15:05
odyssey4mecloudnull ^ done15:06
odyssey4methanks for going through that, I know it's a bit complicated15:06
openstackgerritNolan Brubaker proposed openstack/openstack-ansible-os_swift master: Don't run sync in upgrades  https://review.openstack.org/47904115:10
cshenis anybody familiar with dhcp problem in openstack? we saw an IP was released with "unknown lease" reason in dnsmasq. So the VM was booting, it could not get the IP. How can I force dnsmasq to offer the IP again?15:16
cshenlog is like "Jul  4 16:24:18 dnsmasq-dhcp[8732]: DHCPRELEASE(tap0665484b-42) 10.131.2.75 fa:16:3e:3e:5f:c3 unknown lease."15:16
*** udesale has quit IRC15:18
*** schwicht has quit IRC15:23
cloudnullcshen: if you can console into the vm you can run dhclient15:25
cloudnullor you can wait for the dhcp timeout.15:25
cloudnullor reboot the instance15:25
cshencloudnull: I did reboot the VM, it didn't work. It seems that the dhcp IP was already RELEASED in dnsmasq.15:26
cshenbut in horizon, you still see it was assigned to that VM.15:26
*** schwicht has joined #openstack-ansible15:26
cloudnullthis is dnsmasq running within the neutron agent container within a specific network namespace ?15:26
openstackgerritMajor Hayden proposed openstack/openstack-ansible-specs master: [Spec] Add software firewall  https://review.openstack.org/47941515:27
cshencloudnull: domain is openstacklocal.15:27
cloudnullcan you console to the instance ?15:28
cshenI can't at the moment. But I brutally restarted all the services in neutron-agent.15:29
cshenand it seems the that 10.131.2.75 was offered by dnsmasq again.15:29
*** aszeszo has joined #openstack-ansible15:30
cshenJul  5 15:14:39 dnsmasq-dhcp[17568]: DHCPOFFER(tap0665484b-42) 10.131.2.75 fa:16:3e:a3:3a:8215:30
cloudnullyou could kill the dnsmasq process within the neutron-agent container(s), rm the leases file (/var/lib/neutron/dhcp/$NETWORK_ID/leases), and restart the neutron-agents.15:30
cshencloudnull: thanks for your great tips.15:31
cshenI will try.15:31
*** cNilesh has joined #openstack-ansible15:31
*** perniciouscaffei has joined #openstack-ansible15:34
*** lucasxu has joined #openstack-ansible15:35
cshencloudnull: I believe your tips will do the trick. since I restarted all services in neutron-agent, and the DHCP IP was offered again. The VM can see this offered IP now. thanks for help.15:38
*** dixiaoli has joined #openstack-ansible15:39
andymccrif somebody is about to review: https://review.openstack.org/#/c/467222/ would be appreciated! It's the last thing standing in the way of the translations check site :)15:42
*** dixiaoli has quit IRC15:44
palendaeAre docs jobs broken at the moment?15:45
palendaePIL's failing to run15:45
andymccrpalendae: yeah unfortunately. i believe a fix is being worked on by the infra people15:46
palendaeOk, thanks15:46
palendaeWanted to be sure it wasn't something I had to do for the swift role15:46
andymccrhaha yeah i think you're ok15:47
palendaeAlright, will keep an eye out then15:48
palendaeThanks15:48
mhaydenodyssey4me: could you give me a little nudge on this one? https://review.openstack.org/48057315:49
mhaydenandymccr: gotcha15:49
odyssey4meall docs/linters jobs will fail until https://review.openstack.org/480577 merges palendae andymccr15:50
odyssey4meall docs/linters jobs for *master*15:50
odyssey4mestable branches are just fine15:51
andymccrahh so we'll need to sha bump requirements?15:51
odyssey4memhayden done15:51
odyssey4meandymccr hmm, no the integrated repo should be fine actually15:51
odyssey4methe breaking change in u-c merged this morning15:51
*** qcrivera has quit IRC15:51
andymccrodyssey4me: ahh ok yeah that makes sense15:52
odyssey4meso it's only the docs/linters because they grab straight from the head of the branch, not our fixed SHA's15:52
*** schwicht has quit IRC15:52
palendaeodyssey4me: Thanks, I'll add a depends-on I guess15:53
odyssey4mepalendae that won't help, I don't think15:53
palendaeNo, but it'll communicate status15:53
palendaeOr maybe jsut a comment15:53
odyssey4methe docs/linters use the published u-c15:53
openstackgerritLogan V proposed openstack/openstack-ansible-plugins master: Add container-aware connection plugin tests  https://review.openstack.org/46674315:54
odyssey4meif you like - it'll be merged within the hour I reckon, so it's kinda pointless15:54
mhaydenodyssey4me: gracias15:54
*** armaan has quit IRC15:55
*** schwicht has joined #openstack-ansible15:56
*** cshen has quit IRC15:56
*** BjoernT has joined #openstack-ansible15:58
*** itdependsnetwork has joined #openstack-ansible15:59
*** perniciouscaffei has quit IRC16:04
*** perniciouscaffei has joined #openstack-ansible16:04
*** jamesden_ has joined #openstack-ansible16:04
*** jamesdenton has quit IRC16:04
openstackgerritMerged openstack/ansible-hardening master: [Docs] Adjust hardening domains page  https://review.openstack.org/48057316:05
*** schwicht has quit IRC16:06
*** qcrivera has joined #openstack-ansible16:06
mgariepyandymccr, great success :D http://logs.openstack.org/36/480236/3/check/gate-openstack-ansible-openstack-ansible-ceph-centos-7-nv/f666d97/console.html#_2017-07-05_14_10_06_40655316:08
*** tobberyd_ has joined #openstack-ansible16:08
*** tobberydberg has quit IRC16:08
*** tobberyd_ has quit IRC16:09
*** aszeszo1 has joined #openstack-ansible16:09
andymccrmgariepy: hopefully now all the periodicals are mostly in a working state! at least on master16:09
*** tobberydberg has joined #openstack-ansible16:09
*** aszeszo has quit IRC16:10
mgariepyocata failed this morning16:11
mgariepyfor centos**16:11
*** aszeszo has joined #openstack-ansible16:11
andymccrmgariepy: ugh :P16:12
odyssey4meI have a bash-ism question which I'm trying to puzzle out, perhaps someone can assist? cc hughsaunders16:13
odyssey4mecurrently https://github.com/openstack/openstack-ansible/blob/master/scripts/gate-check-commit.sh#L187 is not working as I expect it to16:13
odyssey4mewhen I execute the script and background it by hand, it works exactly how it should16:13
*** aszeszo1 has quit IRC16:13
odyssey4mehowever in the gate it's behaving more like it does when I execute it using a subshell, eg: bash ./tests/data-plane-test.sh &16:14
odyssey4methe behaviour is that it appears to be running in the background, but it does nothing16:14
odyssey4meany thoughts on what might be going wrong there?16:14
andymccrhmm mgariepy seems to be some cinder attach bits. i'll take a look tomorrow if it fails again and debug more16:17
*** smatzek has quit IRC16:17
openstackgerritAndy McCrae proposed openstack/openstack-ansible master: Add translations test-site deployment option  https://review.openstack.org/46722216:19
mgariepyandymccr, ok great16:19
*** galstrom is now known as galstrom_zzz16:19
mgariepyseems consistant with the upgrade test tho.16:19
andymccrahh thats new then. i wonder if something changed16:20
andymccrahh the new sha's merged16:20
*** markvoelker has quit IRC16:24
*** galstrom_zzz is now known as galstrom16:24
*** lucasxu has quit IRC16:27
*** jamesdenton has joined #openstack-ansible16:29
*** jamesden_ has quit IRC16:29
openstackgerritMerged openstack/openstack-ansible-os_glance stable/ocata: Only sync glance database on major versions  https://review.openstack.org/47938416:30
*** tobberydberg has quit IRC16:30
*** markvoelker has joined #openstack-ansible16:30
*** tobberydberg has joined #openstack-ansible16:30
mbuilguys, I am trying to deploy now with 'mini' and I am getting a problem when installing rabbitmq. it gets a urlopen error when trying to download "http://172.29.236.11:8181/os-releases/master/ubuntu-16.04-x86_64/get-pip.py"16:32
openstackgerritMerged openstack/openstack-ansible-os_glance stable/ocata: Ensure that services restart in a particular order  https://review.openstack.org/47960416:32
openstackgerritMerged openstack/openstack-ansible-os_glance stable/ocata: Use listen instead of chained notifiers  https://review.openstack.org/47960516:32
openstackgerritMerged openstack/openstack-ansible-os_glance stable/newton: Only sync glance database on major versions  https://review.openstack.org/47938516:32
*** rodolof has quit IRC16:32
evrardjpmgariepy: mhayden I didn't follow the whole conversation16:32
mbuilThat file does not exist. In that directory I only have requirements.txt  requirements_constraints.txt. Is there something wrong when creating the repo?16:32
evrardjpbut I think we could have a ok-ish approach, a good enough approach, and a nice approach16:33
evrardjplet's start with the nice16:33
evrardjpwe could have a repo that is solely focusing on writing selinux modules, that basically allow the compilation of policies16:34
*** schwicht has joined #openstack-ansible16:34
evrardjpwe can therefore include this as meta dependency to our roles with no overhead16:34
evrardjpeach of the roles could then ship their own files and use this module16:35
evrardjpthe good enough approach, is we ship those tasks in some role, and include the role in playbooks or roles16:35
openstackgerritMerged openstack/openstack-ansible-os_neutron stable/newton: Ensure that services restart in a particular order  https://review.openstack.org/47889516:36
openstackgerritMerged openstack/openstack-ansible-os_neutron stable/newton: Resolve migration issues  https://review.openstack.org/47891716:36
*** mbuil has quit IRC16:37
evrardjpthe ok-ish approach is (for haproxy + keepalived matter) we store the haproxy policy file for keepalived into haproxy role, and define a group var that basically say "Hey the file to compile your keepalived policy is stored there..." and keepalived tasks will use it.16:37
*** toddnni has quit IRC16:39
openstackgerritMerged openstack/openstack-ansible stable/ocata: Consolidate final two keystone plays  https://review.openstack.org/47923916:41
*** toddnni has joined #openstack-ansible16:44
*** smatzek has joined #openstack-ansible16:53
*** galstrom is now known as galstrom_zzz17:05
*** galstrom_zzz is now known as galstrom17:08
*** cNilesh has quit IRC17:08
*** cNilesh has joined #openstack-ansible17:09
*** chyka has joined #openstack-ansible17:10
*** cNilesh has quit IRC17:17
*** cNilesh has joined #openstack-ansible17:18
*** aszeszo has quit IRC17:22
*** galstrom is now known as galstrom_zzz17:29
*** galstrom_zzz is now known as galstrom17:30
*** PerilousApricot has quit IRC17:34
*** cpuga_ has quit IRC17:35
*** cathrichardson has quit IRC17:35
odyssey4meanyone care to give this a nudge? not sure if jmccrory is around: https://review.openstack.org/47887717:40
* mhayden has already applied nudging17:42
*** cathrichardson has joined #openstack-ansible17:43
*** cpuga has joined #openstack-ansible17:45
*** craigs has joined #openstack-ansible17:47
*** gkadam has joined #openstack-ansible17:48
openstackgerritNolan Brubaker proposed openstack/openstack-ansible master: Install swift serially for rolling updates  https://review.openstack.org/47858017:49
*** cNilesh has quit IRC17:52
craigs@here will the patch - for erlang be merged to OSA newton? and if so what tag?17:57
craigssame patch but merged to ocata- https://review.openstack.org/#/c/474362/117:58
*** cshen has joined #openstack-ansible18:01
*** electrofelix has quit IRC18:02
*** cshen has quit IRC18:03
*** cshen has joined #openstack-ansible18:03
*** stuartgr has left #openstack-ansible18:08
*** BjoernT has quit IRC18:13
*** admin0 has joined #openstack-ansible18:15
admin0hi .. what variable to change openstack.local to my own domain ?18:15
*** hw_wutianwei has quit IRC18:16
*** gkadam has quit IRC18:19
*** rodolof has joined #openstack-ansible18:19
odyssey4meadmin0 I'd prefer to teach you to grep, then you can figure it out yourself. ;)18:20
odyssey4mecd /opt/openstack-ansible; find . -name "*.yml" -exec grep -r 'openstack.local' {} \;18:21
*** craigs has quit IRC18:25
*** schwicht has quit IRC18:26
*** aszeszo has joined #openstack-ansible18:26
*** schwicht has joined #openstack-ansible18:28
*** cshen has quit IRC18:32
admin0oh .. i searched the same, but in /etc/ansible/os_* :D18:33
*** qcrivera has quit IRC18:33
*** perniciouscaffei has quit IRC18:35
openstackgerritMerged openstack/openstack-ansible-os_keystone stable/newton: Implement serialisable rolling upgrade  https://review.openstack.org/47887718:40
openstackgerritMerged openstack/openstack-ansible-os_keystone stable/newton: Use run_once instead of inventory scoping  https://review.openstack.org/47924118:40
*** cathrichardson has quit IRC18:45
*** craigs_ has joined #openstack-ansible18:47
*** perniciouscaffei has joined #openstack-ansible18:49
*** jamesdenton has quit IRC18:55
*** jamesdenton has joined #openstack-ansible18:56
*** aszeszo1 has joined #openstack-ansible18:56
*** jamesdenton has quit IRC18:59
*** aszeszo has quit IRC18:59
*** jamesdenton has joined #openstack-ansible18:59
*** schwicht has quit IRC19:05
*** jamesdenton has quit IRC19:06
*** rromans has quit IRC19:06
*** schwicht has joined #openstack-ansible19:09
*** jamesdenton has joined #openstack-ansible19:11
*** hybridpollo has joined #openstack-ansible19:19
*** cathrichardson has joined #openstack-ansible19:23
*** aszeszo has joined #openstack-ansible19:25
*** cathrich_ has joined #openstack-ansible19:26
*** aszeszo1 has quit IRC19:27
*** cathrichardson has quit IRC19:28
*** rromans has joined #openstack-ansible19:29
*** Wayne_ has joined #openstack-ansible19:37
*** cpuga has quit IRC19:38
*** cshen has joined #openstack-ansible19:39
*** askb has joined #openstack-ansible19:40
*** galstrom is now known as galstrom_zzz19:49
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance stable/newton: Perform an atomic policy file change  https://review.openstack.org/47927519:55
*** ThiagoCMC has joined #openstack-ansible19:57
ThiagoCMCGuys, just curious, is anyone working in an "os_senlin" role for OSA?19:57
ThiagoCMCMaybe for Pike?19:57
odyssey4medmsimard cloudnull jmccrory evrardjp So, I did a thing and hope you can help me finish it. :) https://review.openstack.org/48071619:58
odyssey4meany cores around ready to give this one a nudge? https://review.openstack.org/47927519:59
odyssey4meThiagoCMC I'm not aware of anyone actually working on one.19:59
ThiagoCMCodyssey4me, Hmm... Okay, I'll give it a try... Might be a good way for learning about OSA internals...   =)20:00
dmsimardodyssey4me: I'll help review some more as it gets a little bit more fleshed out (minimum with integration tests? just making sure it at least runs and we recover the subunit file in logs), I'm in firefighting mode right now20:00
*** woodard_ has joined #openstack-ansible20:02
*** galstrom_zzz is now known as galstrom20:02
*** tobberyd_ has joined #openstack-ansible20:04
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance stable/newton: Ensure that services restart in a particular order  https://review.openstack.org/48072020:04
*** woodard__ has joined #openstack-ansible20:04
odyssey4medmsimard sure, not expecting much - but thanks for that pointer :)20:05
*** woodard_ has quit IRC20:05
*** woodard has quit IRC20:06
*** tobberydberg has quit IRC20:07
odyssey4meI just wanted to give a heads-up that it's on the way, finally.20:07
dmsimardodyssey4me: and it's awesome \o/20:08
dmsimardodyssey4me: after the meeting I'm on, I'll explain some ideas that have been brewing20:08
*** tobberyd_ has quit IRC20:09
*** itdependsnetwork has quit IRC20:09
*** jamesdenton has quit IRC20:10
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Implement rolling upgrades for glance  https://review.openstack.org/47938620:13
*** woodard__ has quit IRC20:15
*** PerilousApricot has joined #openstack-ansible20:15
*** woodard has joined #openstack-ansible20:15
PerilousApricotHi all, does OSA support using OVS bridges, or should deployments only use regular linux bridges20:16
*** woodard has quit IRC20:16
*** woodard has joined #openstack-ansible20:17
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for keystone  https://review.openstack.org/47924420:17
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for keystone  https://review.openstack.org/47924420:20
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Allow keystone playbook to work with limits  https://review.openstack.org/47924620:22
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Allow keystone playbook to work with limits  https://review.openstack.org/47924620:23
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Consolidate final two keystone plays  https://review.openstack.org/48072820:24
*** tobberydberg has joined #openstack-ansible20:24
*** tobberydberg has quit IRC20:27
*** tobberyd_ has joined #openstack-ansible20:27
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for neutron  https://review.openstack.org/47924820:32
*** perniciouscaffei has quit IRC20:33
*** shardy has quit IRC20:35
*** cshen has quit IRC20:36
odyssey4mePerilousApricot yes, OVS is a supported configuration20:38
odyssey4mea quick google gave me http://trumant.github.io/openstack-ansible-openvswitch.html20:38
PerilousApricotodyssey4me: I found the same. I get an error under centos, so I was trying to sort out which direction to work on solving it20:40
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for neutron  https://review.openstack.org/47924820:40
odyssey4meheh, no centos is unlikely to work with ovs20:41
odyssey4meno-one's put the effort into that20:41
odyssey4mefor ocata and master, centos support is still new - many, many bugs are still shaking out20:42
openstackgerritMerged openstack/openstack-ansible-os_glance stable/newton: Perform an atomic policy file change  https://review.openstack.org/47927520:42
odyssey4meI don't think OVS for centos is tested at all20:42
odyssey4mein OSA I mean20:42
PerilousApricotright20:42
*** jwitko has quit IRC20:43
PerilousApricotYeah, I get something weird with trying to attach bridges to the containers. We're primarily a centos shop, so I tried installing w/centos, but I guess i'll blow away the machines and put ubuntu on them instead20:44
PerilousApricotThanks for the help!20:45
*** PerilousApricot has quit IRC20:46
*** itdependsnetwork has joined #openstack-ansible20:47
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for glance  https://review.openstack.org/47938720:48
openstackgerritMajor Hayden proposed openstack/ansible-hardening master: Conditinally install EPEL if needed  https://review.openstack.org/48073420:49
openstackgerritMajor Hayden proposed openstack/ansible-hardening master: Allow epel-release package name customization  https://review.openstack.org/48073720:54
*** thorst_ has quit IRC20:55
*** esberglu has quit IRC20:55
*** galstrom is now known as galstrom_zzz20:55
*** thorst has joined #openstack-ansible20:57
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Implement rolling upgrades for keystone  https://review.openstack.org/47924420:59
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Allow keystone playbook to work with limits  https://review.openstack.org/47924620:59
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Consolidate final two keystone plays  https://review.openstack.org/48072820:59
*** thorst has quit IRC21:01
openstackgerritMajor Hayden proposed openstack/ansible-hardening master: Re-organize defaults/main.yml  https://review.openstack.org/48073921:04
*** galstrom_zzz is now known as galstrom21:08
dmsimardodyssey4me: so, this idea21:09
dmsimardodyssey4me: more and more I want to introduce the concept of driver in ARA -- for sending data in and getting data out21:10
*** woodard_ has joined #openstack-ansible21:10
dmsimardodyssey4me: for example, there's output in subunit, junit, html -- there's someone working on graphite right now and I also eventually want to be able to send data to elasticsearch for use in ELK21:11
openstackgerritNolan Brubaker proposed openstack/openstack-ansible-os_swift master: Use playbook hosts for ring synchronization  https://review.openstack.org/48074221:12
dmsimardAnd then there's input streams, right now there's a callback that sends data to sql but upstream is interested in having ARA pick up data from mqtt (firehose) since that's where zuul will be dropping messages21:12
openstackgerritNolan Brubaker proposed openstack/openstack-ansible master: Install swift serially for rolling updates  https://review.openstack.org/47858021:13
dmsimardI guess ARA would more or less become a common "database" format for ansible playbook runs or something like that21:13
palendaedmsimard: I believe twitter did something like that21:13
palendaeThey stored retries in a DB21:13
palendaeNot necessarily metrics21:13
dmsimardand then you'd have an API to abstract drivers to send things in and send things out21:14
*** woodard has quit IRC21:14
dmsimardthis assumes that ARA would have an API in the first place, which it doesn't really right now but it's also something I want to flesh out21:14
*** woodard_ has quit IRC21:15
dmsimardpalendae: Yeah, I don't think ARA is in the business of storing metrics -- but it has data that can be used in metric backends and interpreters21:15
palendaeSure21:15
palendaeWell, it reports on what happened...it's not directly handling stuff like retries21:16
dmsimardpalendae: they'd be translated into the process but things like, I don't know, task durations, amount of tasks and their results, etc.21:16
*** smatzek has quit IRC21:16
palendaeSure21:16
palendaeI guess what I'm saying is there's precedent in the ansible world21:16
palendaeMaybe not open sourced, but twitter definitely did a presentation on it21:16
dmsimardpalendae: trying to search about ansible and twitter on google is a bit of a nightmare :)21:18
*** itdependsnetwork has quit IRC21:18
palendaeOof, yeah21:18
palendaeIt's been a few years21:18
palendaeHere it is https://www.youtube.com/watch?v=fwGrKXzocg421:19
palendae2014, so pretty datd in Ansible-land21:19
dmsimardpalendae: thanks, I'll definitely watch it21:21
*** jamesdenton has joined #openstack-ansible21:27
*** thorst has joined #openstack-ansible21:31
dmsimardpalendae: looks like this is where they talk about it: https://www.youtube.com/watch?v=fwGrKXzocg4&t=122221:33
dmsimardpalendae: from what I can gather, this is mostly no longer a problem today (retrying only the things that failed)21:34
*** thorst has quit IRC21:35
*** fergus has quit IRC21:59
*** fergus has joined #openstack-ansible21:59
*** schwicht has quit IRC22:05
*** klamath has quit IRC22:14
*** esberglu has joined #openstack-ansible22:21
*** acormier has quit IRC22:22
*** acormier has joined #openstack-ansible22:22
*** schwicht has joined #openstack-ansible22:24
*** galstrom is now known as galstrom_zzz22:25
*** esberglu has quit IRC22:26
*** acormier has quit IRC22:27
*** andries has quit IRC22:35
*** admin0 has quit IRC22:37
*** andries has joined #openstack-ansible22:38
*** schwicht has quit IRC22:49
*** schwicht has joined #openstack-ansible22:49
*** schwicht has quit IRC22:55
*** thorst has joined #openstack-ansible23:12
*** thorst has quit IRC23:12
*** thorst has joined #openstack-ansible23:19
*** schwicht has joined #openstack-ansible23:22
*** rodolof has quit IRC23:34
*** chyka has quit IRC23:34
*** aszeszo has quit IRC23:35
*** rodolof has joined #openstack-ansible23:46
*** hw_wutianwei has joined #openstack-ansible23:53
openstackgerritLogan V proposed openstack/openstack-ansible-os_neutron master: Fix DB expand delegated facts  https://review.openstack.org/48077223:54
*** hybridpollo has quit IRC23:55
*** rodolof has quit IRC23:56
*** rodolof has joined #openstack-ansible23:56

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!