Friday, 2016-04-15

*** asettle has quit IRC00:06
*** skamithi13 has quit IRC00:18
*** saneax is now known as saneax_AFK00:20
*** yarkot1 has joined #openstack-ansible00:29
*** yarkot1 has quit IRC00:30
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Enable SSL termination for all services  https://review.openstack.org/27719900:32
cloudnulladmin0 thats the missing pieces that i had in a local branch now being forward ported ^00:32
cloudnulli've tested most, not all00:33
cloudnullplease give it a spin when you can00:33
cloudnullor want to00:33
cloudnulli got to go , off to homedepot00:33
mrdahomedepot is fun :)00:34
*** yarkot1 has joined #openstack-ansible00:34
openstackgerritJimmy McCrory proposed openstack/openstack-ansible-rabbitmq_server: Multi-distro framework for rabbitmq_server role  https://review.openstack.org/28628200:41
*** yarkot1 has quit IRC00:44
*** skamithi13 has joined #openstack-ansible00:45
*** yarkot1 has joined #openstack-ansible01:33
*** fishcried has joined #openstack-ansible01:47
*** fishcried has quit IRC01:48
*** fishcried has joined #openstack-ansible01:50
*** asettle has joined #openstack-ansible01:56
*** thorst has quit IRC01:56
*** thorst has joined #openstack-ansible01:57
*** v1k0d3n has joined #openstack-ansible02:03
*** severion has quit IRC02:05
*** thorst has quit IRC02:05
*** fishcried1 has joined #openstack-ansible02:24
*** fishcried has quit IRC02:25
*** marcoamorales has quit IRC02:25
*** fishcried1 is now known as fishcried02:25
*** asettle has quit IRC02:26
*** asettle has joined #openstack-ansible02:33
*** Mudpuppy has joined #openstack-ansible02:50
asettleodyssey4me and cloudnull - thanks for both your reviews on the Ironic docs. Just a note, this was not at all edited from Kevin's original content. I will get around to it, this was just a first off patch. Reviews would be appreciated again after I go through properly today :)02:52
*** b3rnard0_away is now known as b3rnard003:00
*** thorst has joined #openstack-ansible03:03
openstackgerritMerged openstack/openstack-ansible: Fix some typos  https://review.openstack.org/30592603:04
*** thorst has quit IRC03:11
*** yatin has joined #openstack-ansible03:11
*** sdake_ has quit IRC03:14
*** sdake has joined #openstack-ansible03:16
*** weezS has quit IRC03:19
*** weezS has joined #openstack-ansible03:19
*** marcoamorales has joined #openstack-ansible03:22
*** daneyon has joined #openstack-ansible03:24
*** yatin_ has joined #openstack-ansible03:24
*** fishcried1 has joined #openstack-ansible03:24
*** sdake has quit IRC03:27
*** izaakk_ has joined #openstack-ansible03:27
*** severion has joined #openstack-ansible03:28
*** Mudpuppy has quit IRC03:28
*** Mudpuppy has joined #openstack-ansible03:29
*** mattoliverau has joined #openstack-ansible03:30
*** v1k0d3n has quit IRC03:30
*** stevelle_ has joined #openstack-ansible03:31
*** weezS has quit IRC03:32
*** yatin has quit IRC03:32
*** fishcried has quit IRC03:32
*** stevelle has quit IRC03:32
*** daneyon_ has quit IRC03:32
*** izaakk has quit IRC03:32
*** matt6434 has quit IRC03:32
*** bogeyon18 has quit IRC03:32
*** spotz_zzz has quit IRC03:32
*** fishcried1 is now known as fishcried03:32
*** stevelle_ is now known as stevelle03:32
*** izaakk_ is now known as izaakk03:32
*** bogeyon18 has joined #openstack-ansible03:34
*** sdake has joined #openstack-ansible03:35
openstackgerritMerged openstack/openstack-ansible-os_swift: Expose and set a default fallocate_reserve value  https://review.openstack.org/28791203:36
*** sdake_ has joined #openstack-ansible03:39
*** sdake has quit IRC03:40
*** gonzalo2kx has joined #openstack-ansible03:48
*** gonzalo2kx has quit IRC04:00
openstackgerritMerged openstack/openstack-ansible: Add installation support for os_ironic  https://review.openstack.org/29377904:03
*** Mudpuppy has quit IRC04:03
*** b3rnard0 is now known as b3rnard0_away04:03
*** spotz_zzz has joined #openstack-ansible04:04
*** yatin_ has quit IRC04:05
*** thorst has joined #openstack-ansible04:11
*** hybridpollo has quit IRC04:14
*** thorst has quit IRC04:16
*** ChrisBenson has joined #openstack-ansible04:21
*** saneax_AFK is now known as saneax04:30
*** javeriak has joined #openstack-ansible04:30
*** yatin has joined #openstack-ansible04:37
*** ChrisBenson has quit IRC04:59
*** ChrisBenson has joined #openstack-ansible05:00
*** ChrisBenson1 has joined #openstack-ansible05:02
*** gonzalo2kx has joined #openstack-ansible05:03
*** ChrisBenson has quit IRC05:04
*** javeriak has quit IRC05:10
*** shausy has joined #openstack-ansible05:12
*** asettle has quit IRC05:12
*** thorst has joined #openstack-ansible05:13
*** asettle has joined #openstack-ansible05:13
*** admin0 has joined #openstack-ansible05:18
*** thorst has quit IRC05:20
*** fishcried1 has joined #openstack-ansible05:25
*** finchd has quit IRC05:26
*** fishcried has quit IRC05:27
*** larsks has quit IRC05:27
*** fishcried1 is now known as fishcried05:27
*** falanx has quit IRC05:28
*** jlvillal has quit IRC05:28
*** markvoelker has quit IRC05:28
*** admin0 has quit IRC05:30
*** jlvillal has joined #openstack-ansible05:34
openstackgerritAlexandra Settle proposed openstack/openstack-ansible: DOCS WIP: Adding Ironic configuration docs to Ansible install guide  https://review.openstack.org/30558605:35
*** larsks has joined #openstack-ansible05:36
*** markvoelker has joined #openstack-ansible05:36
*** finchd has joined #openstack-ansible05:36
*** Oku_OS-away has quit IRC05:36
asettleodyssey4me and cloudnull - questions and comments inline both in patch set 1 and 2. Patch set 3 is available. This is still a WIP with regards to the written content. Let's just try and get the technical content up and correct :)05:37
*** Oku_OS-away has joined #openstack-ansible05:39
*** falanx has joined #openstack-ansible05:42
*** LanceHaig has joined #openstack-ansible05:49
*** yatin has quit IRC06:00
*** mikelk has joined #openstack-ansible06:00
*** yatin has joined #openstack-ansible06:01
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Enable SSL termination for all services  https://review.openstack.org/27719906:05
*** sdake_ has quit IRC06:07
*** sdake has joined #openstack-ansible06:07
*** thorst has joined #openstack-ansible06:18
*** thorst has quit IRC06:25
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Enable SSL termination for all services  https://review.openstack.org/27719906:30
*** asettle has quit IRC06:38
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Enable SSL termination for all services  https://review.openstack.org/27719906:47
*** jokeric has joined #openstack-ansible06:53
*** fishcried1 has joined #openstack-ansible06:57
*** jlvillal has quit IRC06:58
*** LanceHaig has quit IRC06:58
*** LanceHaig has joined #openstack-ansible06:58
*** daneyon has quit IRC07:02
*** yatin has quit IRC07:02
*** yatin has joined #openstack-ansible07:03
*** mikelk has quit IRC07:04
*** fishcried has quit IRC07:04
*** fishcried1 is now known as fishcried07:04
*** larsks has quit IRC07:08
*** jlvillal has joined #openstack-ansible07:11
*** jamielennox is now known as jamielennox|away07:11
*** larsks has joined #openstack-ansible07:14
*** mikelk has joined #openstack-ansible07:14
*** javeriak has joined #openstack-ansible07:15
*** markvoelker has quit IRC07:16
*** thorst has joined #openstack-ansible07:23
*** thorst has quit IRC07:30
*** pcaruana has joined #openstack-ansible07:31
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Enable SSL termination for all services  https://review.openstack.org/27719907:48
matttcloudnull: bro seriously07:49
evrardjp:)07:50
*** Oku_OS-away is now known as Oku_OS07:50
evrardjpeverbody needs to eventually sleep you know cloudnull?07:50
*** admin0 has joined #openstack-ansible07:52
*** andymccr has quit IRC07:52
evrardjpand good morning everyone07:58
*** asettle has joined #openstack-ansible08:05
*** openstackgerrit has quit IRC08:08
*** asettle has quit IRC08:09
admin0good morning all08:10
*** asettle has joined #openstack-ansible08:11
*** jamielennox|away is now known as jamielennox08:11
*** asettle has quit IRC08:11
*** javeriak_ has joined #openstack-ansible08:11
evrardjpmorning admin008:11
admin0morning evrardjp08:11
*** andymccr has joined #openstack-ansible08:13
*** javeriak has quit IRC08:13
*** andymccr has quit IRC08:15
*** andymccr has joined #openstack-ansible08:16
admin0what shall I break today (TM)08:16
*** openstackgerrit has joined #openstack-ansible08:18
*** thorst has joined #openstack-ansible08:28
*** thorst has quit IRC08:36
*** jamielennox is now known as jamielennox|away08:43
*** fishcried has quit IRC08:44
openstackgerritDarren Birkett proposed openstack/openstack-ansible-galera_server: ensure mysql defaults file is sourced  https://review.openstack.org/30629808:49
*** asettle has joined #openstack-ansible08:49
mancdaz^ evrardjp odyssey4me08:49
*** asettle has quit IRC08:50
*** javeriak_ has quit IRC08:50
*** goretoxo has joined #openstack-ansible08:50
openstackgerritDarren Birkett proposed openstack/openstack-ansible-galera_server: ensure mysql defaults file is sourced  https://review.openstack.org/30629808:50
*** javeriak has joined #openstack-ansible08:55
*** tiagogomes has joined #openstack-ansible09:03
-openstackstatus- NOTICE: Gerrit is going to be restarted because is not processing new changes09:04
*** fishcried has joined #openstack-ansible09:05
*** fishcried has quit IRC09:07
*** sdake has quit IRC09:10
*** markvoelker has joined #openstack-ansible09:17
-openstackstatus- NOTICE: No jobs are being processed by gerrit and zuul . We are working to solve the problem, please be aware that no changes have been sent to the queue in the last hour, so you will need to recheck jobs for that period.09:18
*** ChanServ changes topic to "No jobs are being processed by gerrit and zuul . We are working to solve the problem, please be aware that no changes have been sent to the queue in the last hour, so you will need to recheck jobs for that period."09:18
*** markvoelker has quit IRC09:22
mancdazoh I was just about to ask why there are no jobs being processed by gerrit and zuul :)09:24
*** jokeric has quit IRC09:25
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: updated ironic playbook inclusion for tftp network  https://review.openstack.org/30365909:29
odyssey4meo/09:30
*** yatin has quit IRC09:32
*** thorst has joined #openstack-ansible09:33
javeriakodyssey4me hey09:35
javeriakwe cant backport directly from master to liberty anymore right?09:35
javeriaki should push a new patch09:35
*** neilus has quit IRC09:37
openstackgerritJaveria Khan proposed openstack/openstack-ansible: Update SSL key / cert distribution for Horizon  https://review.openstack.org/30631909:38
odyssey4mejaveriak yeah, you can't cherry-pick - so the convention is to create a new patch, but reference the source with something like 'This is a backport of <change-id>'09:38
javeriakodyssey4me yep just did that09:38
odyssey4mewhat you've done is perfectly fine09:39
javeriakcool09:39
*** thorst has quit IRC09:41
*** neilus has joined #openstack-ansible09:42
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server: ensure mysql defaults file is sourced  https://review.openstack.org/30629809:44
odyssey4memancdaz evrardjp ^ small change to the commit message to ensure that the bug referenced is appropriately handled09:44
*** jokeric has joined #openstack-ansible09:45
odyssey4memancdaz jmccrory fantastic find, thanks for taking the time to figure that one out!09:46
*** ChanServ changes topic to "Austin Design Summit Schedule: https://goo.gl/WSRblf || Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: https://goo.gl/tTmdgs"09:47
-openstackstatus- NOTICE: Zuul and gerrit are working normally now. Please recheck any jobs that may have been affected by this failure.09:47
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Fix configuration string for haproxy  https://review.openstack.org/30632409:48
*** tiagogomes has quit IRC09:51
admin0morning openstackstatus10:00
admin0morning odyssey4me10:01
admin0:D10:01
admin0i need a bit of help to understand what i did caused the build to fail: http://logs.openstack.org/27/303427/7/check/gate-openstack-ansible-dsvm-commit/71d3ce2/console.html10:04
openstackgerritSashi Dahal proposed openstack/openstack-ansible: make hostname,network and ip-address on all examples consistent  https://review.openstack.org/30342710:07
*** czunker has joined #openstack-ansible10:07
*** javeriak has quit IRC10:12
*** gfa_ is now known as gfa10:12
odyssey4meadmin0 I'm not sure if it's something you did, but this is the cause of the failure: http://logs.openstack.org/27/303427/7/check/gate-openstack-ansible-dsvm-commit/71d3ce2/console.html#_2016-04-14_20_39_53_33010:13
admin0thanks .. though we opted to use aio for documentation, i think the test system is configured to see aio1 .. so i resubmitted another one marking it as aio1 ..10:14
admin0the error message is not very descriptive ( at least to me )10:15
admin0on where/why it failed10:15
admin0i suspect hostname10:15
admin0question .. the patches that i do . is it possible for me to test it also ( like how this system is testing )10:16
admin0in that way, i would be able to know that it passes tests10:16
*** pjm6 has joined #openstack-ansible10:18
odyssey4mehmm, the AIO is not good to use for examples because it does funky networking and is not real-world at all in the way that it does things10:19
odyssey4meI haven't checked the patch yet though, will look later10:19
pjm6morning guys :D10:20
*** tiagogomes has joined #openstack-ansible10:23
evrardjpmorning pjm610:24
pjm6evrardjp: o/10:25
admin0morning10:27
odyssey4memattt so you did this on an Ubuntu host? and it looks like it made a change to another place in the role-requirements?10:29
matttodyssey4me: yeah i did it on a trusty instance10:29
matttodyssey4me: it just tacks on that 0.4.3 at the end for some reason10:29
matttwhen i changed the second open to 'w', it didn't do that10:29
matttso ?_O_?10:29
odyssey4meit does actually edit other places, because it sorts the dicts on output - so that's intentional10:29
*** pjm6 has quit IRC10:29
odyssey4meoh, weird10:29
matttyeah the update happens, which is good, just that trailing 0.4.3 which i don't understand10:30
odyssey4memattt interestingly, that does not happen on my mac10:31
matttodyssey4me: maybe python bug?10:31
odyssey4meok, well no harm - I'll do a quick patch as your suggestion is *more* correct10:31
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Adjust ansible-role-requirements-editer file open options  https://review.openstack.org/30635210:34
odyssey4memattt ^10:34
odyssey4meif that works for you, I'll consolidate the same patch into the backport10:35
matttodyssey4me: testing now10:36
*** johnmilton has joined #openstack-ansible10:37
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone: Remove venv activation code  https://review.openstack.org/28781210:38
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-security: Doc updates  https://review.openstack.org/30277610:39
*** thorst has joined #openstack-ansible10:39
*** pjm6 has joined #openstack-ansible10:39
*** jokeric has quit IRC10:42
*** palendae has quit IRC10:44
*** thorst has quit IRC10:45
odyssey4memattt updated https://review.openstack.org/30601610:46
*** jokeric has joined #openstack-ansible10:47
matttodyssey4me: LGTM10:48
*** pjm6 has quit IRC10:48
odyssey4methanks for validating mattt :)10:49
matttodyssey4me: didn't really test the other bit ... since it was trying to submit stuff w/ git review :P10:50
matttcould be nice to add a flag that doesn't do that by default10:50
odyssey4meoh yeah, well it works for me - and seeing as I'm the only one that uses it right now I guess that's enough ;)10:50
matttthat's true :)10:50
matttbut if someone accidentally runs it then it creates a bit of a mess10:51
odyssey4mewell, for it to actually work they would have to have git review configured and the right gerrit creds10:51
odyssey4meand patches can always be abandoned10:51
odyssey4mebut yeah, I've been thinking that maybe some of these scripts should rather be put into a seperate repo10:52
*** admin0 has quit IRC10:52
openstackgerritMerged openstack/openstack-ansible-security: Doc updates  https://review.openstack.org/30277610:52
*** admin0 has joined #openstack-ansible10:53
matttodyssey4me: probably ok to leave for the mean time, like you say it's not the end of the world if someone generates a few reviews10:53
evrardjpodyssey4me I think it should be in another repo yeah10:53
evrardjpdeployers don't need to see the tools not relevant to the product IMO10:54
evrardjpbut it's painful to configure new repos etc, so I guess you can leave it like that for now, because that's what we always did right?10:55
odyssey4mefor now it's fine, but perhaps in the newton timeframe we should split some of this stuff out - it's not crucial though, there's no real driving need10:56
*** palendae has joined #openstack-ansible10:56
openstackgerritMatt Thompson proposed openstack/openstack-ansible-security: Add ability to enable unattended upgrades  https://review.openstack.org/30409610:58
*** asettle has joined #openstack-ansible10:59
*** asettle has quit IRC10:59
*** electrofelix has joined #openstack-ansible11:11
*** markvoelker has joined #openstack-ansible11:18
openstackgerritMerged openstack/openstack-ansible-os_keystone: Remove venv activation code  https://review.openstack.org/28781211:18
*** pjm6 has joined #openstack-ansible11:19
*** schwicht has joined #openstack-ansible11:21
*** xek has joined #openstack-ansible11:22
*** markvoelker has quit IRC11:24
*** schwicht has quit IRC11:25
*** Bofu2U has quit IRC11:28
admin0odyssey4me: it was the hostname change .. now the new patch fixed it :)11:28
admin0thanks for helping how to figure out11:29
*** schwicht has joined #openstack-ansible11:31
*** thorst has joined #openstack-ansible11:37
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Switch defaults/tests to use master branch  https://review.openstack.org/30637311:39
*** fignew has quit IRC11:46
*** dmsimard has quit IRC11:46
*** dmsimard has joined #openstack-ansible11:47
*** retreved has joined #openstack-ansible11:51
*** openstack has quit IRC11:51
*** openstack has joined #openstack-ansible11:54
*** asettle has joined #openstack-ansible11:59
*** asettle has quit IRC12:00
*** skamithi14 has joined #openstack-ansible12:04
*** weshay has joined #openstack-ansible12:05
*** javeriak has joined #openstack-ansible12:06
*** skamithi13 has quit IRC12:06
*** pjm6 has quit IRC12:07
*** tiagogomes has quit IRC12:09
*** gfa is now known as gfa_12:11
*** gfa_ is now known as gfa12:12
*** Oku_OS is now known as Oku_OS-away12:12
*** woodard has joined #openstack-ansible12:13
*** woodard has quit IRC12:14
*** busterswt has joined #openstack-ansible12:15
*** woodard has joined #openstack-ansible12:16
*** markvoelker has joined #openstack-ansible12:16
*** tiagogomes has joined #openstack-ansible12:16
*** saneax is now known as saneax_AFK12:19
*** schwicht has quit IRC12:26
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Adding modularity to keepalived configuration  https://review.openstack.org/30597112:28
*** busterswt has quit IRC12:28
mhaydenmorning12:29
evrardjpmorning mhayden12:29
*** javeriak_ has joined #openstack-ansible12:29
*** tiagogomes has quit IRC12:30
*** tiagogomes has joined #openstack-ansible12:30
*** fignew has joined #openstack-ansible12:30
*** javeriak has quit IRC12:31
*** goretoxo has quit IRC12:31
*** busterswt has joined #openstack-ansible12:32
*** keedya has joined #openstack-ansible12:34
*** retreved_ has joined #openstack-ansible12:41
*** keedya has quit IRC12:43
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Adjust ansible-role-requirements-editor file open options  https://review.openstack.org/30635212:43
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Adjust ansible-role-requirements-editor file open options  https://review.openstack.org/30635212:45
*** retreved has quit IRC12:45
*** goretoxo has joined #openstack-ansible12:46
*** Oku_OS-away is now known as Oku_OS12:49
*** mgoddard__ has joined #openstack-ansible12:51
*** mgoddard_ has quit IRC12:54
*** rohanp has joined #openstack-ansible12:54
*** rohanp_ has joined #openstack-ansible12:59
*** Bjoern_ has joined #openstack-ansible13:00
*** Bjoern_ is now known as Bjoern_zZzZzZzZ13:00
*** Bjoern_zZzZzZzZ is now known as Bjoern_13:02
*** automagically has joined #openstack-ansible13:04
automagicallymorning all13:04
*** admin0 has quit IRC13:08
*** b3rnard0_away is now known as b3rnard013:13
*** goretoxo has quit IRC13:15
odyssey4mehowdy automagically - would you mind taking a peek at https://review.openstack.org/306298 ? I'd like to try and get that resolved asap13:17
odyssey4methere's a bunch of investigation info in the referenced bug if you're curious13:17
automagicallyI followed most of the discussion yesterday13:18
automagicallyodyssey4me: Have a second to give me some more insight into your comments here? https://review.openstack.org/#/c/306028/13:19
*** joker__ has quit IRC13:20
*** sdake has joined #openstack-ansible13:20
*** sdake_ has joined #openstack-ansible13:22
*** sdake has quit IRC13:26
odyssey4meautomagically I'm not sure of the best approach, but an option could be to only set pip_links if the group repo_all has more than 0 members?13:27
automagicallyThat results in different functionality. I have 0 members in that group, since my repo is centralized13:28
automagicallyBut I still have links and still want to lockdown pip13:28
odyssey4mepart of the issue is that the pip_lock_down role is included as a meta dep by almost all the roles, so even with your patch I think it might get set regardless13:28
automagicallyYep, except for galera and rabbitmq interestingly where its done at the playbook level13:29
automagicallyI’m trying to build out a centralized Keystone and Horizon infra for a multi-region install13:29
odyssey4mehmm, ok - I see that none of the galera_server deps include the lock down13:30
automagicallyExactly13:30
automagicallyNor do the rabbit ones to my knowledge13:30
odyssey4meyep, confirmed13:31
automagicallyGiven my use case, I’d say that relying upon the openstack_repo var is probably the most sane path forward13:31
automagicallyAnd it is backware compatible for all existing operators/deployers13:31
automagicallyWithout placing such a hard dependency on the inventory13:31
odyssey4mehmm, ok so https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/hosts.yml#L62-L63 will always set the links - and we're ok with that13:32
odyssey4methen you just override the repo url13:32
automagicallyExactly13:32
odyssey4meand yeah, it needs to not rely on a repo_server being in the inventory13:33
automagicallyYou got it, that’s what I’m going for there13:33
odyssey4methen actually I think that perhaps your approach is right13:33
automagicallyWe’ve got inventory group conditions spread all over the place. I understand why, but they can be very vague, when we already have more fine-grained vars implemented13:33
automagicallyI’ll update the patch to rely upon openstack_repo_url13:34
automagicallyAppreciate your time odyssey4me13:34
odyssey4meI'm trying to think if there's a case where pip_links would be set, but the repo server would not be around13:34
automagicallyYes, that is my case13:34
automagicallywell, from an inventory standpoint the repo_server isn’t around13:34
odyssey4meyeah13:34
automagicallyFrom an openstack_repo_url standpoint it is13:34
odyssey4memaybe the conditional should be something like if (groups['repo_all'] | length) >= 1) or (openstack_repo_url is defined) ?13:35
odyssey4mewhy was that conditional there in the first place - it's a little odd13:36
automagicallyAgreed, no idea why it was there13:37
odyssey4meit came in via https://github.com/openstack/openstack-ansible/commit/6f443d69713723810a448a9c83cdf603c07352d013:37
odyssey4mewe should find out from cloudnull why, because the commit doesn't explain it13:38
automagicallyBecause of this, we shouldn’t need to test the repo_all host group in the condition: https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/hosts.yml#L3813:38
odyssey4meit seems silly to perpetuate an edge case for no reason - if that's there for a situation we don't support then it gives you more freedom to change the conditional to support your specific use-case, and your commit can explain the use-case13:38
*** jthorne has joined #openstack-ansible13:39
automagicallyAh, this gives a hint to why: https://github.com/openstack/openstack-ansible/commit/6f443d69713723810a448a9c83cdf603c07352d0#diff-5e3fc411764d3375c98601ca2203fae5L3313:42
automagicallyAnd Galera was never locking down pip13:43
odyssey4meah, it was an early IRR13:43
odyssey4mewe handle things differently in the other roles13:43
odyssey4medoes the handling method in the other roles cause trouble for your use-case?13:44
automagicallyThe existing method like: https://github.com/openstack/openstack-ansible-os_cinder/blob/master/meta/main.yml#L33 would work fine if applied to galera_server and rabbitmq13:45
automagicallyHappy to abandon my current patch and submit those changes as an alternative13:45
odyssey4mehmm, the trouble with that method is that there is no developer_mode for those roles13:46
*** pjm6 has joined #openstack-ansible13:46
automagicallyThat’s fine, than it just becomes a simple dep with no condition13:47
automagicallySorry, I should have been more clear with my reference13:47
odyssey4meperhaps they should key off from whether openstack_repo_url is defined ?13:47
automagicallyThey could, but ultimately, that’s a side effect once you get to the role level. The only thing that really matters to the pip lockdown role is the value of pip_links13:49
automagicallyWhich is why I started where I did13:49
automagicallyIt just so happens that the default value of pip_links references openstack_repo_url13:49
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Improved documentation about LVM overwrite behaviour  https://review.openstack.org/30643613:49
odyssey4meyeah, inside the role pip_links is the key thing to key off from - agreed13:49
openstackgerritMerged openstack/openstack-ansible-galera_server: ensure mysql defaults file is sourced  https://review.openstack.org/30629813:50
odyssey4memancdaz ^ can you please backport https://review.openstack.org/306298 to stable/mitaka and liberty?13:50
odyssey4meautomagically ok I think it makes sense to make the change in the role and key off of pip_links, then adjust the play to remove the lock down role execution13:51
odyssey4methat makes a little more sense, and keeps things a little more consistent13:51
odyssey4mein fact it may make better sense to key all the role's use as meta deps based on pip_links ?13:52
*** jayc__ has joined #openstack-ansible13:52
automagicallyOr hide that logic within pip_lockdown role itself13:53
automagicallyAnd let the roles merrily depend on it all they want, but the role is a no-op when pip_links is empty13:53
odyssey4meyeah, I'm sort-of a fan of having pip_lock_down be a part of pip_install and the lock down and config bits only get implemented if there actually are pip_links13:54
odyssey4methat would mean only having one role dep and it's smarter13:54
odyssey4meand it handles all the pip things13:54
odyssey4mewhich to me seems more logical13:54
*** woodard has quit IRC13:56
*** pjm6 has quit IRC13:57
*** ametts has joined #openstack-ansible13:59
automagicallyCool and leads to quite a bit of simplicity13:59
automagicallyAnd also feels much more logical13:59
automagicallyIt may severely complicate the developer_mode logic though, I’ll have to think through it a bit more14:00
*** javeriak_ has quit IRC14:01
automagicallyAFK for a few minutes14:01
*** javeriak has joined #openstack-ansible14:01
*** schwicht has joined #openstack-ansible14:05
cloudnullmorning14:05
evrardjpmorning cloudnull14:05
cloudnullmattt evrardjp sleep is for the weak :p14:05
evrardjp:D14:05
evrardjpI've started to look at your haproxy change14:06
evrardjpto enable the ssl termination14:06
cloudnullcool14:06
evrardjpthere are some quirks right now, but I'd like to work with you14:06
evrardjpit's really a good thing14:06
*** Mudpuppy has joined #openstack-ansible14:06
evrardjpI like it14:06
cloudnullodyssey4me: what change needs explaining ?14:06
cloudnullindeed. lets do it14:07
openstackgerritDarren Birkett proposed openstack/openstack-ansible: ensure mysql defaults file is sourced  https://review.openstack.org/30645014:11
*** sigmavirus24_awa is now known as sigmavirus2414:12
odyssey4mecloudnull I think we managed to decipher it ;)14:14
odyssey4mecloudnull evrardjp yeah, agreed - those changes look good on paper but I haven't had the chance to test yet14:14
cloudnullbrushed up on your cloudnullise ?14:14
odyssey4mecloudnull yeah, I think that automagically is learning to speak cloudnull :p14:14
evrardjpodyssey4me I got issues with testing too, that's why we could work on this14:15
evrardjpjust ending a quick doc commit14:15
cloudnullI have it running within a 12 node lab from an api standpoint its working well.14:15
evrardjpI guess cloudnull is already doing lots of stuff on it anyway14:15
evrardjpcloudnull cool14:15
cloudnullhowever i have to go figure out how to make tempest happy14:15
odyssey4mecloudnull you may note my question regarding whether enabling the ssl header by default in all the roles has any side effects?14:16
evrardjpcloudnull I saw an issue on horizon: it seems there is no bind on port 80 anymore @haproxy, so no redirect happening from 80 to 44314:16
odyssey4mebasically if we can do away with the vars and just hard set it in the templates, if it works for ssl and non-ssl then I'd prefer that14:16
odyssey4meit's simpler and has less logic tied to it14:17
evrardjpI didn't test the nova console yet14:17
cloudnullodyssey4me:  thats working fine.14:17
odyssey4meevrardjp yep, there's actually already a patch in for that14:17
cloudnullyea i was looking for that patch14:17
cloudnulllogan-: made it14:17
evrardjpI'll check all of this afterwards14:17
odyssey4meevrardjp https://review.openstack.org/28949214:17
cloudnulli need to rebase that patch off of master instead of being dependent on my ssl patch so i can pull in those changes.14:17
odyssey4meit's based on the ssl termination patch14:18
cloudnullyea we should undo that14:18
cloudnullor squash the two14:18
odyssey4mecan that patch stand alone without your patch?14:18
odyssey4meyeah, I'm thinking just squash14:18
odyssey4meadd logan- as a co-author and squash14:18
cloudnullright now if you go to IP:80 it just tell you to gfy14:19
cloudnullw/ his bits it does proper redirection14:19
cloudnulllogan-:  do you mind if i squash your commit into the ssl patch and add you as a co-author ?14:20
*** jayc__ has quit IRC14:22
*** jayc_ has joined #openstack-ansible14:23
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: ensure mysql defaults file is sourced  https://review.openstack.org/30645014:24
*** jayc_ is now known as jayc14:25
*** spotz_zzz is now known as spotz14:28
*** gfa is now known as gfa_14:28
odyssey4mejmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr there are several stable/mitaka backports in the queue, I'd appreciate some reviews: https://review.openstack.org/#/q/project:%255Eopenstack/openstack-ansible.*+status:open+branch:stable/mitaka14:32
openstackgerritMerged openstack/openstack-ansible: Adjust ansible-role-requirements-editor file open options  https://review.openstack.org/30635214:32
automagicallyTaking a look14:34
* cloudnull looking now14:35
*** phalmos has joined #openstack-ansible14:37
*** gonzalo2kx has quit IRC14:37
*** electrofelix has quit IRC14:43
odyssey4mecloudnull with regards to https://review.openstack.org/306328 surely it won't deploy or do anything different unless the environment includes 'ironic-infra_hosts' and the virt_type is set to 'ironic' ?14:44
cloudnullit will not do anything different if those hosts are undefined. but that include give the impression that it works.14:45
cloudnullim not adverse to having it there later once the docs are online14:45
cloudnulland the network bits are worked out14:45
odyssey4meok, so we should rather focus on finalising master, then doing a combined backport with all the related patches included in one review?14:46
*** mgoddard has joined #openstack-ansible14:46
odyssey4meI think I prefer that option.14:46
*** mgoddard__ has quit IRC14:46
*** neilus has quit IRC14:47
cloudnulli think its fine to pull in the intial work but omit that include14:47
cloudnullthis is similar to how we did the ceilometer inclusion14:48
cloudnulland when there were tests we added it to the integrated build14:48
odyssey4meok, for now I'll hold it back as a combined backport will make more sense anyway14:49
*** Brew has joined #openstack-ansible14:51
*** Xu_Chao has joined #openstack-ansible14:51
*** electrofelix has joined #openstack-ansible14:51
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Added a documentation section for the cinder-horizon  https://review.openstack.org/30648214:52
cloudnull+114:53
evrardjp^ doc specialists would probably fix my english14:53
evrardjpok now14:53
*** gonzalo2kx has joined #openstack-ansible14:53
spotzevrardjp: did I hear my role? Which is funny as I'm a linux systems engineer:)14:54
evrardjpare you native english? that's good enough for me14:55
evrardjp:D14:55
*** weezS has joined #openstack-ansible14:55
spotzevrardjp: hehe14:56
evrardjpwhich proves engineers can do everything14:56
evrardjpthanks then spotz if you take the leap :D14:56
spotzevrardjp: Yep I've got it up. let me get a DB moving at work then I'll look14:57
evrardjpI'm definitely not in a hurry14:57
*** pjm6 has joined #openstack-ansible14:58
*** rohanp_ has quit IRC15:00
*** john has joined #openstack-ansible15:00
pjm6anyone used cinder with lvm?15:00
*** john is now known as h1nch15:00
pjm6i'm in doubt about the "cinderAZ_1" variable ( http://docs.openstack.org/developer/openstack-ansible/mitaka/install-guide/configure-cinder.html#configuring-cinder-to-use-lvm )15:01
pjm6this is the zone of the cinder or related with nova avaiability zone?15:01
evrardjpto make your life easier you can make them equal pjm615:02
evrardjpbut that's another topic15:02
pjm6if I didn't put that var, will go the default?15:02
pjm6I'm trying to put openfiler (to simulate a storage) comunnicating with cinder, in this case i wanted to share an iSCSI connection from openfiler to the cinder15:03
rromansevrardjp: do you mind if I change the wording a little bit and upload a new patch? or would you prefer comments?15:03
pjm6for this, the openfiler server, must be in the same storage network (it's not necessary have the container host) right?15:03
evrardjprromans go ahead15:03
rromansevrardjp: ok, thanks15:04
*** gonzalo2kx has quit IRC15:04
evrardjppjm6 technically you're not forced to be on the same network, as long as your backend is reachable15:04
evrardjprromans I thank you!15:05
evrardjppjm6 but you probably want that15:05
pjm6but in terms of "good practice" would be in being in the same network, and then all my volumes will be stored in that storage15:05
*** javeriak has quit IRC15:06
pjm6and if cinder "dies" i can reallocate volumes to a new agent and then I get rid of the SPOF15:06
evrardjpgood practice is to put your storage as close as possible to your consumption15:06
*** pcaruana has quit IRC15:06
evrardjppjm6 that's not that simple15:06
evrardjpone does not simply move a volume to a different cinder instance15:07
evrardjpif that's what you meant15:07
pjm6no no, what i understood (but could be wrong)15:07
pjm6are imagine15:07
evrardjpI didn't test that though15:07
pjm6I have two cinder agents right?15:07
pjm6if one of them dies, all the volumes managed by that agent "die" and all info will disappear from the openstack15:08
evrardjpI can imagine anything, so two cinder policemen seems doable15:08
evrardjpgo on15:08
pjm6but if I use a storage (with LVM controller) I can change in the database15:08
pjm6that the cinder agent that will maange that volume will be the other policemen that aren't sleeping xD15:09
pjm6*that is not sleeping (or dead in this case)15:09
logan-cloudnull: sure @ squash15:10
logan-was afk15:10
cloudnullcool.15:10
cloudnullthanks logan-15:10
pjm6cloudnull: o/15:10
evrardjppjm6 but you still have one backend storage? so how do you improve this reliability?15:10
cloudnullo/15:10
evrardjpcloudnull I'll wait for the squash then15:11
*** meteorfox has quit IRC15:11
*** Tebro has quit IRC15:11
*** serverascode has quit IRC15:11
*** etoews has quit IRC15:11
pjm6evrardjp: backend storage could be a physical storage or openfiler in this case, right?15:11
pjm6btw cloudnull I tested your modifications of vpnaas and haven't lucky :\15:11
*** sdake_ has quit IRC15:11
*** alextricity25 has quit IRC15:12
*** bryan_att has quit IRC15:12
evrardjpif you have openfiler, you plan to run cinder volume on it15:12
evrardjpI guess15:12
evrardjpI'm not aware of an openfiler backend15:12
*** jmccrory has quit IRC15:12
evrardjpfor cinder volumes15:12
evrardjpmaybe nfs backend?15:12
evrardjp<- not a cinder expert15:13
*** flwang has quit IRC15:13
*** erikwilson has quit IRC15:13
*** odyssey4me has quit IRC15:13
*** kong has quit IRC15:13
*** LiftedKilt has quit IRC15:13
*** haasn has quit IRC15:13
*** mhayden has quit IRC15:13
pjm6evrardjp: i'm not too, but i'm thinking in use a iSCSI connector15:13
pjm6http://docs.openstack.org/developer/openstack-ansible/mitaka/install-guide/configure-cinder.html#configuring-cinder-to-use-lvm15:13
pjm6iscsi_ip_address15:13
pjm6LVM_iSCSI15:14
*** phalmos has quit IRC15:16
*** pjm6_ has joined #openstack-ansible15:16
*** bustersw_ has joined #openstack-ansible15:16
*** mgoddard_ has joined #openstack-ansible15:16
*** v1k0d3n has joined #openstack-ansible15:17
*** odyssey4me has joined #openstack-ansible15:17
*** erikmwilson has joined #openstack-ansible15:17
*** electrofelix has quit IRC15:18
evrardjppjm6 I thought that LVM_iSCSI used LVM on the cinder volume host and exports logical volumes to the nova/libvirt nodes through iscsi.15:18
evrardjpnot the opposite15:18
*** LiftedKilt has joined #openstack-ansible15:18
*** kong has joined #openstack-ansible15:19
evrardjpbut you could use nfs backend maybe15:19
*** schwicht_ has joined #openstack-ansible15:19
*** sdake has joined #openstack-ansible15:20
*** mgoddard has quit IRC15:20
*** jayc has quit IRC15:20
evrardjphttp://docs.openstack.org/liberty/config-reference/content/NFS-driver.html15:20
*** javeriak has joined #openstack-ansible15:20
*** AndroUser2 has joined #openstack-ansible15:20
*** haasn has joined #openstack-ansible15:21
pjm6_yes evrardjp, i think is that15:21
pjm6_I was confusing :S15:21
*** rohanp has quit IRC15:21
*** matt6434 has joined #openstack-ansible15:21
pjm6_but yeah, that way of exporting from cinder to the nova/libvirt through iscsi dispends the use of storage node (after attributing)15:22
*** v1k0d3n has quit IRC15:22
*** flwang has joined #openstack-ansible15:22
*** v1k0d3n has joined #openstack-ansible15:22
*** meteorfox has joined #openstack-ansible15:23
*** etoews has joined #openstack-ansible15:23
*** alextricity25 has joined #openstack-ansible15:23
*** jlvillal_ has joined #openstack-ansible15:24
*** pjm6 has quit IRC15:25
*** schwicht has quit IRC15:25
*** ametts has quit IRC15:25
*** busterswt has quit IRC15:25
*** jokeric has quit IRC15:25
*** mikelk has quit IRC15:25
*** jlvillal has quit IRC15:25
*** falanx has quit IRC15:25
*** mattoliverau has quit IRC15:25
*** severion has quit IRC15:25
*** grumpycatt has quit IRC15:25
*** pabelanger has quit IRC15:25
*** nhadzter has quit IRC15:25
*** jlvillal_ is now known as jlvillal15:27
*** jlvillal is now known as Guest8106215:27
*** mhayden has joined #openstack-ansible15:28
*** falanx has joined #openstack-ansible15:29
*** ametts has joined #openstack-ansible15:29
*** jmccrory has joined #openstack-ansible15:30
openstackgerritRobb Romans proposed openstack/openstack-ansible: Added horizon documentation section for the cinder  https://review.openstack.org/30648215:30
*** klamath has joined #openstack-ansible15:30
*** pabelanger has joined #openstack-ansible15:30
*** phalmos has joined #openstack-ansible15:30
*** grumpycatt has joined #openstack-ansible15:30
*** pabelanger has quit IRC15:30
*** pabelanger has joined #openstack-ansible15:30
*** Tebro has joined #openstack-ansible15:31
*** nhadzter has joined #openstack-ansible15:31
*** phalmos has quit IRC15:32
openstackgerritRobb Romans proposed openstack/openstack-ansible: Added horizon documentation section for cinder  https://review.openstack.org/30648215:32
*** jayc has joined #openstack-ansible15:33
*** Xu_Chao has quit IRC15:36
*** evanstoner has joined #openstack-ansible15:37
evrardjpthanks rromans15:38
pjm6_well evrardjp i think i was thinking bad, that configuration will share an iscsi controller from storage node15:39
pjm6_(cinder) to the nova15:39
*** raddaoui has joined #openstack-ansible15:39
rromansevrardjp: welcome!15:39
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Added horizon documentation section for cinder  https://review.openstack.org/30648215:40
*** bryan_att has joined #openstack-ansible15:40
evrardjpquick reedit to explain I'm not re-writing default horizon doc :D15:41
evrardjpI think it's fine now, your text is really better15:41
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Improved logging for memcached (OSA calling part)  https://review.openstack.org/30516515:42
*** phalmos has joined #openstack-ansible15:43
*** skamithi14 has quit IRC15:44
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible: Improved documentation about LVM overwrite behaviour  https://review.openstack.org/30643615:45
*** skamithi13 has joined #openstack-ansible15:45
*** shausy has quit IRC15:47
*** sdake has quit IRC15:49
*** javeriak has quit IRC15:50
*** sdake has joined #openstack-ansible15:50
*** serverascode has joined #openstack-ansible15:50
*** git-harry has quit IRC15:50
*** javeriak has joined #openstack-ansible15:54
*** git-harry has joined #openstack-ansible15:54
*** gfa_ is now known as gfa15:58
mhaydencloudnull: did you mention something about doing OSA on OnMetal at some point? i was just talking with antonym about that16:00
*** mgoddard__ has joined #openstack-ansible16:01
*** git-harry has left #openstack-ansible16:01
mhaydeni coulda sworn you had a repo somewhere16:01
palendaemhayden: https://github.com/cloudnull/osa-multi-node-aio16:02
*** gfa is now known as gfa_16:03
*** gfa_ is now known as gfa16:03
mhaydenantonym: ^^16:04
cloudnullpalendae: +1 ^ thats it16:04
*** mgoddard_ has quit IRC16:04
antonymah, interesting16:05
antonymthanks, i'll poke around on it16:05
*** javeriak has quit IRC16:08
*** bustersw_ has quit IRC16:08
*** busterswt has joined #openstack-ansible16:08
*** jayc has quit IRC16:10
*** serverascode has quit IRC16:13
*** woodard has joined #openstack-ansible16:15
*** yarkot has quit IRC16:15
*** yarkot1 has quit IRC16:17
*** serverascode has joined #openstack-ansible16:17
cloudnullcool antonym PRs welcome :)16:19
antonymyeah, i got an aio spun up on a single node, but sounds like this might be a better way to go if i want to play around with networking16:20
cloudnullyes16:20
cloudnullim using OM IO to build a 14 node cloud16:21
cloudnullw/in KVM16:21
cloudnullits using haproxy for the LB i'd like to integrate with something like vyos however i've not done that yet16:21
openstackgerritTravis Truman (automagically) proposed openstack/openstack-ansible-pip_install: WIP/POC - Combine pip_install and pip_lockdown roles  https://review.openstack.org/30652916:23
antonymcool, yeah, just wanting to get up to speed on the latest and hack on it some16:23
*** bryan_att has quit IRC16:23
cloudnullthat'll get you there. its fake cloud :)16:23
cloudnullautomagically: +116:24
automagicallycloudnull and odyssey4me - https://review.openstack.org/306529 came from the chat that Jesse and I were having earlier about a better solution to the problem highlighted by https://review.openstack.org/#/c/306028/16:24
antonymperfect16:24
*** jayc has joined #openstack-ansible16:24
*** bryan_att has joined #openstack-ansible16:25
odyssey4mepjm6_ if you're using an NFS, iSCSI or Ceph back-end for cinder, then all the volume service does is inform the hypervisor of the config it needs to use - therefore the service only needs to be running when a volume is created/deleted and when it's attached/detached to an instance.16:28
pjm6_yes odyssey4me, that was i was thinking (if not in mistake)16:28
pjm6_but my doubt is, the iSCSI back-end is of the cinder service, or could be an existent iSCSI that i have, for example, the iqn of the openfiler16:29
odyssey4mepjm6_ yes you can use an existing iSCSI storage solution if openstack has a driver for it16:30
*** skamithi13 has quit IRC16:30
pjm6_hmm I see16:30
pjm6_here in volume drivers16:30
pjm6_http://docs.openstack.org/liberty/config-reference/content/section_cinder.conf.html16:30
pjm6_i have several storages (but as i'm using a server for testing propuses) maybe the solution is attach a iSCSI volume to the cinder host16:31
pjm6_and then share that LVM_iSCSI to the nova, right odyssey4me?16:31
pjm6_sorry if i'm confusing16:31
*** jayc_ has joined #openstack-ansible16:32
*** jayc_ has quit IRC16:33
*** jayc has quit IRC16:34
*** cfarquhar has quit IRC16:37
*** AndroUser2 has quit IRC16:39
*** Oku_OS is now known as Oku_OS-away16:41
odyssey4mepjm6_ lvm-backed cinder is not meant for production at all16:43
*** cfarquhar has joined #openstack-ansible16:43
*** cfarquhar has quit IRC16:43
*** cfarquhar has joined #openstack-ansible16:43
*** stevelle has quit IRC16:43
pjm6_really? :O so whats used in production?16:43
odyssey4mepjm6_ you should be looking here: http://docs.openstack.org/liberty/config-reference/content/section_volume-drivers.html16:43
odyssey4methere's a NetApp driver, seeing as you have a filer16:43
cloudnullif anyone has time to eval https://review.openstack.org/#/c/304840/ i'd appreciate it , it'd be good to get that rolling for newton16:44
cloudnullalso odyssey4me had you seen https://review.openstack.org/#/c/304840/15/tox.ini ?16:44
cloudnullI added tests since the last patch set16:44
automagicallyWill take another look at the virtualenv work cloudnull16:45
pjm6_Yes odyssey4me i see that, but i didn't know the netapp driver16:45
cloudnullthanks automagically16:45
pjm6_I will see that if do16:45
*** stevelle has joined #openstack-ansible16:45
odyssey4mecloudnull tests for the cli tooling? heh, that's funny16:45
cloudnullyup16:45
cloudnullhttps://review.openstack.org/#/c/304840/15/scripts/openstack-ansible.sh@3416:45
odyssey4meI haven't had a chance yet and am behind on my objectives for the day16:45
*** tiagogomes has quit IRC16:46
automagicallyI’m behind on my objectives for the day/week/month/year, but its cloudnull :)16:46
cloudnullthanks automagically :)16:46
automagicallyAnd I can’t wait to move to Ansible 2+16:46
pjm6_ NetApp unified driver is a generic driver for multiple existing storages?16:46
pjm6_odyssey4me:16:46
cloudnullso this tool came from my 2.x work16:47
automagicallyThanks to 1.9 I’m still stuck dealing with this: https://bugs.launchpad.net/openstack-ansible/+bug/155697516:47
openstackLaunchpad bug 1556975 in openstack-ansible "Keystone playbook fails on remote venv checksum when accessed through a proxy" [Wishlist,Confirmed]16:47
odyssey4mepjm6_ it's a driver for NetApps, if you have a NetApp16:47
openstackgerritTravis Truman (automagically) proposed openstack/openstack-ansible-pip_install: WIP/POC - Combine pip_install and pip_lockdown roles  https://review.openstack.org/30652916:47
cloudnulli was testing osa w/ and switching ansible versions to ensure it works in 1.9.x | 2.1.x16:47
odyssey4mepjm6_ if you have a NetApp, or more than one NetApp, then you setup the config as described in http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-cinder.html#configuring-cinder-to-use-a-netapp-appliance as an example... NetApp has their own documentation for the setup of the NetApp storage itself16:48
pjm6_yeah, thanks odyssey4me. I will see16:48
pjm6_in the future I will use a storage (dell equallogic)16:48
cloudnullautomagically: I wonder if thats good to go in 2.1.x(devel)16:49
pjm6_but for testing, as I don't have other storage for testing, was trying to use another that used iscsi16:49
odyssey4mepjm6_ if you're not using cinder with the lvm driver, then you can switch cinder-volume to run in containers on your infra hosts and you can have multiple back-ends per cinder-volume service16:49
cloudnullyes it seems so16:50
cloudnullhadn't read the whole bug16:50
cloudnullso # openstack-ansible -rtv git/devel ...16:50
*** gfa is now known as gfa_16:50
*** gfa_ is now known as gfa16:50
*** tiagogomes has joined #openstack-ansible16:51
odyssey4meautomagically hmm, so the issue for that bug doesn't pertain to rpc-repo specifically but instead your general upstream repo ?16:52
odyssey4meie you have a repo for multiple regions, but each region accesses it through a proxy?16:52
*** stevelle has quit IRC16:52
*** jayc has joined #openstack-ansible16:52
pjm6_odyssey4me: thanks :) will try to research more about it16:53
*** stevelle has joined #openstack-ansible16:53
pjm6_odyssey4me: (sorry for the questions) imagine that I have a storage node (bare-metal)16:53
pjm6_and I connect the /cinder-volumes16:53
pjm6_to an storage via iSCSI16:53
cloudnullautomagically: http://cdn.pasteraw.com/3udz2rczsc5egcs3lex2qpvf7ugcm87 that'd at least get you out from dealing with the proxy mess.16:53
pjm6_its possible to use that volume in the cinder for providing volumes ??16:54
*** rohanp has joined #openstack-ansible16:54
odyssey4mepjm6_ for testing you may just wish to use the NFS driver16:55
odyssey4meset up an NFS server, then point cinder at it16:55
odyssey4mepjm6_ yes, that's possible - but then that server is a single point of failure16:55
odyssey4mepjm6_ the point of using networked storage and not a server is that networked storage typically can be upgraded and maintained without any down time - a server is not like that16:56
pjm6_yes that true16:57
pjm6_and then its better use the backend of a storage for better results and avoid SPOF?16:57
odyssey4mepjm6_ yes, for production you really, really want to avoid any SPOF - and especially anything that may inhibit your ability to upgrade without causing data plane down time16:58
odyssey4mepjm6_ so ideally you should have a shared storage back-ends for nova (instance OS/ephemeral disks), glance (images) and cinder (volumes)16:59
odyssey4meautomagically if we switch up the use of the uri module and rather use http://docs.ansible.com/ansible/get_url_module.html does it work around the problem?17:00
odyssey4meit'll be more cumbersome, but if it solves the problem then I think we should do that until we're using Ansible 2.x17:01
*** permalac has joined #openstack-ansible17:01
*** Guest81062 is now known as jlvillal17:04
*** daneyon has joined #openstack-ansible17:05
*** daneyon_ has joined #openstack-ansible17:06
*** tiagogomes has quit IRC17:06
*** tiagogomes has joined #openstack-ansible17:09
*** mgoddard__ has quit IRC17:10
*** mgoddard has joined #openstack-ansible17:10
*** daneyon has quit IRC17:11
pjm6_yes odyssey4me17:13
pjm6_but something that are confusing, is why there is not a generic driver for iSCSI ?17:13
pjm6_that would not be  able to manage almost every17:14
pjm6_storages?17:14
odyssey4mepjm6_ because no-one has developed one :p17:14
pjm6_odyssey4me: ok, fair enough xD17:14
pjm6_btw: between an NFS share or an Dell EqualLogic17:16
pjm6_what would be better? If we use a driver of a specific storage would have more advantage (i think)17:16
pjm6_well i have to run, back later, and thanks :)17:17
*** Mudpuppy_ has joined #openstack-ansible17:17
odyssey4meon consideration - iSCSI is a connection protocol, whereas the driver do more than just connect17:17
odyssey4methey connect to the service API to do things like snapshot, create volumes, etc - iSCSI can't do those things17:18
odyssey4methe specific storage drivers are designed to take advantage of the API's available in the storage units which do those things17:18
pjm6_hmm yeah that make senses17:19
pjm6_and do you know any operating system that had driver in cinder? As far as I saw only exists hardware17:19
*** git-harry has joined #openstack-ansible17:20
odyssey4meif you're using a server to simulate a storage environment, then I'd suggest using the NFS/ZFS driver17:20
pjm6_Ok thanks odyssey4me :)17:21
*** daneyon has joined #openstack-ansible17:21
odyssey4mejust setup a server with an NFS export for each service, then consume those17:21
*** stevelle_ has joined #openstack-ansible17:21
*** b3rnard0 is now known as b3rnard0_away17:22
pjm6_and when moving to the storage, I will just use the default =)17:22
pjm6_gonna try that17:22
*** raddaoui has quit IRC17:22
*** jmccrory has quit IRC17:22
*** mhayden has quit IRC17:22
odyssey4meyeah, when going for production use the driver that most fits with the real storage back-end17:22
pjm6_the problem would be to validate, because without hardware i can "simulate" if my configuration is working17:23
*** v1k0d3n_ has joined #openstack-ansible17:24
odyssey4mepjm6_ some of the storage vendors have virtual appliances which you can use, and some have evaluation versions so that you can use them for a limited time17:24
odyssey4meI would suggest speaking to your storage vendor to find out whether there are any options like that17:24
pjm6_hmmm nice :D will see17:25
*** woodard has quit IRC17:25
pjm6_its Dell17:25
pjm6_I will contact them17:25
*** andymccr_ has joined #openstack-ansible17:26
*** woodard has joined #openstack-ansible17:26
*** sdake_ has joined #openstack-ansible17:26
*** daneyon_ has quit IRC17:26
*** permalac has quit IRC17:26
*** stevelle has quit IRC17:26
*** jayc has quit IRC17:26
*** cfarquhar has quit IRC17:26
*** bryan_att has quit IRC17:26
*** serverascode has quit IRC17:26
*** stevelle_ is now known as stevelle17:26
*** phalmos_ has joined #openstack-ansible17:27
*** phalmos has quit IRC17:27
*** ametts has quit IRC17:27
*** etoews has quit IRC17:27
*** v1k0d3n has quit IRC17:27
*** h1nch has quit IRC17:27
*** Mudpuppy has quit IRC17:27
*** markvoelker has quit IRC17:27
*** openstackgerrit has quit IRC17:27
*** andymccr has quit IRC17:27
*** marcoamorales has quit IRC17:27
*** sdake has quit IRC17:29
pjm6_cya odyssey4me, and once again thanks very much :)17:29
*** mhayden has joined #openstack-ansible17:30
*** pjm6_ has quit IRC17:30
automagicallyodyssey4me and I have been talking about OSA cross-project needs lately so I decided to start this: https://etherpad.openstack.org/p/openstack-ansible-cross-project17:30
automagicallyConsider it a dumping ground for items that our community wants to see discussed in the weekly Cross Project meeting: https://wiki.openstack.org/wiki/Meetings/CrossProjectMeeting17:31
odyssey4meautomagically thanks for taking the initiative17:32
automagicallyEh, just making a list ;)17:32
*** Bjoern_ is now known as Bjoern_zZzZzZzZ17:35
*** markvoelker has joined #openstack-ansible17:35
*** jayc has joined #openstack-ansible17:36
*** etoews has joined #openstack-ansible17:36
*** jmccrory has joined #openstack-ansible17:37
*** tiagogomes has quit IRC17:37
*** yarkot has joined #openstack-ansible17:37
*** sdake_ has quit IRC17:38
*** cfarquhar has joined #openstack-ansible17:39
*** cfarquhar has quit IRC17:39
*** cfarquhar has joined #openstack-ansible17:39
*** raddaoui has joined #openstack-ansible17:42
*** sdake has joined #openstack-ansible17:43
*** eil397 has joined #openstack-ansible17:44
*** ametts has joined #openstack-ansible17:44
*** daneyon has quit IRC17:44
*** openstackgerrit has joined #openstack-ansible17:48
*** Bjoern_zZzZzZzZ is now known as Bjoern_17:50
*** yarkot1 has joined #openstack-ansible17:51
*** serverascode has joined #openstack-ansible17:54
*** javeriak has joined #openstack-ansible17:56
*** raginbajin has quit IRC17:57
*** javeriak_ has joined #openstack-ansible17:59
*** javeriak has quit IRC18:01
*** jayc has quit IRC18:01
*** bryan_att_ has joined #openstack-ansible18:01
*** raginbajin has joined #openstack-ansible18:02
*** bryan_att_ is now known as bryan_att18:04
*** KLevenstein has joined #openstack-ansible18:05
*** zhangjn has quit IRC18:05
*** czunker has quit IRC18:06
*** h1nch has joined #openstack-ansible18:08
*** zhangjn has joined #openstack-ansible18:12
openstackgerritMerged openstack/openstack-ansible-galera_server: Reorganize test playbooks  https://review.openstack.org/30309018:12
automagicallyWoot ^ Only took ~20 rechecks18:16
openstackgerritRobb Romans proposed openstack/openstack-ansible: Added horizon documentation section for cinder  https://review.openstack.org/30648218:16
jmccroryhaha18:17
odyssey4mejmccrory automagically heh, yeah - now that https://review.openstack.org/306298 is merged other patches are going in much more reliably18:17
odyssey4meafk for a bit18:18
jmccroryyeah it's great, empty queue for the role finally18:18
cloudnullhahaha18:18
cloudnulltime to make moar!18:18
jmccroryheh18:19
cloudnullso tempest on the gate is mad with the ssl changes, 12 physical nodes get along just fine.18:19
* cloudnull may have to dig into some of the test assumptions18:19
*** gfa is now known as gfa_18:19
*** marcoamorales has joined #openstack-ansible18:20
*** zhangjn has quit IRC18:21
openstackgerritMerged openstack/openstack-ansible: ensure mysql defaults file is sourced  https://review.openstack.org/30645018:21
*** asettle has joined #openstack-ansible18:23
*** eil397 has quit IRC18:26
automagicallyjmccrory: Would appreciate your thoughts on https://review.openstack.org/#/c/306529/ if you have the time18:26
automagicallyAnd I bet Kevin would appreciate you taking a gander at https://review.openstack.org/#/c/304840/ as well18:27
jmccrorysure, already liking that description18:27
*** asettle has quit IRC18:27
*** rETROpunK has quit IRC18:35
*** basilAB has quit IRC18:35
*** basilAB has joined #openstack-ansible18:36
*** rETROpunK has joined #openstack-ansible18:37
openstackgerritSteve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing  https://review.openstack.org/30488718:43
*** sdake_ has joined #openstack-ansible18:49
*** pjm6 has joined #openstack-ansible18:51
*** sdake has quit IRC18:52
*** b3rnard0_away is now known as b3rnard018:53
*** schwicht_ has quit IRC18:56
*** ametts has quit IRC18:56
stevelleany ideas for why all requests to the service end up with 404s? https://review.openstack.org/#/c/304887/18:58
stevelleeither apache access or gnocchi service (when mod_wsgi mode is disabled) logs the 404 so it is reaching the host. I can verify by running this by hand.18:59
*** javeriak_ has quit IRC19:02
*** dolphm has quit IRC19:03
stevelleno other info in the logs except some keystone middleware warnings19:03
*** dolphm has joined #openstack-ansible19:03
*** hybridpolio has joined #openstack-ansible19:07
jmccrorystevelle: get a 200 if you hit /v1/status ?19:08
jmccroryhttp://gnocchi.xyz/rest.html#status19:09
stevellejmccrory: every url is 404 (unless I skip sending a token, which gives 401)19:09
*** sdake_ has quit IRC19:10
odyssey4meFYI Ubuntu Xenial experimental jobs - proposed review: https://review.openstack.org/30656419:10
stevelleall of / and /v1 and /v1/status and ...19:10
jmccroryhmm strange, i'll try a local deploy in a bit19:10
odyssey4mealso FYI prposed review for higher timeout for the roles: https://review.openstack.org/30657319:10
stevellejmccrory: I'm running an integrated (see ext/bootstrap.sh) now to verify that is working still but it did last I checked.19:11
*** Mudpuppy_ is now known as Mudpuppy19:12
*** sdake has joined #openstack-ansible19:12
*** mgoddard has quit IRC19:14
*** keedya has joined #openstack-ansible19:15
*** asettle has joined #openstack-ansible19:18
jmccrorystevelle: oh, your api-paste is missing a few things https://github.com/openstack/gnocchi/blob/master/etc/gnocchi/api-paste.ini19:18
*** v1k0d3n__ has joined #openstack-ansible19:19
*** Mudpuppy_ has joined #openstack-ansible19:19
*** markvoelker_ has joined #openstack-ansible19:19
prometheanfireosa should probably use the new ansible19:21
prometheanfirefor sec reasons19:22
*** stevelle_ has joined #openstack-ansible19:22
prometheanfire1.9.6 was released19:22
*** raddaoui_ has joined #openstack-ansible19:22
*** asettle has quit IRC19:22
*** serverascode has quit IRC19:22
*** raddaoui has quit IRC19:22
*** yarkot has quit IRC19:22
*** stevelle has quit IRC19:22
*** v1k0d3n_ has quit IRC19:22
*** Mudpuppy has quit IRC19:22
*** stevelle_ is now known as stevelle19:22
*** markvoelker has quit IRC19:22
*** raddaoui_ is now known as raddaoui19:23
prometheanfireCVE-2016-309619:23
*** Mudpuppy_ is now known as Mudpuppy19:23
*** schwicht has joined #openstack-ansible19:24
*** serverascode has joined #openstack-ansible19:25
odyssey4meprometheanfire I'm not sure that it matters directly to OSA, especially considering that deployers can choose their own versions easily - but thanks for the pointer. We tried updating to 1.9.5 but it was busted.19:27
prometheanfireah, boo19:27
prometheanfireI thought we hard coded to 1.9.419:27
odyssey4meprometheanfire all it takes is one env var to change that: https://github.com/openstack/openstack-ansible/blob/master/scripts/bootstrap-ansible.sh#L2519:28
*** woodard has quit IRC19:28
prometheanfireok19:28
*** woodard has joined #openstack-ansible19:29
*** schwicht has quit IRC19:29
openstackgerritMatthew Thode proposed openstack/openstack-ansible: update ansible version to 1.9.6 for CVE-2016-3096  https://review.openstack.org/30660919:33
prometheanfireodyssey4me: well, guess we can see if it works at least19:33
odyssey4meprometheanfire ok, I'll update that patch - I was just writing a release note19:34
prometheanfireupdate that patch?19:34
odyssey4meyeah, it needs a release note too - and I'm half-way done writing one19:34
prometheanfireah19:34
prometheanfireya, haven't submitted osa code in a while :P19:35
*** mhayden has quit IRC19:35
*** schwicht has joined #openstack-ansible19:36
*** mhayden has joined #openstack-ansible19:37
odyssey4meprometheanfire do you have a mitre link for the CVE?19:37
prometheanfiresec19:38
prometheanfireno, not here either https://bugzilla.redhat.com/show_bug.cgi?id=132292519:38
openstackbugzilla.redhat.com bug 1322925 in vulnerability "CVE-2016-3096 ansible: Code execution vulnerability in lxc_container" [High,Assigned] - Assigned to security-response-team19:38
odyssey4meyeah, that's the only decent one I could fine19:39
odyssey4mehow annoying19:39
prometheanfirejust the placeholder19:39
*** ametts has joined #openstack-ansible19:42
*** KLevenstein has quit IRC19:47
*** KLevenstein has joined #openstack-ansible19:47
prometheanfirewell, that failed quick19:51
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Update ansible version to 1.9.6 for CVE-2016-3096  https://review.openstack.org/30660919:52
openstackgerritAmy Marrich (spotz) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host  https://review.openstack.org/30484019:52
*** schwicht has quit IRC19:56
*** cfarquhar has quit IRC20:03
*** ChrisBenson1 has quit IRC20:04
*** ChrisBenson has joined #openstack-ansible20:05
*** cfarquhar has joined #openstack-ansible20:06
*** cfarquhar has quit IRC20:06
*** cfarquhar has joined #openstack-ansible20:06
*** hybridpolio has quit IRC20:09
*** evanstoner has quit IRC20:15
*** sigmavirus24 is now known as sigmavirus24_awa20:17
*** sigmavirus24_awa is now known as sigmavirus2420:17
*** hybridpollo has joined #openstack-ansible20:18
*** automagically has quit IRC20:20
*** basilAB has quit IRC20:22
*** schwicht has joined #openstack-ansible20:26
*** basilAB has joined #openstack-ansible20:29
*** rohanp has quit IRC20:32
cloudnullwe should probably rev forward however that CVE does not effect us.20:45
Bjoern_cloudnull: Strange I expected that the openstack/venvs directory is bind mounted to each container but it seems like it's local to each container20:45
*** johnmilton has quit IRC20:46
cloudnullnope no bind mount20:46
Bjoern_whys that ? Different packages inside the same venvs  ?20:46
*** Bjoern_ is now known as BjoernT20:46
cloudnullcould / can be20:46
cloudnullIE plumgrid packages in a neutron linuxbridge agent container20:47
cloudnullbut not neutron server20:47
BjoernTtok20:47
BjoernTok20:47
cloudnullsame is with the extra cinder volume drivers, etc20:47
cloudnullhowever every venv is crated on the repo server and then downloaded to the target location20:48
cloudnulland those extra bits get installed within the venv based on the requirements20:48
BjoernTyeah we could have implemented all venvs have the same packages across all container types and then bind mount. But it does not matter that much20:48
cloudnulli looked at doing it as a layer using unioning fs however there was almost no performance improvement20:49
cloudnulland being that the /openstac/venvs dir will now have versioned venvs it gives deployers a rather simple rollback plan to a known functional state, regardless of the extra bits that may be added .20:51
cloudnullall that assumes theres no DB migration that would impact roll back.20:51
cloudnullbut thats a larger openstack problem than our project would care to tackle alone20:51
* cloudnull my 2cents20:52
*** Mudpuppy has quit IRC20:57
*** ametts has quit IRC20:58
*** admin0 has joined #openstack-ansible20:59
*** admin0 has quit IRC21:10
*** admin0 has joined #openstack-ansible21:10
mhaydenspotz: https://blueprints.launchpad.net/openstack-ansible/+spec/install-guide https://review.openstack.org/#/c/241037/21:10
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible-ironic: Changed the swift storage URL retrieval command  https://review.openstack.org/30663221:14
cloudnullany cores around to give that a spin?  -- odyssey4me andymccr_ ?21:15
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: Add `ironic_swift_temp_url_secret_key` the secrets  https://review.openstack.org/30663321:21
*** retreved_ has quit IRC21:21
*** b3rnard0 is now known as b3rnard0_away21:25
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible-ironic: Changed the swift storage URL retrieval command  https://review.openstack.org/30663221:28
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible: updated ironic playbook inclusion for tftp network  https://review.openstack.org/30365921:32
*** KLevenstein has quit IRC21:37
*** Brew has quit IRC21:39
*** thorst has quit IRC21:43
*** sigmavirus24 is now known as sigmavirus24_awa21:43
*** sigmavirus24_awa is now known as sigmavirus2421:43
*** thorst has joined #openstack-ansible21:43
*** thorst_ has joined #openstack-ansible21:46
*** schwicht has quit IRC21:47
*** thorst has quit IRC21:48
*** spotz is now known as spotz_zzz21:50
*** thorst_ has quit IRC21:50
*** busterswt has quit IRC21:56
*** eil397 has joined #openstack-ansible21:57
*** raginbajin has quit IRC22:00
*** thorst has joined #openstack-ansible22:00
*** b3rnard0_away is now known as b3rnard022:02
*** thorst has quit IRC22:05
*** raginbajin has joined #openstack-ansible22:05
*** sigmavirus24 is now known as sigmavirus24_awa22:05
*** thorst has joined #openstack-ansible22:06
*** phalmos_ has quit IRC22:07
*** thorst has quit IRC22:14
*** weezS has quit IRC22:18
*** falanx has quit IRC22:20
*** weezS has joined #openstack-ansible22:20
*** schwicht has joined #openstack-ansible22:22
*** sonus has quit IRC22:22
*** sonus has joined #openstack-ansible22:22
*** odyssey4me has quit IRC22:22
*** dalees has quit IRC22:23
*** dalees has joined #openstack-ansible22:26
*** odyssey4me has joined #openstack-ansible22:27
*** falanx has joined #openstack-ansible22:34
*** thorst has joined #openstack-ansible22:38
*** schwicht has quit IRC22:41
*** thorst has quit IRC22:47
*** falanx has quit IRC22:50
*** falanx has joined #openstack-ansible23:07
*** falanx has quit IRC23:14
*** pjm6 has quit IRC23:16
*** BjoernT has quit IRC23:17
*** admin0 has quit IRC23:19
*** woodard has quit IRC23:23
*** falanx has joined #openstack-ansible23:31
*** weezS has quit IRC23:32
*** markvoelker_ has quit IRC23:37
*** asettle has joined #openstack-ansible23:41
*** asettle has quit IRC23:43
*** thorst has joined #openstack-ansible23:44
*** busterswt has joined #openstack-ansible23:46
*** eil397 has left #openstack-ansible23:46
*** busterswt has quit IRC23:50
*** thorst has quit IRC23:52

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!