| *** arif-ali has quit IRC | 00:06 | |
| openstackgerrit | Andrey Grebennikov proposed openstack/openstack-ansible-galera_server master: Added Galera SSL support https://review.openstack.org/425997 | 00:07 |
|---|---|---|
| *** rob0t_ has joined #openstack-ansible | 00:08 | |
| *** hfu has joined #openstack-ansible | 00:12 | |
| *** rob0t_ has quit IRC | 00:12 | |
| *** esberglu has joined #openstack-ansible | 00:16 | |
| *** esberglu has quit IRC | 00:17 | |
| *** esberglu has joined #openstack-ansible | 00:17 | |
| MasterOfBugs | Hi All | 00:18 |
| MasterOfBugs | I want to deploy 3 controller and 50+ compute with OVS on an openstack-ansible environment | 00:19 |
| *** adrian_otto has joined #openstack-ansible | 00:19 | |
| MasterOfBugs | Is there a documentation which explains me that in detail? | 00:19 |
| MasterOfBugs | me -* | 00:19 |
| *** rob0t_ has joined #openstack-ansible | 00:20 | |
| *** jperry has quit IRC | 00:20 | |
| *** arif-ali has joined #openstack-ansible | 00:20 | |
| *** esberglu has quit IRC | 00:22 | |
| *** agrebennikov__ has quit IRC | 00:24 | |
| *** rob0t_ has quit IRC | 00:24 | |
| *** LinStatSDR has quit IRC | 00:36 | |
| *** chyka has quit IRC | 00:36 | |
| *** weezS has joined #openstack-ansible | 00:41 | |
| *** weezS has quit IRC | 00:41 | |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible stable/mitaka: Bump requirements SHA https://review.openstack.org/428511 | 00:43 |
| openstackgerrit | Tom Jose Kalapura proposed openstack/openstack-ansible stable/newton: Use https to retrieve upper-constraints https://review.openstack.org/428513 | 00:45 |
| *** LinStatSDR has joined #openstack-ansible | 00:47 | |
| *** acormier has joined #openstack-ansible | 00:49 | |
| *** hw_wutianwei has joined #openstack-ansible | 00:51 | |
| *** acormier has quit IRC | 00:54 | |
| *** pjh03 has joined #openstack-ansible | 00:59 | |
| *** retreved has quit IRC | 00:59 | |
| *** retreved has joined #openstack-ansible | 00:59 | |
| *** weezS has joined #openstack-ansible | 01:03 | |
| *** pjh03 has quit IRC | 01:03 | |
| *** jamesdenton has joined #openstack-ansible | 01:07 | |
| *** jamesdenton has quit IRC | 01:09 | |
| *** pjh03 has joined #openstack-ansible | 01:12 | |
| cloudnull | evenings | 01:15 |
| logan- | hello | 01:15 |
| cloudnull | hows it logan- | 01:16 |
| logan- | great. messing with M->N finally | 01:17 |
| logan- | you? | 01:17 |
| *** pjh03 has quit IRC | 01:17 | |
| logan- | kind of amazing how slow this is http://cdn.pasteraw.com/ryb4m96x0cshk5mithblq31572sb9pv | 01:17 |
| cloudnull | wow 17 sec | 01:18 |
| cloudnull | internal routing? | 01:18 |
| cloudnull | doing well, just playing catch-up -- the forever battle | 01:18 |
| LinStatSDR | no sdn :( | 01:18 |
| cloudnull | LinStatSDR: ? | 01:18 |
| LinStatSDR | software defined networking | 01:19 |
| LinStatSDR | like contrail | 01:19 |
| cloudnull | ah, yea. no contrail | 01:20 |
| cloudnull | we have nuage and plumgrid | 01:20 |
| logan- | yeah local boxes low latency etc etc. was trying to think of a way to make it loop thru hostvars faster | 01:20 |
| *** thorst_ has quit IRC | 01:20 | |
| *** crushil has joined #openstack-ansible | 01:20 | |
| LinStatSDR | mm | 01:20 |
| cloudnull | and calico | 01:20 |
| LinStatSDR | yeah i use calico | 01:20 |
| LinStatSDR | i feel | 01:21 |
| *** thorst_ has joined #openstack-ansible | 01:21 | |
| cloudnull | logan-: hum... | 01:21 |
| logan- | even targeting that debug at localhost is 17sce | 01:21 |
| logan- | sec* | 01:21 |
| logan- | i guess it is some ansible thing | 01:21 |
| cloudnull | maybe its the inventory processing ? | 01:21 |
| LinStatSDR | 17s i would look into kinda long | 01:21 |
| LinStatSDR | i came in late, what exactly is taking 17s? | 01:22 |
| cloudnull | curious if a static inventory with a limited set of hosts would be faster? | 01:22 |
| LinStatSDR | hmmm lots of ips in there | 01:22 |
| logan- | looping thru an 81 host group in hostvars LinStatSDR: http://cdn.pasteraw.com/ryb4m96x0cshk5mithblq31572sb9pv | 01:22 |
| LinStatSDR | yeah i see tha tnow | 01:23 |
| LinStatSDR | all internal tho | 01:23 |
| logan- | it isn't even connecting to them, seems to be laggy inventory | 01:24 |
| LinStatSDR | interesting | 01:24 |
| * LinStatSDR shrugs | 01:24 | |
| cloudnull | also curious if a jinja loop using {% for i in ... %}{% endfor %} is faster ? | 01:24 |
| logan- | nope tested that too, 17 seconds also | 01:24 |
| cloudnull | hum. | 01:24 |
| LinStatSDR | if you keep getting the same times | 01:24 |
| cloudnull | ansible 2.2.x ? | 01:25 |
| *** weezS has quit IRC | 01:25 | |
| *** thorst_ has quit IRC | 01:25 | |
| *** evilt0ne has quit IRC | 01:25 | |
| logan- | newton 2.1.4.0 | 01:25 |
| *** cmart has quit IRC | 01:27 | |
| logan- | i'll try latest just for kicks | 01:28 |
| *** retreved has quit IRC | 01:28 | |
| logan- | maybe this will resolve itself for ocata :) | 01:28 |
| *** sdake has joined #openstack-ansible | 01:28 | |
| cloudnull | logan-: http://cdn.pasteraw.com/fcal8gmytl08ewkygbodf1znsxr52jo | 01:30 |
| cloudnull | 13 sec | 01:30 |
| cloudnull | so not much better | 01:30 |
| cloudnull | and thats on ansible 2.3 | 01:30 |
| logan- | same here on 2.2.1.0 | 01:30 |
| logan- | 13.487 | 01:30 |
| logan- | :( | 01:30 |
| *** portdirect is now known as portdirect_away | 01:30 | |
| *** rob0t_ has joined #openstack-ansible | 01:36 | |
| cloudnull | so it looks like it's slow to process the inventory in the loop | 01:37 |
| cloudnull | inventory load time seems good. | 01:38 |
| cloudnull | though the jinja loop seems like it's slow | 01:38 |
| logan- | mitaka 1.9.6 is at 5.379s so it was still pretty slow but not quite as bad on 1.9 | 01:39 |
| logan- | yeah it seems to be when you loop hostvars via jinja you hit this | 01:40 |
| *** chris_hultin|AWA is now known as chris_hultin | 01:41 | |
| *** LinStatSDR has quit IRC | 01:42 | |
| *** evilt0ne has joined #openstack-ansible | 01:43 | |
| *** evilt0ne has quit IRC | 01:43 | |
| *** evilt0ne has joined #openstack-ansible | 01:43 | |
| *** chris_hultin is now known as chris_hultin|AWA | 01:43 | |
| *** kberger has quit IRC | 01:44 | |
| cloudnull | https://groups.google.com/forum/#!topic/ansible-devel/UU0Tpw-qlhY | 01:45 |
| cloudnull | seems like its a known which is "managable" according to the last pst | 01:46 |
| logan- | hmm that PR they merged is included as of 2.2.1.0 so maybe thats the 17s->13s we're seeing | 01:49 |
| *** acormier has joined #openstack-ansible | 01:49 | |
| *** acormier has quit IRC | 01:50 | |
| *** acormier has joined #openstack-ansible | 01:50 | |
| *** thorst_ has joined #openstack-ansible | 02:00 | |
| *** thorst_ has quit IRC | 02:00 | |
| *** adrian_otto has quit IRC | 02:06 | |
| *** rob0t_ has quit IRC | 02:07 | |
| *** gouthamr has joined #openstack-ansible | 02:09 | |
| *** cathrich_ has joined #openstack-ansible | 02:13 | |
| *** cathrichardson has quit IRC | 02:13 | |
| *** rob0t_ has joined #openstack-ansible | 02:14 | |
| boxrick1 | Is there currently a good branch which is safe to build from ? | 02:17 |
| *** Mahe has quit IRC | 02:23 | |
| *** Mahe has joined #openstack-ansible | 02:25 | |
| *** thorst_ has joined #openstack-ansible | 02:32 | |
| *** thorst_ has quit IRC | 02:32 | |
| *** woodard has quit IRC | 02:39 | |
| *** woodard has joined #openstack-ansible | 02:40 | |
| cloudnull | logan-: that may be | 02:41 |
| cloudnull | sorry was getting dinner | 02:41 |
| cloudnull | boxrick1: the last deployment I did was from 14.0.6 | 02:42 |
| boxrick1 | How long ago ? | 02:42 |
| cloudnull | which was from the stable/newton branch | 02:42 |
| cloudnull | two weeks | 02:42 |
| cloudnull | I've done several labs since then on the head of stable newton | 02:42 |
| cloudnull | are you running into things ? | 02:43 |
| boxrick1 | Yea, was having this problem with "master" about 5 hours ago: https://gist.github.com/boxrick/b40cb2c6589876d4f2e19c6553327633 | 02:46 |
| boxrick1 | Someone said there were fixes being pushed at the time though | 02:46 |
| boxrick1 | Doesn't matter too much this exact second. Just cleaning up some of these roles to replace Cobbler to push upstream and just making sure things still work with my stuff. | 02:47 |
| cloudnull | mcarden: ^ you were seeing something with the run playbooks script ? | 02:47 |
| *** MasterOfBugs has quit IRC | 02:47 | |
| cloudnull | boxrick1: is this for a multi-node deploy or are you running this on a cluster ? | 02:47 |
| boxrick1 | Yea, this is a multinode deploy. | 02:47 |
| mcarden | I was having a problem, but it wasn't that problem and also the patch to fix the problem I was having, has merged. | 02:48 |
| boxrick1 | Ill give 14.0.6 another go | 02:49 |
| boxrick1 | And see where I get | 02:49 |
| *** pmannidi has quit IRC | 02:57 | |
| *** pmannidi has joined #openstack-ansible | 02:58 | |
| *** whiteveil has joined #openstack-ansible | 02:59 | |
| *** whiteveil has quit IRC | 03:02 | |
| *** crushil has quit IRC | 03:05 | |
| *** cmart has joined #openstack-ansible | 03:14 | |
| mhayden | andymccr: so it seems that CentOS is 2-3x slower in the gate on tasks across the board | 03:23 |
| cloudnull | :'( | 03:24 |
| mhayden | some of them are ansible heavy | 03:24 |
| mhayden | but some are just easy stuff, like git clones | 03:24 |
| mhayden | i wonder if there's a config difference in the gate for centos | 03:24 |
| cloudnull | could be the image ? | 03:24 |
| mhayden | i'm wondering if so | 03:24 |
| mhayden | because my deployment on centos seemed to be the same speed as ubuntu | 03:25 |
| mhayden | i would have noticed if it wasn't! | 03:25 |
| *** crushil has joined #openstack-ansible | 03:25 | |
| mhayden | cloudnull: https://gist.github.com/major/7933b918723bff2216190a38bf2c9572 | 03:29 |
| mhayden | repo build was almost 2x | 03:29 |
| cloudnull | are those the gate timings? | 03:30 |
| cloudnull | ah, well this kinda makes sense. i think (for repo-build that is) | 03:30 |
| mhayden | yes | 03:30 |
| cloudnull | the image we used is from the upstream gate. | 03:31 |
| cloudnull | which caches just about everything for devstack | 03:31 |
| cloudnull | I suspect they dont do the same level of caching for centos | 03:31 |
| cloudnull | our gat ejobs tries to get and use the cached bits to make things faster if present | 03:31 |
| mhayden | quick script i hacked together https://gist.github.com/major/b090eaa865b5e86ec9676610acb6b47f | 03:32 |
| mhayden | hmm, i wonder what would happen if we disabled fastestmirror | 03:32 |
| mhayden | but yeah, this appears to be happening with anything that depends heavily on the network | 03:33 |
| mhayden | create neutron ha task is 41 sec vs 52 sec | 03:33 |
| mhayden | fairly close | 03:33 |
| mhayden | compressing static files in horizon is within 15-20 sec | 03:34 |
| mhayden | perhaps i'll cobble a patch to disable fastestmirror just in case | 03:38 |
| mhayden | i disable it on quite a few centos machines | 03:38 |
| cloudnull | there are a mess of packages that gets installed | 03:38 |
| cloudnull | so that could be it | 03:38 |
| mhayden | sometimes the fastestmirror check hangs for a long time | 03:40 |
| *** sdake has quit IRC | 03:41 | |
| *** crushil has quit IRC | 03:41 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: corrected infra wheel mirror URL https://review.openstack.org/428558 | 03:42 |
| cloudnull | ^ found a bug :) | 03:42 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-openstack_hosts master: Disable yum fastestmirror plugin https://review.openstack.org/428560 | 03:43 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-lxc_hosts master: Copy over fastestmirror configuration https://review.openstack.org/428561 | 03:46 |
| mhayden | we can try those out | 03:46 |
| mhayden | mgariepy: i linked to those patches in 407690 | 03:49 |
| mhayden | cloudnull: but centos should be using the pip mirror from openstack, right? | 03:52 |
| cloudnull | yes it looks like the current mirror selector should cover it | 03:53 |
| mhayden | well, my .pip directory in my repo server doesn't specify a mirror | 03:56 |
| mhayden | err, i guess it wouldn't on a non-gate run | 03:56 |
| cloudnull | yea it should only be run when detected to be running w/in nodepool | 03:57 |
| *** thorst_ has joined #openstack-ansible | 03:58 | |
| *** thorst_ has quit IRC | 03:58 | |
| mhayden | i'll go run one in a 8gb vm and see | 04:00 |
| *** hfu has quit IRC | 04:00 | |
| *** hfu has joined #openstack-ansible | 04:02 | |
| mhayden | imma go sleepytime while that runs | 04:03 |
| mhayden | night folks | 04:03 |
| *** crushil has joined #openstack-ansible | 04:07 | |
| *** sdake has joined #openstack-ansible | 04:07 | |
| *** acormier has quit IRC | 04:10 | |
| cloudnull | cheers mhayden | 04:10 |
| *** sdake_ has joined #openstack-ansible | 04:11 | |
| boxrick1 | cloudnull using 14.0.6 gives this: ERROR! Unexpected Exception: The 'packaging>=16.8' distribution was not found and is required by setuptools | 04:11 |
| boxrick1 | Which I think was fixed in a more recent build? | 04:11 |
| cloudnull | ah. that was fixed in the branch | 04:12 |
| cloudnull | sorry. | 04:12 |
| cloudnull | stable/newton | 04:12 |
| cloudnull | I forgot about that issue | 04:12 |
| *** sdake has quit IRC | 04:12 | |
| *** udesale has joined #openstack-ansible | 04:13 | |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in Apache2 role - Ubuntu 16.04 https://review.openstack.org/428566 | 04:13 |
| *** esberglu has joined #openstack-ansible | 04:17 | |
| *** gouthamr has quit IRC | 04:17 | |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in DHCPD role - Ubuntu 16.04 https://review.openstack.org/428569 | 04:18 |
| *** thorst_ has joined #openstack-ansible | 04:21 | |
| *** thorst_ has quit IRC | 04:21 | |
| *** esberglu has quit IRC | 04:21 | |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in TFTP Configure role - Ubuntu 16.04 https://review.openstack.org/428570 | 04:22 |
| *** jcannava has quit IRC | 04:23 | |
| *** jcannava has joined #openstack-ansible | 04:25 | |
| boxrick1 | Hopefully if these end up looking OK we can start thinking about Cobbler going away. | 04:26 |
| cloudnull | +1 they look pretty good. | 04:26 |
| cloudnull | tomorrow i'm going to give it all a go. | 04:26 |
| *** furlongm has quit IRC | 04:27 | |
| boxrick1 | I have all the glue which brings it together, but lets make sure this looks all good first then ill push that next. | 04:29 |
| boxrick1 | Right, its rather late here time to sleep. I shall speak to you later. | 04:31 |
| *** furlongm has joined #openstack-ansible | 04:32 | |
| *** admin0_ has joined #openstack-ansible | 04:32 | |
| *** crushil has quit IRC | 04:33 | |
| *** crushil has joined #openstack-ansible | 04:33 | |
| *** acormier has joined #openstack-ansible | 04:34 | |
| *** crushil has quit IRC | 04:39 | |
| cloudnull | take care boxrick1, ttyt | 04:39 |
| *** crushil has joined #openstack-ansible | 04:41 | |
| *** lucas_ has joined #openstack-ansible | 04:49 | |
| *** drifterza has joined #openstack-ansible | 04:52 | |
| *** lucas_ has quit IRC | 04:52 | |
| *** acormier has quit IRC | 04:52 | |
| *** lucas_ has joined #openstack-ansible | 04:53 | |
| *** haasn has quit IRC | 05:00 | |
| *** lucas_ has quit IRC | 05:01 | |
| *** haasn has joined #openstack-ansible | 05:05 | |
| *** shausy has joined #openstack-ansible | 05:09 | |
| *** adrian_otto has joined #openstack-ansible | 05:15 | |
| *** crushil has quit IRC | 05:16 | |
| *** crushil has joined #openstack-ansible | 05:16 | |
| *** lucas_ has joined #openstack-ansible | 05:17 | |
| *** lucas_ has quit IRC | 05:22 | |
| *** cmart has quit IRC | 05:23 | |
| *** shausy has quit IRC | 05:25 | |
| *** shausy has joined #openstack-ansible | 05:26 | |
| *** Jolrael has quit IRC | 05:53 | |
| *** Jolrael has joined #openstack-ansible | 05:57 | |
| *** phalmos has quit IRC | 05:58 | |
| *** Jack_Iv has joined #openstack-ansible | 05:58 | |
| *** MasterOfBugs has joined #openstack-ansible | 06:01 | |
| *** pramodrj07 has joined #openstack-ansible | 06:04 | |
| *** MasterOfBugs has quit IRC | 06:05 | |
| *** admin0_ has quit IRC | 06:10 | |
| *** kiriakos has joined #openstack-ansible | 06:12 | |
| *** crushil has quit IRC | 06:13 | |
| *** thorst_ has joined #openstack-ansible | 06:22 | |
| *** thorst_ has quit IRC | 06:27 | |
| *** Andrew_jedi has joined #openstack-ansible | 06:46 | |
| *** kiriakos has quit IRC | 06:47 | |
| *** unlaudable has joined #openstack-ansible | 06:47 | |
| *** sdake_ has quit IRC | 06:52 | |
| *** Oku_OS-away is now known as Oku_OS | 07:07 | |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron master: Implement Dragonflow deployment https://review.openstack.org/391524 | 07:07 |
| *** jrobinson has quit IRC | 07:13 | |
| evrardjp[m] | Good morning everyone | 07:15 |
| *** adrian_otto has quit IRC | 07:20 | |
| *** adrian_otto has joined #openstack-ansible | 07:26 | |
| *** adrian_otto has quit IRC | 07:33 | |
| *** Andrew_jedi has quit IRC | 07:40 | |
| *** Jack_Iv has quit IRC | 07:44 | |
| *** Jack_Iv has joined #openstack-ansible | 07:45 | |
| *** esberglu has joined #openstack-ansible | 07:50 | |
| *** esberglu has quit IRC | 07:54 | |
| *** woodard has quit IRC | 08:00 | |
| *** admin0_ has joined #openstack-ansible | 08:11 | |
| *** pramodrj07 has quit IRC | 08:18 | |
| *** Jack_Iv has quit IRC | 08:21 | |
| *** thorst_ has joined #openstack-ansible | 08:23 | |
| *** thorst_ has quit IRC | 08:28 | |
| *** Andrew_jedi has joined #openstack-ansible | 08:31 | |
| *** fxpester has joined #openstack-ansible | 08:55 | |
| andymccr | mgariepy: ive moved the nova CentOS gate to non-voting again, unfortunately its timing out about 50% of the time so merging things became impossible - we need to figure out the centos time discrepancy and then re-enable it as voting. | 08:59 |
| *** rob0t_ has quit IRC | 08:59 | |
| * andymccr burns a goat at the altar of the gate-lords | 09:00 | |
| evrardjp | haha | 09:04 |
| *** kiriakos has joined #openstack-ansible | 09:04 | |
| kiriakos | Hi again there, on Yesterday's try I had an issue where machines could not access the default gateway of their internal subnet. They could ping the .2 IP though Which I suppose is the dnsmaq service? Subnet was like 10.30.1.0/24 Clients on 10.30.1.7&8 DG 10.30.1.1 (assigned by default by Openstack & not pingable), DHCP on 10.30.1.2 (pingable) | 09:05 |
| *** admin0_ has quit IRC | 09:06 | |
| *** evilkry has quit IRC | 09:07 | |
| *** admin0_ has joined #openstack-ansible | 09:08 | |
| kiriakos | So far I understand that sort of traffic should becompletely handled by the neutron agent in the compute node but when pinging 10.30.1.1. tcpdump showed that ARPs were sent by .8 to discover .1's mac but no response was recorded at all. | 09:10 |
| kiriakos | The Arpscame from the neutron linuxbridge interface | 09:13 |
| *** sawblade6 has quit IRC | 09:14 | |
| *** sawblade6 has joined #openstack-ansible | 09:23 | |
| *** karimb has joined #openstack-ansible | 09:27 | |
| *** udesale__ has joined #openstack-ansible | 09:28 | |
| *** udesale has quit IRC | 09:30 | |
| *** skuda has joined #openstack-ansible | 09:34 | |
| *** udesale__ has quit IRC | 09:46 | |
| *** udesale has joined #openstack-ansible | 09:46 | |
| *** udesale has quit IRC | 09:46 | |
| *** hfu has quit IRC | 09:47 | |
| *** udesale has joined #openstack-ansible | 09:52 | |
| hachi | Hello, | 09:56 |
| hachi | I am trying to do ceph OSA integration on stable/newton. for ceph i am using this repo: | 09:56 |
| hachi | https://github.com/ceph/ceph-ansible/tree/v2.1.5 | 09:56 |
| hachi | My problem is that the ceph-common role checks the existing of keyring before it gets created!, this the error : | 09:57 |
| hachi | http://paste.openstack.org/show/597516/ | 09:58 |
| hachi | I am not sure whether this error is due to misconfiguration on my intergartion or in ceph-ansible itself, here is how my user_ceph.yml user variables looks: | 09:59 |
| *** evilt0ne has quit IRC | 09:59 | |
| hachi | http://paste.openstack.org/show/597517/ | 09:59 |
| hachi | any idea ? | 10:00 |
| *** thorst_ has joined #openstack-ansible | 10:24 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/mitaka: Add dummy tox config to workaround bad CI config https://review.openstack.org/428677 | 10:24 |
| andymccr | hachi: i dont think any ceph experts are online atm! So you might get more assistance later on - in master we have an integrated ceph build - you may be able to gather some ideas from there, and some info about the required config: https://github.com/openstack/openstack-ansible/commit/06ede41645550b1260f0c457aeca7ab7feffe918 | 10:26 |
| odyssey4me | andymccr I'm finding more and more of those rogue jobs on the mitaka branch | 10:26 |
| odyssey4me | jmccrory ^ | 10:26 |
| odyssey4me | trying to figure it out | 10:27 |
| hachi | andymccr thank you for your response, i will review the link you mentioned and come back later on. | 10:29 |
| *** thorst_ has quit IRC | 10:29 | |
| andymccr | odyssey4me: thanks! yeah i noticed the neutron one, meant to put in a patch to just dummy it for now | 10:29 |
| openstackgerrit | Merged openstack/openstack-ansible-security master: Configure pam_faildelay on Ubuntu https://review.openstack.org/426892 | 10:29 |
| odyssey4me | andymccr like this one? https://review.openstack.org/#/c/428677/ | 10:29 |
| andymccr | hachi: sorry about that :) but that should help because that gate on master is working and passing! | 10:29 |
| andymccr | odyssey4me: yeah exactly! maybe thats the simplest way | 10:29 |
| odyssey4me | sure, we can implement patches like those for now while we work out the CI stuff, if we ever do | 10:30 |
| *** udesale__ has joined #openstack-ansible | 10:30 | |
| andymccr | odyssey4me: yeah part of me thinks so few patches go in its not worth risking messing up master/the work involved in untangled that spaghetti web | 10:31 |
| odyssey4me | yeah, agreed | 10:31 |
| odyssey4me | andymccr this one is an easy fix though: https://review.openstack.org/428675 | 10:32 |
| hachi | andymccr Ok then :), i think it would be good if i can test the intergation on the master then on the stable/newton branch | 10:32 |
| *** udesale__ has quit IRC | 10:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server master: Update galera running check for CentOS https://review.openstack.org/426934 | 10:33 |
| andymccr | hachi: yeah id recommend using stable/newton now, but it should help to get it going on stable/newton :) | 10:33 |
| *** udesale has quit IRC | 10:33 | |
| hachi | Ok i will review and apply directly on stable/newton | 10:34 |
| odyssey4me | hachi you oculd try that, although then you'd have to maintain your fork | 10:36 |
| odyssey4me | I think andymccr meant that all the settings needed are there and you can learn from them. | 10:36 |
| andymccr | yeah ^ that :) | 10:37 |
| odyssey4me | andymccr are there any patches that need reviews for the RC? | 10:38 |
| andymccr | odyssey4me: not atm, basically the one is merging now (centos7 moved to non-voting) | 10:39 |
| andymccr | then i just need to update the rc patch for the os_nova sha and then it'll need a reeview | 10:39 |
| andymccr | *review | 10:39 |
| andymccr | tested it manually and it works so hopefully that'll be all good :) | 10:39 |
| odyssey4me | yeah, I see mhayden put in https://review.openstack.org/428560 and https://review.openstack.org/428561 to try and speed up the infra job | 10:40 |
| odyssey4me | I think that's an important thing to do regardless - a deployment should only use the mirror it has configured, not some random mirror from elsewhere. | 10:41 |
| hachi | odyssey4me so by maintaining the fork you mean cloning the appropriate ceph roles that works with stable/newton instead of master OSA, right ? :) | 10:41 |
| odyssey4me | hachi if you use the patch from master you would have to cherry-pick it to your fork of the github repo | 10:41 |
| andymccr | odyssey4me: ahh i see. thanks mhayden :D lets give it a go | 10:41 |
| odyssey4me | from then on in order to get updates for newton you'd have to maintain that | 10:42 |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Disable yum fastestmirror plugin https://review.openstack.org/428560 | 10:44 |
| *** hw_wutianwei has quit IRC | 10:44 | |
| hachi | odyssey4me Ah got it now. for now i will just work on getting the deployment right on a testing environment, based on that i will try to adjust it for somthing easy to maintain on production | 10:46 |
| odyssey4me | hmm, I think I have the bug in the CI config andymccr | 10:47 |
| openstackgerrit | Pedro Magalhães (pjm6) proposed openstack/openstack-ansible-os_cinder stable/newton: [DOCS] Adding configuration for use Dell EQLX in cinder https://review.openstack.org/428687 | 10:47 |
| odyssey4me | I can put in a test patch to validate | 10:47 |
| odyssey4me | I think the first inclusion is getting a broader inclusion from https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L3287 | 10:48 |
| odyssey4me | so lemme try and remove the liberty/mitaka branches from there to see what happens | 10:48 |
| openstackgerrit | Pedro Magalhães (pjm6) proposed openstack/openstack-ansible-os_cinder master: [DOCS] Adding missing config parameter of Dell EQLX in cinder https://review.openstack.org/428688 | 10:50 |
| pjm6 | good morning all | 10:51 |
| *** stuartgr has joined #openstack-ansible | 10:55 | |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts master: Copy over fastestmirror configuration https://review.openstack.org/428561 | 10:55 |
| pjm6 | I'm using cinder with dell eqlx volume driver | 10:57 |
| pjm6 | for some reason, when I create an instance booted with volume, sometimes works and sometimes not | 10:58 |
| pjm6 | here is the logs of a success and failure operation | 10:58 |
| pjm6 | succcess: pastebin.com/WgV9SJ6P | 10:58 |
| pjm6 | failure: pastebim.com/sEZcwUpw | 10:59 |
| pjm6 | it seems that for some reason, he cannot found image from glance | 10:59 |
| pjm6 | odyssey4me: could you take a look when you have time? | 11:00 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate master: Enable current Designate services https://review.openstack.org/427810 | 11:00 |
| odyssey4me | andymccr I think we should get this in before we release os_designate: https://review.openstack.org/426904 | 11:00 |
| odyssey4me | pjm6 what glance store is configured? | 11:01 |
| odyssey4me | are you using ceph, swift, or file? | 11:01 |
| pjm6 | odyssey4me: is file | 11:01 |
| openstackgerrit | Merged openstack/openstack-ansible-os_cinder master: [DOCS] Adding missing config parameter of Dell EQLX in cinder https://review.openstack.org/428688 | 11:01 |
| odyssey4me | do you have more than one infra server? | 11:01 |
| pjm6 | and i'm not using NFS | 11:01 |
| pjm6 | yes | 11:01 |
| odyssey4me | well, that's the problem right there | 11:01 |
| pjm6 | hmmm, there is only in one server | 11:01 |
| pjm6 | and as I don't have NFS | 11:01 |
| pjm6 | he fails? | 11:01 |
| odyssey4me | the LB will send the request to one of three glance stores | 11:01 |
| odyssey4me | only one will have the image, as there is no shared or replicated store | 11:01 |
| odyssey4me | so one of three requests will fail | 11:02 |
| odyssey4me | so best to implement a shared store, or a replicated store | 11:02 |
| odyssey4me | it doesn't have to be NFS - it can just be the folder mounted via any means | 11:03 |
| pjm6 | hmm I see | 11:03 |
| pjm6 | i will see you do a replicated store | 11:03 |
| odyssey4me | I have often thought about putting a little hack in similar to the repo server | 11:03 |
| pjm6 | (for now i'm doing tests with storage, so I suposse that will be more easy to do) | 11:03 |
| pjm6 | odyssey4me: for having one place with all files, right? | 11:04 |
| odyssey4me | yep | 11:04 |
| odyssey4me | so basically setup a LUN on your Dell storage, then mount that LUN from each of your glance containers | 11:04 |
| odyssey4me | you'll have to move the existing data over | 11:04 |
| odyssey4me | but that's your simplest option right now | 11:05 |
| drifterza | swift/ceph for me is best for glance | 11:05 |
| drifterza | Hey Jesse, could you give this another pass ? https://review.openstack.org/#/c/397313/ | 11:08 |
| pjm6 | odyssey4me: that should do manually of I could use OSA to do that? | 11:08 |
| pjm6 | drifterza: I don't have ceph :\ | 11:08 |
| *** yolanda has quit IRC | 11:08 | |
| drifterza | oh so you're not going to do object storage at all ? | 11:08 |
| pjm6 | drifterza: no, for now the aim is just to use Block Storage | 11:09 |
| openstackgerrit | Merged openstack/openstack-ansible-os_cinder stable/newton: [DOCS] Adding configuration for use Dell EQLX in cinder https://review.openstack.org/428687 | 11:09 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Add nova-placement-api service and cell_v2 setup https://review.openstack.org/418457 | 11:09 |
| odyssey4me | pjm6 can you backport https://review.openstack.org/428688 to newton quickly please | 11:09 |
| pjm6 | suer | 11:09 |
| pjm6 | sure | 11:09 |
| drifterza | yeah well then the only options is NFS or sharing the same lun | 11:09 |
| odyssey4me | pjm6 you'd need to do it manually | 11:09 |
| openstackgerrit | Pedro Magalhães (pjm6) proposed openstack/openstack-ansible-os_cinder stable/newton: [DOCS] Adding missing config parameter of Dell EQLX in cinder https://review.openstack.org/428695 | 11:10 |
| evrardjp | placement-api service merged \o/ | 11:10 |
| odyssey4me | pjm6 technically I think the store is bind-mounted to the hosts anyway, so you can pre-setup the mount on the host | 11:10 |
| *** yolanda has joined #openstack-ansible | 11:10 | |
| pjm6 | btw: but sharing the same lun using iSCSI will not have concurrency problems? | 11:10 |
| odyssey4me | andymccr figure out the bug in the CI jobs :) https://review.openstack.org/428691 | 11:11 |
| odyssey4me | pjm6 you will have to ensure that the lun can be mounted by multiple hosts | 11:11 |
| drifterza | pjm6: it only hurts when you have concurrent writes to the same files on the lun | 11:11 |
| drifterza | but glance is read-only no ? | 11:11 |
| odyssey4me | glance itself will not write to the same file from multiple sources | 11:11 |
| pjm6 | thanks odyssey4me, drifterza | 11:11 |
| andymccr | odyssey4me: nice! | 11:11 |
| pjm6 | oh its only in the same file? | 11:11 |
| odyssey4me | glance writes snapshots and when you upload images | 11:12 |
| pjm6 | I was thinking that was not regarding any file | 11:12 |
| *** hfu has joined #openstack-ansible | 11:13 | |
| pjm6 | btw: do you know where the images are stored by default? | 11:13 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Fix SHA's for rc1 + SHA bump https://review.openstack.org/428212 | 11:13 |
| drifterza | usually /var/lib/glance/images | 11:13 |
| pjm6 | thanks drifterza, thats its =) | 11:14 |
| drifterza | or in the venv | 11:14 |
| drifterza | ah cool | 11:14 |
| drifterza | ;) | 11:14 |
| *** hfu has quit IRC | 11:16 | |
| odyssey4me | andymccr I think we should also make sure this one goes in for the release: https://review.openstack.org/427810 | 11:16 |
| *** sdake has joined #openstack-ansible | 11:18 | |
| andymccr | odyssey4me: its getting close to time - im not sure we can wait that long, it can just go in rc2 surely? | 11:20 |
| odyssey4me | yeah, it can | 11:24 |
| odyssey4me | asettle or spotz another vote for https://review.openstack.org/428695 would be good | 11:25 |
| boxrick1 | Is OSA still in an odd state of flux? Seeing errors like this on stable/newton https://gist.github.com/boxrick/b3ce22bd4bb577fb051257b465e1f155 | 11:25 |
| evrardjp | boxrick1: this error looks like your ansible deploy node couldn't get an url | 11:26 |
| evrardjp | can your deploy node connect to http://10.29.236.200:8181/venvs/14.0.7/ubuntu/keystone-14.0.7-x86_64.checksum ? | 11:27 |
| *** Jack_Iv has joined #openstack-ansible | 11:27 | |
| odyssey4me | andymccr we're still missing the patch that overrides the URL for the config_template URL for the policy files | 11:27 |
| boxrick1 | Seems not evrardjp | 11:27 |
| andymccr | odyssey4me: yeah i know - ive only had time to fix the nova stuff - that will also have to wait for rc2 | 11:27 |
| asettle | odyssey4me: kk | 11:28 |
| odyssey4me | ta asettle | 11:28 |
| evrardjp | boxrick1: well if you can't curl http://10.29.236.200:8181/venvs/14.0.7/ubuntu/keystone-14.0.7-x86_64.checksum from your deploy node, that's your issue :p | 11:28 |
| evrardjp | the reason is a different thing | 11:28 |
| boxrick1 | Ok cheers :) | 11:28 |
| evrardjp | is the file present in your repo, can you deploy node connect to this network, etc... | 11:29 |
| evrardjp | I don't know your config, so hard to guess :p | 11:29 |
| *** hfu has joined #openstack-ansible | 11:30 | |
| boxrick1 | Ill figure this one out, was more just wondering how things were generally so I wasn't chasing my own tail. | 11:30 |
| openstackgerrit | Merged openstack/openstack-ansible-os_cinder stable/newton: [DOCS] Adding missing config parameter of Dell EQLX in cinder https://review.openstack.org/428695 | 11:30 |
| *** sdake_ has joined #openstack-ansible | 11:43 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate master: Configure pools.yaml inside Designate role https://review.openstack.org/426904 | 11:43 |
| openstackgerrit | Merged openstack/openstack-ansible master: Implementation Neutron SR-IOV https://review.openstack.org/415909 | 11:45 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-tests master: Remove work-around for reserved_host_disk_mb bug https://review.openstack.org/428705 | 11:46 |
| *** sdake has quit IRC | 11:46 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova master: Implementation Neutron SR-IOV https://review.openstack.org/415907 | 11:46 |
| odyssey4me | wow, these CentOS jobs do rather take their time | 11:52 |
| *** hfu has quit IRC | 11:53 | |
| pjm6 | odyssey4me: when the creation of a instance fails (nova service) and there's not error logs in any of the compute nodes and storage node | 11:54 |
| pjm6 | and not even in the nova scheduler | 11:54 |
| pjm6 | where's the next step I could use? | 11:54 |
| *** pmannidi has quit IRC | 11:54 | |
| pjm6 | well i think i found it nova conductor | 11:56 |
| *** Hosam has joined #openstack-ansible | 11:56 | |
| *** crushil has joined #openstack-ansible | 11:56 | |
| *** Hosam_ has joined #openstack-ansible | 11:57 | |
| odyssey4me | argh, it would seem that we allowed a breaking change to merge into stable/newton: https://review.openstack.org/#/q/I817ff372c2d58e999a89496fce3d25dc617216f0 | 11:58 |
| odyssey4me | also, palendae we may have an inventory bug? http://lists.openstack.org/pipermail/openstack-operators/2017-February/012573.html | 11:58 |
| odyssey4me | pjm6 yep, nova has many moving parts | 11:59 |
| pjm6 | yeah odyssey4me, and the error is so generic :\ | 11:59 |
| pjm6 | it says | 11:59 |
| pjm6 | No valid host was found. Tehre are not enough hosts avaiable | 11:59 |
| pjm6 | but i have a compute node that only has 1 vm | 11:59 |
| pjm6 | and 1 vcpu used from 4 :| | 12:00 |
| *** Hosam has quit IRC | 12:01 | |
| *** rob0t_ has joined #openstack-ansible | 12:02 | |
| pjm6 | there exist a VCPU limit? | 12:04 |
| pjm6 | in this case, i can exceed the vcpus used | 12:05 |
| *** jwitko has quit IRC | 12:06 | |
| *** rob0t_ has quit IRC | 12:07 | |
| *** jwitko has joined #openstack-ansible | 12:08 | |
| odyssey4me | you'd have to debug the scheduler to see why | 12:09 |
| odyssey4me | the limits are typically cpu, ram or disk | 12:09 |
| odyssey4me | other reasons could be that it's wanting a specific hardware profile and not finding it | 12:10 |
| pjm6 | I think that could be problem with vcpu ratio | 12:10 |
| pjm6 | the other compute node I saw that have insufficient ram :\ | 12:10 |
| odyssey4me | you can mess with the ratio's - they're set in nova.conf | 12:11 |
| pjm6 | yeah i was seeing this config | 12:11 |
| pjm6 | http://docs.openstack.org/developer/openstack-ansible-os_nova/ | 12:11 |
| pjm6 | and to update i only need to run setup-nova.conf --tags nova-config ? | 12:11 |
| odyssey4me | yeah, that should do the trick | 12:13 |
| pjm6 | thanks odyssey4me | 12:14 |
| pjm6 | reading the compute scheduler make senses why is not creating | 12:14 |
| pjm6 | the default cpu_allocation_ratio is 2.0 in OSA | 12:15 |
| pjm6 | i have 4vcpu in the compute node that have available ram | 12:15 |
| pjm6 | so the limit is 8 VMs | 12:15 |
| *** rob0t_ has joined #openstack-ansible | 12:17 | |
| *** esberglu has joined #openstack-ansible | 12:18 | |
| *** rob0t_ has quit IRC | 12:21 | |
| *** Jack_Iv_ has joined #openstack-ansible | 12:22 | |
| *** esberglu has quit IRC | 12:22 | |
| *** Jack_Iv has quit IRC | 12:23 | |
| *** rob0t_ has joined #openstack-ansible | 12:23 | |
| *** DanyC has joined #openstack-ansible | 12:27 | |
| *** sdake has joined #openstack-ansible | 12:27 | |
| *** rob0t_ has quit IRC | 12:28 | |
| *** mcrafty has quit IRC | 12:29 | |
| *** mcrafty has joined #openstack-ansible | 12:29 | |
| *** sdake_ has quit IRC | 12:30 | |
| *** Jack_Iv_ has quit IRC | 12:33 | |
| *** Jack_Iv has joined #openstack-ansible | 12:35 | |
| Andrew_jedi | Folks, have you tried deploying OpenStack on OpenStack VMs. Were you able to get internet connectivity on the instances? | 12:35 |
| *** thorst_ has joined #openstack-ansible | 12:37 | |
| *** Jack_Iv_ has joined #openstack-ansible | 12:40 | |
| *** schwicht has joined #openstack-ansible | 12:41 | |
| *** Jack_Iv has quit IRC | 12:42 | |
| *** Jack_Iv_ has quit IRC | 12:44 | |
| *** Jack_Iv has joined #openstack-ansible | 12:49 | |
| *** drifterza has quit IRC | 12:49 | |
| *** Jack_Iv has quit IRC | 12:49 | |
| *** Jack_Iv has joined #openstack-ansible | 12:49 | |
| *** rob0t_ has joined #openstack-ansible | 12:52 | |
| evrardjp | Andrew_jedi: I guess it depends on the VM and the network | 12:55 |
| andymccr | odyssey4me: hughsaunders any other cores around: would appreciate a +2w https://review.openstack.org/#/c/428212/ | 12:55 |
| evrardjp | for example, in a default OpenStack network, you are not allowed to have multiple mac addresses behind a port, which makes things harder | 12:56 |
| evrardjp | so it's hard to say without knowing a little more | 12:56 |
| evrardjp | andymccr: ok | 12:56 |
| *** rob0t_ has quit IRC | 12:57 | |
| odyssey4me | andymccr done | 12:57 |
| evrardjp | andymccr: don't want to bring the latest? It has some release notes issue fixing for keystone, not that it really matters | 12:58 |
| odyssey4me | evrardjp that'll come in the next rc | 12:59 |
| evrardjp | naturally :p | 12:59 |
| spotz | I got beat to the DOC one, but got to +2 and workflow one of cloudniull's I was waiting on jenkins for:) | 12:59 |
| *** kiriakos_ has joined #openstack-ansible | 12:59 | |
| *** kiriakos has quit IRC | 12:59 | |
| logan- | wonder why ceph failed tempest on the sha bump one :/ | 12:59 |
| *** askb has quit IRC | 12:59 | |
| andymccr | evrardjp: yeah the gate will take 3+ hours if i make a change (assuming it will not time out) - id rather not make changes unless they are critical, anything missing can go in rc2 | 13:00 |
| odyssey4me | oanson evrardjp I see that https://review.openstack.org/391524 is failing the tests, any idea whether we'll get resolution for that soon? | 13:00 |
| oanson | odyssey4me, I am working on it - currently I am trying to get AIO pass tempest | 13:00 |
| andymccr | im mostly just glad we got the placement stuff in and were able to sha bump on ceilometer. | 13:01 |
| oanson | As a quick fix, I can add the failing tests to the skip-list | 13:01 |
| odyssey4me | oanson ok, great - I just wanted to make sure that we're still on top of it | 13:01 |
| odyssey4me | oanson if you'd like dragonflow included in the ocata release, then I'd advise that | 13:01 |
| oanson | Yes. The deployment issue if my first priority, currently. | 13:01 |
| odyssey4me | that'll buy time to figure out the fixes and implement them one at a time, with backports | 13:02 |
| oanson | Yes. | 13:02 |
| oanson | Some of these issues are real dragonflow issues. | 13:02 |
| evrardjp | I think we all agree | 13:02 |
| evrardjp | oanson: yes, because it used to pass, and at one point (hack week in dragonflow before feature freeze?) it stopped working | 13:03 |
| odyssey4me | ok, so in my mind if it deploys and all the services are running, that's good enough for the feature to land in an experimental status | 13:03 |
| odyssey4me | everything that comes after that is a bug fix | 13:03 |
| oanson | I'll definitely get it done by the beginning of next week. I'll try today, but I can't promise | 13:03 |
| evrardjp | odyssey4me: we have to remember that we depend on upstream, so if the deploy itself works, we could consider it as "ok to integrate" | 13:03 |
| odyssey4me | evrardjp yep, that's what I mean | 13:03 |
| evrardjp | ok | 13:04 |
| *** rob0t_ has joined #openstack-ansible | 13:04 | |
| odyssey4me | oanson it'll be up to andymccr if we can allow it to be ported back to ocata if it doesn't make the branch cut which will be happening this week or next | 13:04 |
| evrardjp | also the docs change can be done later, and logan- and I had about the same way to configure it | 13:04 |
| evrardjp | so the review of the docs gonna be easy | 13:04 |
| odyssey4me | considering this is an opt-in feature, I would support a backport personally | 13:05 |
| evrardjp | integration in OSA main directly shouldn't be that hard | 13:05 |
| logan- | ping me if you need a hand wiring up stuff to the AIO oanson. i will be around most of the day | 13:05 |
| oanson | Thanks. I'll do my best. | 13:07 |
| mgariepy | morning everyone ! | 13:08 |
| *** rob0t_ has quit IRC | 13:09 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 13:11 |
| mgariepy | andymccr, sad for nova. | 13:11 |
| mgariepy | did your goat thing fixed the issue ? haha | 13:12 |
| *** rob0t_ has joined #openstack-ansible | 13:12 | |
| andymccr | mgariepy: yeah :( it was just becomming impossible to get anything merged in when it needs 2 x successful gates and 50% chance on either! | 13:12 |
| mgariepy | yeah I understand. | 13:12 |
| andymccr | ive asked mhayden to take a look at the gate times for CentOS to see if we can do something. also the infrastructure is quite random, sometimes its fast (45 mins or so) and passes no problem, others its timing out. | 13:13 |
| mgariepy | while it's 45 minutes for centos it's mostly only 30 for ubunut.. | 13:13 |
| andymccr | hoping to use ara to debug that a bit | 13:16 |
| *** rob0t_ has quit IRC | 13:16 | |
| mgariepy | on centos I started to see a lot of warning in ansible. | 13:17 |
| mgariepy | le ssh connection gets closed and it needs to re-open it. | 13:17 |
| mgariepy | every time it happens it take like a second. | 13:18 |
| odyssey4me | mgariepy not sure if this will help, but perhaps we should try and get the lxc compilation to be packaged | 13:18 |
| mgariepy | lxc-compile is not that long. | 13:18 |
| odyssey4me | last I discussed it with cloudnull he said the packages available did not have the same CLI options | 13:18 |
| mgariepy | the version compiled is like 1.x | 13:18 |
| odyssey4me | hmm, bad ssh connections - that's weird | 13:19 |
| mgariepy | also the ceilometer restart failing takes 5 minutes. | 13:19 |
| odyssey4me | argh | 13:19 |
| mgariepy | ceilomter-collector | 13:19 |
| mgariepy | both ubuntu and centos. | 13:19 |
| mgariepy | for the ocata release, i've looked up as far as I could, and it was like that back december also. | 13:20 |
| *** Jack_Iv has quit IRC | 13:20 | |
| *** rob0t_ has joined #openstack-ansible | 13:22 | |
| odyssey4me | mgariepy yeah, I remember discussion about it | 13:24 |
| odyssey4me | I think even for the newton release we struggled with it | 13:25 |
| mgariepy | newton is ok today. | 13:25 |
| *** jheroux has joined #openstack-ansible | 13:26 | |
| *** rob0t_ has quit IRC | 13:26 | |
| mgariepy | every tasks on centos takes longer lol | 13:27 |
| mgariepy | http://logs.openstack.org/74/426774/1/check/gate-openstack-ansible-os_nova-ansible-func-centos-7-nv/9a3d80a/logs/ara/playbook/127f5ae2-c39a-4aa1-a89a-da620cfa75aa/index.html#plays | 13:27 |
| mgariepy | http://logs.openstack.org/74/426774/1/check/gate-openstack-ansible-os_nova-ansible-func-ubuntu-xenial/4de7d45/logs/ara/playbook/9e85d19a-bca4-442f-87cb-0b04b1c1dbcf/index.html | 13:27 |
| mgariepy | exceot destroy container haha | 13:28 |
| mgariepy | for obvious reason. | 13:28 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Implementation Neutron SR-IOV https://review.openstack.org/415907 | 13:29 |
| mgariepy | is there a reason why both tests don't run on the same hardware ? | 13:29 |
| mgariepy | the flavor is the same but the cpu aren't. | 13:29 |
| *** markvoelker_ has joined #openstack-ansible | 13:29 | |
| Andrew_jedi | evrardjp: I have a 4 node setup. Controller, Compute, Network and Cinder nodes in OpenStack VMs. All VMs have two NICs, ens3 and ens4. I am using vxlan for configuring all the OSA bridges and all the bridge traffic is passing through "ens3" except for br-vlan. The br-vlan is configured on Network and Compute node. vxlan and Flat network are used. On the Network node i have used allowed... | 13:30 |
| Andrew_jedi | ...address pair to allow "0.0.0.0/24". | 13:30 |
| andymccr | mgariepy: i think it depends which cloud provider you get - so its kinda a "luck" thing sometimes :/ | 13:30 |
| mgariepy | 2017-02-03 13:30:28.890698 | [WARNING]: sftp transfer mechanism failed on [172.29.236.100]. Use | 13:30 |
| mgariepy | 2017-02-03 13:30:28.890780 | ANSIBLE_DEBUG=1 to see detailed information | 13:30 |
| mgariepy | that's the issue I have on centos. | 13:31 |
| *** crushil has quit IRC | 13:32 | |
| *** markvoelker has quit IRC | 13:32 | |
| Andrew_jedi | evrardjp: Instance MTU is 1350, here is the network configuration http://paste.openstack.org/show/597559/ | 13:32 |
| mgariepy | do you have a aio running on ubuntu ? | 13:32 |
| mgariepy | what is brctl show like ? | 13:32 |
| *** woodard has joined #openstack-ansible | 13:33 | |
| *** hfu has joined #openstack-ansible | 13:33 | |
| *** hfu has quit IRC | 13:33 | |
| *** rob0t_ has joined #openstack-ansible | 13:34 | |
| *** woodard has quit IRC | 13:34 | |
| odyssey4me | all cores, please see http://lists.openstack.org/pipermail/openstack-dev/2017-February/111585.html | 13:34 |
| *** woodard has joined #openstack-ansible | 13:34 | |
| mgariepy | omg, didn't expect that :D | 13:35 |
| Andrew_jedi | evrardjp: Some more info about namespaces and neutron http://paste.openstack.org/show/597560/ | 13:38 |
| *** rob0t_ has quit IRC | 13:39 | |
| asettle | mgariepy: I think that means that odyssey4me likes you :p | 13:39 |
| asettle | +1 from me dude :) | 13:40 |
| mgariepy | hehe | 13:40 |
| *** rob0t_ has joined #openstack-ansible | 13:42 | |
| *** wdirons has quit IRC | 13:43 | |
| evrardjp | yeah +1 ofc | 13:43 |
| andymccr | hmm that ceilometer collector thing is interesting - its not happening like that in role gate. | 13:43 |
| andymccr | it seems to get a shutdown that takes ages and then it starts up again | 13:43 |
| mgariepy | yeah on the ceilometer tests it's ok, | 13:44 |
| mgariepy | difference is that all ceilometer runs in the same container on that check | 13:44 |
| andymccr | mgariepy: ahh - true. well reading their docs we should be running ceilometer on the hosts themselves. i think this role needs a bit of a refresher | 13:45 |
| andymccr | *compute-hosts | 13:45 |
| andymccr | the polling agent at least | 13:45 |
| *** Jolrael has quit IRC | 13:46 | |
| mgariepy | maybe it's an order thing. | 13:46 |
| *** rob0t_ has quit IRC | 13:47 | |
| *** Jolrael has joined #openstack-ansible | 13:47 | |
| andymccr | i think we do need a way to set an order for restarts, this may be an example of that. | 13:48 |
| *** Hosam_ has quit IRC | 13:49 | |
| odyssey4me | andymccr yeah, that's going to become more and more important for services that have multiple daemons | 13:50 |
| odyssey4me | it's definitely a requirement for doing rolling upgrades in nova | 13:50 |
| logan- | https://github.com/ceph/ceph-ansible/commit/40a2df5bbf13685c83ddc0b784b97e8a76414006 ceph-ansible just started doing some handler serialization this week, the delegation stuff there might work as a stopgap until ansible can get serialization as a task level parameter | 13:52 |
| *** crushil has joined #openstack-ansible | 13:53 | |
| logan- | but I guess that's not gonna happen looking at https://github.com/ansible/ansible/issues/12170 | 13:53 |
| fxpester | hi all, my dib started to act strange, it is not taking variables in work anymore | 13:54 |
| *** esberglu has joined #openstack-ansible | 13:54 | |
| fxpester | I run `export DIB_YUM_REPO_CONF=redhat.repo && /usr/bin/disk-image-create`but it says `no repo file passed` | 13:55 |
| fxpester | and I run dib with -t raw, yet it is generating qcow2 for me | 13:55 |
| fxpester | and I pass image name rhel_ironic, but output images named image.* | 13:56 |
| fxpester | how can it be ?:D | 13:56 |
| *** rob0t_ has joined #openstack-ansible | 13:56 | |
| *** kberger has joined #openstack-ansible | 13:56 | |
| *** esberglu has quit IRC | 13:59 | |
| *** crushil has quit IRC | 14:00 | |
| *** rob0t_ has quit IRC | 14:01 | |
| *** nbante has joined #openstack-ansible | 14:02 | |
| *** rob0t_ has joined #openstack-ansible | 14:06 | |
| *** wdirons has joined #openstack-ansible | 14:07 | |
| *** cathrichardson has joined #openstack-ansible | 14:08 | |
| *** drifterza has joined #openstack-ansible | 14:09 | |
| fxpester | I feel like is it a bug in element `rhel ` | 14:09 |
| *** rob0t_ has quit IRC | 14:10 | |
| *** jwitko has quit IRC | 14:10 | |
| *** cathrich_ has quit IRC | 14:11 | |
| fxpester | or not | 14:11 |
| *** crushil has joined #openstack-ansible | 14:12 | |
| fxpester | `/usr/bin/disk-image-create --install-type package --ramdisk-element dracut-ramdisk -x -x -t -o rhel6_ironic -p lvm2 -c rhel baremetal dhcp-all-interfaces rhel-common` | 14:12 |
| *** rob0t_ has joined #openstack-ansible | 14:12 | |
| cloudnull | mornings | 14:12 |
| fxpester | 2017-02-03 09:11:34.806 ERROR diskimage_builder.element_dependencies [-] Element 'rhel6_ironic' not found in '/usr/bin/../share/diskimage-builder/elements' | 14:12 |
| cloudnull | fxpester: you running this w/ openstack-ansible ? | 14:13 |
| cloudnull | maybe see if there were changes in DIB in #openstack-dib ? | 14:13 |
| fxpester | epic fail, whole time the wrong channel) | 14:13 |
| cloudnull | it happens :) | 14:13 |
| *** jwitko has joined #openstack-ansible | 14:13 | |
| *** gouthamr has joined #openstack-ansible | 14:14 | |
| *** jperry has joined #openstack-ansible | 14:14 | |
| *** strattao_ has joined #openstack-ansible | 14:15 | |
| *** rob0t_ has quit IRC | 14:17 | |
| odyssey4me | cloudnull jamesdenton BjoernT logan- nice to see the SRIOV work merged now | 14:18 |
| odyssey4me | it'd be super-great to somehow get an external-CI test for it | 14:18 |
| *** McMurlock1 has joined #openstack-ansible | 14:19 | |
| odyssey4me | I wonder if adreznec / kylek3h_away could make that happen ;) | 14:20 |
| *** drifterza has quit IRC | 14:20 | |
| logan- | would be good to add to the ptg agenda to discuss during that osic ci item. iirc that is going to be bare metal | 14:20 |
| *** Jack_Iv has joined #openstack-ansible | 14:21 | |
| odyssey4me | logan- ah, good point - it might be possible for that team to extend their CI tests beyond just upgrades | 14:21 |
| odyssey4me | luzC ^ | 14:21 |
| *** rob0t_ has joined #openstack-ansible | 14:22 | |
| *** TxGirlGeek has joined #openstack-ansible | 14:22 | |
| *** bauruine has joined #openstack-ansible | 14:24 | |
| *** Jack_Iv has quit IRC | 14:25 | |
| *** rob0t_ has quit IRC | 14:27 | |
| openstackgerrit | Merged openstack/openstack-ansible master: corrected infra wheel mirror URL https://review.openstack.org/428558 | 14:28 |
| odyssey4me | I would recommend that we hold back any integrated build rechecks/+w's until https://review.openstack.org/428212 merges | 14:28 |
| odyssey4me | otherwise andymccr is going to have a very, very late night | 14:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: corrected infra wheel mirror URL https://review.openstack.org/428760 | 14:29 |
| odyssey4me | jmccrory ping me when you're around to chat about the mitaka tests | 14:29 |
| *** esberglu has joined #openstack-ansible | 14:31 | |
| *** esberglu has quit IRC | 14:31 | |
| *** esberglu has joined #openstack-ansible | 14:32 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Fix SHA's for rc1 + SHA bump https://review.openstack.org/428212 | 14:32 |
| *** rob0t_ has joined #openstack-ansible | 14:32 | |
| *** esberglu_ has joined #openstack-ansible | 14:33 | |
| mgariepy | gathering fact with a 300 router deployed is quite long.. on neutron-agent.. | 14:35 |
| *** jperry has quit IRC | 14:36 | |
| *** esbergl__ has joined #openstack-ansible | 14:37 | |
| *** rob0t_ has quit IRC | 14:37 | |
| *** esberglu has quit IRC | 14:37 | |
| *** esberglu has joined #openstack-ansible | 14:38 | |
| *** esberglu_ has quit IRC | 14:38 | |
| *** shausy has quit IRC | 14:39 | |
| odyssey4me | mgariepy haha, yeah | 14:39 |
| odyssey4me | if we don't need the facts, maybe more filtering is needed | 14:39 |
| cloudnull | mgariepy: yea I've seen that be miserable in the OSIC. | 14:40 |
| mgariepy | haha | 14:40 |
| cloudnull | though I've not tackled how to solve it | 14:40 |
| odyssey4me | we can filter the fact gathering - the question is whether the facts are needed for the role execution | 14:41 |
| *** esbergl__ has quit IRC | 14:41 | |
| *** esberglu_ has joined #openstack-ansible | 14:41 | |
| mgariepy | can we filter like get eth? info but not tap and other stuff like that ? | 14:42 |
| odyssey4me | mgariepy filter may not improve speed, but gather_subset may: http://docs.ansible.com/ansible/setup_module.html | 14:43 |
| odyssey4me | I think the filter gathers them, then filters the return | 14:43 |
| odyssey4me | whereas gather_subset filters the request for facts | 14:43 |
| evrardjp | I didn't follow your conversation but if facts suck, just use another fact system or make your own facts :p | 14:44 |
| odyssey4me | currently we do reduce the subset: https://github.com/openstack/openstack-ansible/blob/master/scripts/openstack-ansible.rc#L32 | 14:44 |
| *** esberglu has quit IRC | 14:44 | |
| adreznec | evrardjp: are you saying we should gather alternative facts? | 14:44 |
| odyssey4me | but maybe it could be reduced more | 14:44 |
| odyssey4me | hahaha adreznec | 14:44 |
| odyssey4me | adreznec any possibility for the IBM CI tests to report back as an external CI? | 14:45 |
| odyssey4me | (and perhaps to have an SRIOV scenario)? | 14:45 |
| *** lucas_ has joined #openstack-ansible | 14:45 | |
| adreznec | Hmm yeah, that's something we've been meaning to switch over from the PowerVM side | 14:46 |
| adreznec | Have gotten a bit constrained on CI resources since we're running double CI for Nova now (out-of-tree and in-tree) | 14:46 |
| adreznec | None of our CI hardware has SRIOV in it today though | 14:46 |
| odyssey4me | adreznec any feedback is good, even if it comes after merge | 14:47 |
| *** rob0t_ has joined #openstack-ansible | 14:48 | |
| adreznec | odyssey4me: I'll have to sync up with some of the other Power CI maintainers internally, see if any of them have SRIOV capable hardware | 14:48 |
| odyssey4me | thanks adreznec | 14:49 |
| odyssey4me | just powervm feedback would be nice though :) | 14:49 |
| adreznec | odyssey4me: But yeah, we'll definitely look at getting the powervm driver feedback publish externally | 14:49 |
| adreznec | I'll hit up the guys who were working on that and hopefully have things to report in Atlanta | 14:50 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/mitaka: Use same branch for upper constraints when testing https://review.openstack.org/428297 | 14:50 |
| openstackgerrit | Merged openstack/openstack-ansible-rsyslog_server master: Add Vagrantfile to allow for easier testing https://review.openstack.org/428332 | 14:50 |
| Mr_Smurf | cloudnull: OPENSTACK_KEYSTONE_URL is apperently used as default region when multiple regions are configured (bug 1660344) | 14:51 |
| openstack | bug 1660344 in openstack-ansible "OPENSTACK_KEYSTONE_URL missing when horizon_available_regions is defined." [Undecided,New] https://launchpad.net/bugs/1660344 | 14:51 |
| cloudnull | Mr_Smurf: hum, is that newton ? | 14:53 |
| Mr_Smurf | cloudnull: yes | 14:54 |
| Mr_Smurf | cloudnull: I added a comment in the bug with some more logs | 14:54 |
| cloudnull | I've a multi-region mitaka cloud up and i dont think i've had an issue with that but I remember there being some changes in newton for horizon and maybe this is one of them | 14:54 |
| *** rob0t_ has quit IRC | 14:56 | |
| Mr_Smurf | cloudnull: if you look at the code on line 101 https://github.com/openstack/django_openstack_auth/blob/stable/newton/openstack_auth/forms.py#L101 | 14:59 |
| Mr_Smurf | cloudnull: you can see that it fetches OPENSTACK_KEYSTONE_URL before fetching the list of AVALIABLE_REGIONS | 15:00 |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in network config - removes hardcoded values https://review.openstack.org/428770 | 15:00 |
| *** chris_hultin|AWA is now known as chris_hultin | 15:01 | |
| *** whiteveil has joined #openstack-ansible | 15:01 | |
| mhayden | andymccr: i'm off today, but i dug into some of the centos stuff last night | 15:01 |
| *** weezS has joined #openstack-ansible | 15:01 | |
| mhayden | fastestmirror could be part of it, but cloudnull noted that we might not be using the best yum mirrors for the gate jobs | 15:01 |
| *** pjh03 has joined #openstack-ansible | 15:01 | |
| odyssey4me | mhayden we're using whatever infra gives us | 15:02 |
| odyssey4me | which should be their mirrors | 15:02 |
| *** klamath has joined #openstack-ansible | 15:02 | |
| mhayden | what's odd is that the tasks that don't use the network are fairly close in comparison w/ubuntu | 15:02 |
| odyssey4me | although I suppose we're adding RDO and EPEL, so perhaps we need to also adapt the vars to ensure we use infra repositories for those | 15:02 |
| mhayden | but anything that uses the network (like pip / yum) is quite slow | 15:02 |
| cloudnull | odyssey4me: mhayden: there's not a tun of dev that happens on cent so I suspect that there's less caching within the cent image making some of the installation times slower | 15:02 |
| mhayden | possibly | 15:03 |
| mgariepy | rdo and epel is already installed. | 15:03 |
| mhayden | but we don't install that much from RDO/EPEL | 15:03 |
| *** whiteveil has quit IRC | 15:03 | |
| *** rob0t_ has joined #openstack-ansible | 15:03 | |
| mhayden | the bulk comes from standard centos/updates mirrors | 15:03 |
| odyssey4me | well, that means that we should work with infra to improve that | 15:03 |
| odyssey4me | I'm sure the puppet and kolla communities would enjoy the fruits of that labour | 15:04 |
| *** klamath has quit IRC | 15:04 | |
| *** klamath has joined #openstack-ansible | 15:04 | |
| mhayden | hopefully it's not too labourous | 15:05 |
| mhayden | labouring isn't my favourite | 15:05 |
| mhayden | but it can be a colourful activity | 15:05 |
| * mhayden waits for it | 15:05 | |
| *** pjh03 has quit IRC | 15:06 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_horizon master: ensure OPENSTACK_KEYSTONE_URL is always configured https://review.openstack.org/428772 | 15:06 |
| cloudnull | Mr_Smurf: | 15:06 |
| cloudnull | ^ | 15:06 |
| Mr_Smurf | cloudnull: I'm just checking one thing | 15:08 |
| mgariepy | mhayden, Setting nodepool variablestakes 6 sec on centos while it takes only 2 on ubuntu. | 15:08 |
| *** rob0t_ has quit IRC | 15:08 | |
| Mr_Smurf | cloudnull: It could still be a bug in horizon... | 15:08 |
| mhayden | that's peculiar | 15:08 |
| mgariepy | centos-7-osic-cloud1-disk-7100150 vs ubuntu-xenial-osic-cloud1-s3700-71001 | 15:09 |
| *** smatzek has joined #openstack-ansible | 15:09 | |
| *** chris_hultin is now known as chris_hultin|AWA | 15:10 | |
| *** pcaruana has quit IRC | 15:11 | |
| logan- | from what ive seen the osic disk vs ssd flavors run almost the same gate times fwiw | 15:11 |
| *** acormier has joined #openstack-ansible | 15:11 | |
| mhayden | even nova gates are 34m for ubuntu and > 65m for centos | 15:12 |
| mhayden | that's unusual | 15:12 |
| Mr_Smurf | cloudnull: it looks like a bug in horizon.. L101 should probably be getattr(settings, "OPENSTACK_KEYSTONE_URL", "Default Region") | 15:14 |
| Mr_Smurf | cloudnull: but the most simple solution for now is just to force it to be set anyway | 15:14 |
| *** ssmith has joined #openstack-ansible | 15:15 | |
| *** pjh03 has joined #openstack-ansible | 15:15 | |
| Mr_Smurf | cloudnull: well "Defalut Region" is still not a valid keystone URL so I dont know.. ;) | 15:17 |
| *** pcaruana has joined #openstack-ansible | 15:17 | |
| *** ssmith_ has joined #openstack-ansible | 15:18 | |
| *** pjh03 has quit IRC | 15:19 | |
| *** rob0t_ has joined #openstack-ansible | 15:19 | |
| *** adrian_otto has joined #openstack-ansible | 15:20 | |
| *** ssmith has quit IRC | 15:21 | |
| *** marst has quit IRC | 15:21 | |
| *** strattao_ has quit IRC | 15:23 | |
| *** adrian_otto has quit IRC | 15:25 | |
| *** woodard has quit IRC | 15:25 | |
| *** rob0t_ has quit IRC | 15:26 | |
| luzC | odyssey4me hello, yes, OSIC CI is on 22 baremetal nodes. I'm not very familiar with SRIOV, but let's talk about it on ptg and let me include dankolbrs and kvcobb on the loop | 15:27 |
| *** strattao_ has joined #openstack-ansible | 15:28 | |
| *** whiteveil has joined #openstack-ansible | 15:28 | |
| *** chris_hultin|AWA is now known as chris_hultin | 15:29 | |
| *** strattao_ has quit IRC | 15:29 | |
| *** strattao_ has joined #openstack-ansible | 15:30 | |
| *** marc_ab has joined #openstack-ansible | 15:30 | |
| *** unlaudable has quit IRC | 15:33 | |
| luzC | odyssey4me or before ptg :) | 15:33 |
| *** rob0t_ has joined #openstack-ansible | 15:37 | |
| *** cmart has joined #openstack-ansible | 15:39 | |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in proxy role - Ubuntu 16.04 https://review.openstack.org/428782 | 15:39 |
| *** marst has joined #openstack-ansible | 15:41 | |
| *** rob0t_ has quit IRC | 15:42 | |
| *** admin0_ has quit IRC | 15:43 | |
| *** DanyC has quit IRC | 15:43 | |
| *** unlaudable has joined #openstack-ansible | 15:48 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/mitaka: Bump requirements SHA https://review.openstack.org/428511 | 15:48 |
| *** rob0t_ has joined #openstack-ansible | 15:52 | |
| *** unlaudable has quit IRC | 15:54 | |
| *** whiteveil has quit IRC | 15:55 | |
| *** rob0t_ has quit IRC | 15:57 | |
| *** ssmith__ has joined #openstack-ansible | 15:58 | |
| *** unlaudable has joined #openstack-ansible | 15:58 | |
| *** whiteveil has joined #openstack-ansible | 16:02 | |
| *** ssmith_ has quit IRC | 16:02 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible master: Fix haproxy playbook group target https://review.openstack.org/428787 | 16:04 |
| *** crushil has quit IRC | 16:06 | |
| *** rob0t_ has joined #openstack-ansible | 16:06 | |
| *** crushil has joined #openstack-ansible | 16:06 | |
| *** unlaudable has quit IRC | 16:07 | |
| *** sdake_ has joined #openstack-ansible | 16:08 | |
| *** fxpester has quit IRC | 16:08 | |
| *** sdake has quit IRC | 16:08 | |
| *** sdake has joined #openstack-ansible | 16:09 | |
| *** Andrew_jedi has quit IRC | 16:10 | |
| *** galstrom_zzz is now known as galstrom | 16:11 | |
| *** rob0t_ has quit IRC | 16:11 | |
| *** sdake has quit IRC | 16:12 | |
| *** sdake_ has quit IRC | 16:12 | |
| openstackgerrit | Frank Zhang proposed openstack/openstack-ansible-os_horizon master: Add retries for horizon venv download https://review.openstack.org/428794 | 16:14 |
| *** pjh03 has joined #openstack-ansible | 16:16 | |
| jwitko | hey guys | 16:17 |
| jwitko | is there a way to restart specific LXC containers without rebooting the host? | 16:17 |
| jwitko | I've tried restarting them in the past but they don't seem to come up the exact way that openstack-ansible brings them up, not sure what I'm doing wrong | 16:17 |
| *** dankolbrs has quit IRC | 16:17 | |
| *** pjh03 has quit IRC | 16:18 | |
| *** pjh03 has joined #openstack-ansible | 16:18 | |
| logan- | lxc-attach -n <container> -- reboot or lxc-stop -n <container>; lxc-start -n <container> should work | 16:18 |
| palendae | jwitko: Can use ansible's ad hoc commands and groups to target them. https://thoughtstreams.io/palendae/openstack-ansible-tips-and-tricks/9015/ | 16:18 |
| *** Oku_OS is now known as Oku_OS-away | 16:18 | |
| jwitko | thank you both! | 16:18 |
| palendae | If you're on Newton+ you can target the lxc_hosts group (though there's a bug in that group I'm working on) | 16:19 |
| jwitko | unfortunately this old ass cluster is on Kilo | 16:19 |
| jwitko | trying to get it into shape for an upgrade | 16:19 |
| *** dankolbrs has joined #openstack-ansible | 16:19 | |
| *** rob0t_ has joined #openstack-ansible | 16:19 | |
| *** unlaudable has joined #openstack-ansible | 16:19 | |
| palendae | Ah | 16:20 |
| palendae | Yeah; you could probably pare the "hosts:is_metal" group down a bit by adding ":!nova_compute" | 16:20 |
| palendae | And run your lxc command | 16:20 |
| palendae | Bash command loop in an ad hoc command is doable | 16:21 |
| palendae | Not pretty, but doable | 16:21 |
| jwitko | thanks, appreciate it | 16:21 |
| jwitko | this is just for a single container, wanted to make sure i was rebooting it correctly | 16:21 |
| palendae | Cool | 16:21 |
| jwitko | i have a new compute node which is dropping connections to rabbitmq containers instantly | 16:21 |
| palendae | Yeah, logan's commands in an ad-hoc command (lxc-stop/start) should work | 16:21 |
| jwitko | the only errors coming from rabbitmq are seen here http://cdn.pasteraw.com/izh0zi9bb4g291vr1wb2ug9tcojnnii | 16:22 |
| *** Andrew_jedi has joined #openstack-ansible | 16:22 | |
| oanson | Hi. I found one issue with the dragonflow tests (os_neutron role) - the IP addresses in {{ neutron_local_ip }} is 127.0.0.1. Is there a way to get the container's management IP? | 16:23 |
| oanson | (For a file deployed inside the container?) | 16:24 |
| *** unlaudable has quit IRC | 16:24 | |
| odyssey4me | oanson that's supposed to be overridden by the playbook IIRC | 16:25 |
| oanson | In the deployement created by tox, I'm looking at the file, and the final address is 127.0.0.1 | 16:25 |
| *** unlaudable has joined #openstack-ansible | 16:26 | |
| *** rob0t_ has quit IRC | 16:26 | |
| oanson | The template takes it from {{ neutron_local_ip }}. | 16:26 |
| jwitko | palendae, one more quick question: is there a way to rebuild the inventory JSON file ? | 16:27 |
| *** agrebennikov__ has joined #openstack-ansible | 16:27 | |
| andymccr | oanson: taking a look quick | 16:27 |
| palendae | jwitko: In kilo, execute playbooks/inventory/dynamic_inventory.py | 16:27 |
| jwitko | thanks | 16:28 |
| oanson | andymccr, thanks | 16:28 |
| palendae | Well, in any series actually | 16:28 |
| palendae | Ansible will call that file every time it runs | 16:28 |
| palendae | But you can do that generate stuff "out of band" | 16:28 |
| *** adrian_otto has joined #openstack-ansible | 16:28 | |
| andymccr | oanson: is that set on localhost or one of the containers? | 16:29 |
| logan- | neutron-overrides-ovs.yml:neutron_local_ip: "{{ ansible_host }}" | 16:29 |
| logan- | i suspect we'll need something like that in the dragonflow test overrides | 16:29 |
| logan- | neutron-overrides-dragonflow.yml | 16:29 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_cinder master: re-added cinder_nfs_client for legacy support https://review.openstack.org/428802 | 16:29 |
| jwitko | palendae, so its a little more complicated than that unfortunately. So I have a running cluster but someone deleted the JSON file | 16:29 |
| cloudnull | andymccr: odyssey4me: ^ | 16:30 |
| jwitko | So running dynamic inventory is generating a new one with new IP allocations and newly generated container names | 16:30 |
| jwitko | this of course is no good | 16:30 |
| palendae | jwitko: Yikes | 16:30 |
| jwitko | is there any way to generate the JSON off an existing infrastructure? | 16:30 |
| palendae | Not that I know of :( | 16:30 |
| cloudnull | jwitko: do you have the openstack_inventory.tar ? | 16:30 |
| logan- | jwitko: do you have /etc/openstack_deploy/backup_openstack_inventory.tar | 16:30 |
| palendae | Oh, right | 16:30 |
| palendae | ^^ | 16:30 |
| cloudnull | if so then there should be a running archive of all json files using the date stamp before the inventory was deleted. | 16:31 |
| oanson | logan-, I'll try that. Thanks | 16:31 |
| logan- | no prob | 16:31 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_cinder master: re-added cinder_nfs_client for legacy support https://review.openstack.org/428802 | 16:32 |
| palendae | cloudnull: Is ^ in reference to http://lists.openstack.org/pipermail/openstack-operators/2017-February/012573.html ? | 16:32 |
| jwitko | cloudnull, logan-, palendae, thank you :) that file exists and it just saved me a massive amount of time | 16:33 |
| palendae | jwitko: Phew! | 16:33 |
| jwitko | you guys are the best | 16:33 |
| andymccr | cloudnull: one comment (on patchset one) | 16:34 |
| palendae | jwitko: In terms of *both* those files missing, you'd have to generate a new JSON file and update the IPs manually | 16:34 |
| jwitko | and the container names! | 16:34 |
| jwitko | but this is awesome, thank you again | 16:34 |
| *** ssmith__ has quit IRC | 16:34 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: [WIP] Combine package install tasks https://review.openstack.org/428806 | 16:34 |
| *** rob0t_ has joined #openstack-ansible | 16:35 | |
| odyssey4me | andymccr cloudnull personally, I think we could just push https://review.openstack.org/428802 to newton and leave master as-is? | 16:36 |
| odyssey4me | then the bug andymccr is mentioning should be sorted out in a separate patch perhaps | 16:36 |
| palendae | Is that patch for the ML question? | 16:36 |
| andymccr | palendae: partly | 16:36 |
| odyssey4me | but yeah - actually maybe a newton/ocata deprecation would be good | 16:36 |
| logan- | nice change mhayden | 16:36 |
| andymccr | palendae: its mostly to fix an issue that wa sintroduced by the multiple nfs support | 16:37 |
| mhayden | logan-: will be nice if it works :P | 16:37 |
| logan- | applying that security role pattern more places is cool | 16:37 |
| andymccr | but also to address the fact we ltierally just changed how it works without any migration path | 16:37 |
| palendae | andymccr: Ah; trying to determine if i need to do any spelunking on inventory for ti | 16:37 |
| andymccr | palendae: i don't thinkso - i think what he's doing should be ok, although the main issue is there is no support for multiple backends. so once he moves to 14.0.8 (and this is fixed) it'll work better | 16:37 |
| palendae | andymccr: Gotcha, thanks | 16:38 |
| andymccr | actually odyssey4me i agree on the basis that the current method is massively broken | 16:38 |
| andymccr | if the current method worked nicely i would say put it in master and backport | 16:38 |
| odyssey4me | andymccr I'm confused. | 16:39 |
| *** rob0t_ has quit IRC | 16:39 | |
| odyssey4me | Are you saying it should be a master patch that fixes the lack of deprecation, and the lack of a working system in one go? | 16:40 |
| odyssey4me | Or are you saying there should be seperate patches for each issue? | 16:40 |
| andymccr | odyssey4me: we should push the patch to newton - and then add a second patch to master (to fix the shares issue), that gets backported | 16:40 |
| odyssey4me | ah yes, makes sense | 16:40 |
| *** unlaudable has quit IRC | 16:41 | |
| odyssey4me | is it just me, or is gerrit slow as molasses | 16:41 |
| cloudnull | odyssey4me: andymccr: so what should I do ? | 16:41 |
| cloudnull | odyssey4me: its being slow | 16:41 |
| openstackgerrit | Frank Zhang proposed openstack/openstack-ansible-os_horizon master: Add retries for horizon venv download https://review.openstack.org/428794 | 16:42 |
| *** rob0t_ has joined #openstack-ansible | 16:42 | |
| andymccr | cloudnull: well what are your thoughts? i mean we could leave it in master too - give people an extra cycle to change, but imo the old way is broken in anycase. | 16:42 |
| mhayden | logan-: i assumed in the nova role that the distro pkg install must be done before the pip install, but i could be wrong | 16:42 |
| *** chyka has joined #openstack-ansible | 16:42 | |
| mhayden | i figured some pip modules might depend on certain pkgs | 16:42 |
| *** weezS has quit IRC | 16:43 | |
| *** DanyC has joined #openstack-ansible | 16:44 | |
| cloudnull | and i've been getting lp failures too. | 16:45 |
| *** DanyC has quit IRC | 16:46 | |
| *** rob0t_ has quit IRC | 16:47 | |
| *** whiteveil has quit IRC | 16:50 | |
| *** DanyC has joined #openstack-ansible | 16:51 | |
| *** adrian_otto has quit IRC | 16:51 | |
| *** ddaskal has joined #openstack-ansible | 16:51 | |
| *** rob0t_ has joined #openstack-ansible | 16:52 | |
| *** whiteveil has joined #openstack-ansible | 16:52 | |
| *** ddaskal has quit IRC | 16:53 | |
| *** weezS has joined #openstack-ansible | 16:54 | |
| *** ssmith__ has joined #openstack-ansible | 16:54 | |
| *** DanyC has quit IRC | 16:55 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 16:55 |
| *** ssmith_ has joined #openstack-ansible | 16:56 | |
| *** McMurlock1 has quit IRC | 16:56 | |
| *** rob0t_ has quit IRC | 16:57 | |
| mhayden | mgariepy: congratulations -- it looks like you're on your way to core! :P | 16:58 |
| *** ssmith__ has quit IRC | 16:59 | |
| mhayden | logan-: "async mode is not supported with the package module" | 17:01 |
| mhayden | :| | 17:01 |
| logan- | :( | 17:01 |
| mgariepy | mhayden, thanks :) | 17:01 |
| logan- | theres still some optimization to be had even if you can't async it I guess. | 17:02 |
| *** rob0t_ has joined #openstack-ansible | 17:02 | |
| logan- | just by consolidating the install tasks | 17:02 |
| *** adrian_otto has joined #openstack-ansible | 17:02 | |
| mhayden | true | 17:02 |
| mhayden | logan-: life finds a way | 17:03 |
| *** lucas_ has quit IRC | 17:03 | |
| *** yolanda has quit IRC | 17:03 | |
| logan- | lol | 17:04 |
| *** whiteveil has quit IRC | 17:04 | |
| mgariepy | mhayden, running AIO setup on 2 vm here centos and ubuntu to see the difference in time. | 17:05 |
| mhayden | cool | 17:05 |
| *** yolanda has joined #openstack-ansible | 17:06 | |
| mgariepy | wow. | 17:06 |
| mgariepy | ubuntu is faster.. | 17:07 |
| *** rob0t_ has quit IRC | 17:07 | |
| *** Jack_Iv has joined #openstack-ansible | 17:08 | |
| *** Jack_Iv has quit IRC | 17:08 | |
| *** marc_ab has quit IRC | 17:09 | |
| mgariepy | let's wait for the end to see.. | 17:10 |
| *** Jack_Iv has joined #openstack-ansible | 17:11 | |
| *** Jack_Iv has quit IRC | 17:11 | |
| mgariepy | mhayden, have you seen : [WARNING]: sftp transfer mechanism failed on [172.29.236.100]. Use ANSIBLE_DEBUG=1 to see detailed information on centos? | 17:11 |
| *** Jack_Iv has joined #openstack-ansible | 17:11 | |
| mhayden | whoa | 17:11 |
| mhayden | i haven't | 17:11 |
| mhayden | i also wondered if we ought to disable GSSAPI support | 17:12 |
| mhayden | that has given me more speed in the past | 17:12 |
| *** rob0t_ has joined #openstack-ansible | 17:12 | |
| *** Jack_Iv has quit IRC | 17:12 | |
| odyssey4me | well, for the AIO bootstrap we give the ssh role this config: https://github.com/openstack/openstack-ansible/blob/master/tests/bootstrap-aio.yml#L51-L89 | 17:12 |
| *** Jack_Iv has joined #openstack-ansible | 17:12 | |
| odyssey4me | I don't know whether you guys are looking at AIO deploy speeds or role deploy speeds | 17:13 |
| odyssey4me | and I don't know whether the role tests use the ssh role to reconfigure the host in the same way | 17:13 |
| odyssey4me | it's entirely plausible that we've been using system defaults for both platforms, and they're very different. Perhaps normalising them in the role tests will help. | 17:14 |
| mgariepy | GSSAPIAuthentication was set to Yes. | 17:14 |
| *** Jack_Iv has quit IRC | 17:14 | |
| *** Jack_Iv has joined #openstack-ansible | 17:14 | |
| cloudnull | in the gate i think we normalize it | 17:15 |
| odyssey4me | cloudnull for the integrated gate, yes - but for the role tests? | 17:15 |
| cloudnull | no i dont think so | 17:15 |
| odyssey4me | yeah, I don't think the role tests do it | 17:15 |
| odyssey4me | perhaps they should | 17:15 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible master: Move scenario variables under vars https://review.openstack.org/428820 | 17:16 |
| *** Jack_Iv_ has joined #openstack-ansible | 17:16 | |
| *** Jack_Iv_ has quit IRC | 17:17 | |
| *** rob0t_ has quit IRC | 17:17 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: [WIP] Combine package install tasks https://review.openstack.org/428806 | 17:17 |
| *** Jack_Iv has quit IRC | 17:20 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_cinder master: NFS backend option "shares" should not be templated https://review.openstack.org/428823 | 17:20 |
| odyssey4me | mhayden is the async attempt even necessary? | 17:21 |
| odyssey4me | if you give the package module a list it installs them all at once, rather than through a loop | 17:22 |
| *** rob0t_ has joined #openstack-ansible | 17:22 | |
| mhayden | odyssey4me: no idea -- mainly testing now to see if it's worth keeping the pattern going | 17:23 |
| mhayden | the "action:" stuff makes me cringe | 17:23 |
| mhayden | but i figured it was worth a try | 17:24 |
| odyssey4me | yeah, I'm hating it | 17:24 |
| odyssey4me | I'd rather then revert to using the apt module if that's the package manager, and the yum module on a yum platform | 17:24 |
| odyssey4me | ie two tasks | 17:24 |
| mhayden | that would be possible, too | 17:24 |
| mhayden | let's see if this helps at all, especially for centos, and then i'll prettify this thing into a reusable pattern | 17:25 |
| odyssey4me | but the question still stands - if you use the map to build the list, does the task do a with_items loop or does it push it all in one go | 17:25 |
| *** marc_ab has joined #openstack-ansible | 17:26 | |
| mhayden | it does it in one go -- already tested that in a sample playbook | 17:26 |
| mhayden | the actual package install task is a single step | 17:26 |
| mhayden | that sum() takes care of that | 17:26 |
| odyssey4me | ok awesome, but the async task doesn't work with it? | 17:27 |
| odyssey4me | and you're hoping that the async task can improve run time | 17:27 |
| mhayden | async doesn't work with the package module | 17:27 |
| *** rob0t_ has quit IRC | 17:27 | |
| mhayden | but it does work with dnf/yum/apt | 17:27 |
| odyssey4me | ah | 17:27 |
| mhayden | it seems like something isn't getting passed through into the pkg module :| | 17:27 |
| odyssey4me | in that case I would say let's have two tasks - one for each pkg manager | 17:27 |
| mhayden | that seems like an easy fix, but it might not be if i examine the code ;) | 17:27 |
| mhayden | odyssey4me: agreed | 17:28 |
| mgariepy | should we copy sshd config into containers ? | 17:29 |
| odyssey4me | mgariepy well, we do adjust the config already in the lxc_hosts role | 17:30 |
| mgariepy | centos: GSSAPIAuthentication yes by default | 17:30 |
| mgariepy | on the host. not in the container | 17:30 |
| odyssey4me | also note that we don't actually ssh to the containers | 17:30 |
| odyssey4me | we ssh to the host, then lxc-attach into the container | 17:30 |
| odyssey4me | our special ssh connection plugin does that for us | 17:30 |
| *** weezS has quit IRC | 17:31 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Fix SHA's for rc1 + SHA bump https://review.openstack.org/428212 | 17:32 |
| andymccr | the goat sacrifices have been accepted | 17:32 |
| logan- | finally | 17:32 |
| *** Jack_Iv has joined #openstack-ansible | 17:32 | |
| logan- | lol | 17:32 |
| andymccr | although we have memory issues etc - which is probably caused by additional nova services now - so aside from long running gate we're gonna need to find a way to reduce memory footprint | 17:33 |
| *** rob0t_ has joined #openstack-ansible | 17:34 | |
| *** Jack_Iv has quit IRC | 17:34 | |
| odyssey4me | andymccr it's likely those new services have worker configs which we haven't implemented | 17:35 |
| andymccr | odyssey4me: its just 1 service, i think the nginx config could be tuned though perhaps | 17:35 |
| andymccr | but yeah you're right that could help | 17:35 |
| odyssey4me | the upstream defaults are quite onerous | 17:36 |
| odyssey4me | we need to set the aio to push it way down | 17:36 |
| odyssey4me | we do that all here: https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/templates/user_variables.aio.yml.j2#L32 | 17:36 |
| odyssey4me | and here https://github.com/openstack/openstack-ansible-tests/blob/master/test-vars.yml#L428 | 17:37 |
| *** rob0t_ has quit IRC | 17:39 | |
| *** galstrom is now known as galstrom_zzz | 17:39 | |
| andymccr | ok so stable/ocata is gonna get branched soon | 17:42 |
| andymccr | at that point we're gonna have to start deving on master but backporting all fixes/patches to stable/ocata | 17:42 |
| andymccr | and we'll have to fix up testing too | 17:42 |
| andymccr | so if all cores/contributors could just keep an eye out for that and backport as appropriate. I'll try do a regular run through but would appreciate any help - that'll happen once the release goes in for RC1 | 17:43 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: [WIP] Combine package install tasks https://review.openstack.org/428806 | 17:45 |
| andymccr | and on that note - i'm out :) enjoy your weekends everybody and thanks for all the help and effort - got there in the end. | 17:48 |
| cloudnull | cheers andymccr | 17:48 |
| odyssey4me | cheers andymccr - I'm out too | 17:48 |
| odyssey4me | have a great w/end y'all! | 17:49 |
| *** adrian_otto1 has joined #openstack-ansible | 17:49 | |
| mhayden | enjoy! | 17:49 |
| *** rob0t_ has joined #openstack-ansible | 17:49 | |
| *** adrian_otto has quit IRC | 17:50 | |
| *** Jack_Iv has joined #openstack-ansible | 17:52 | |
| spotz | seeya andymccr and odyssey4me | 17:53 |
| *** galstrom_zzz has quit IRC | 17:56 | |
| *** d34dh0r53 has quit IRC | 17:56 | |
| *** coolj has quit IRC | 17:56 | |
| *** nwonknu has quit IRC | 17:56 | |
| *** coolj has joined #openstack-ansible | 17:56 | |
| *** agarner has quit IRC | 17:56 | |
| *** jwagner has quit IRC | 17:56 | |
| *** SamYaple has quit IRC | 17:56 | |
| *** NostawRm has quit IRC | 17:56 | |
| *** topol has quit IRC | 17:56 | |
| *** nategraf has quit IRC | 17:56 | |
| *** d34dh0r53 has joined #openstack-ansible | 17:56 | |
| *** NostawRm has joined #openstack-ansible | 17:56 | |
| *** galstrom_zzz has joined #openstack-ansible | 17:56 | |
| *** rob0t_ has quit IRC | 17:56 | |
| *** jowisz_ has quit IRC | 17:56 | |
| *** z- has quit IRC | 17:56 | |
| *** dmsimard has quit IRC | 17:56 | |
| *** SamYaple has joined #openstack-ansible | 17:56 | |
| *** topol_ has joined #openstack-ansible | 17:56 | |
| *** jwagner has joined #openstack-ansible | 17:56 | |
| *** jowisz has joined #openstack-ansible | 17:56 | |
| *** nategraf has joined #openstack-ansible | 17:56 | |
| *** dmsimard has joined #openstack-ansible | 17:56 | |
| *** aleph1 has joined #openstack-ansible | 17:59 | |
| *** kiriakos_ has quit IRC | 18:00 | |
| *** nwonknu has joined #openstack-ansible | 18:02 | |
| *** Andrew_jedi has quit IRC | 18:05 | |
| oanson | andymccr, it's set in a container (openstack1) (Sorry, I missed your message in real time) | 18:05 |
| *** rob0t_ has joined #openstack-ansible | 18:06 | |
| *** Jack_Iv has quit IRC | 18:08 | |
| *** woodard has joined #openstack-ansible | 18:10 | |
| *** rob0t_ has quit IRC | 18:10 | |
| *** pjh03 has quit IRC | 18:13 | |
| *** McMurlock1 has joined #openstack-ansible | 18:15 | |
| openstackgerrit | Rick Box proposed openstack/openstack-ansible-ops master: Adds in Playbook and inventory for Ubuntu 16.04 roles https://review.openstack.org/428844 | 18:16 |
| boxrick1 | cloudnull: I have a whole bunch of commits relating to the promised roles. Would you mind having a look when you have a moment? | 18:18 |
| *** mheremans has quit IRC | 18:18 | |
| *** rob0t_ has joined #openstack-ansible | 18:19 | |
| cloudnull | ++ will have a look asap | 18:21 |
| *** mheremans has joined #openstack-ansible | 18:22 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: [WIP] Combine package install tasks https://review.openstack.org/428806 | 18:23 |
| *** lucas_ has joined #openstack-ansible | 18:23 | |
| *** admin0_ has joined #openstack-ansible | 18:25 | |
| *** whiteveil has joined #openstack-ansible | 18:26 | |
| *** rob0t_ has quit IRC | 18:26 | |
| *** rob0t_ has joined #openstack-ansible | 18:36 | |
| *** rob0t_ has quit IRC | 18:41 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 18:42 |
| *** rob0t_ has joined #openstack-ansible | 18:44 | |
| *** chris_hultin is now known as chris_hultin|AWA | 18:44 | |
| *** weezS has joined #openstack-ansible | 18:46 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 18:47 |
| *** rob0t_ has quit IRC | 18:49 | |
| *** Hosam has joined #openstack-ansible | 18:49 | |
| *** Andrew_jedi has joined #openstack-ansible | 18:51 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 18:51 |
| *** rob0t_ has joined #openstack-ansible | 18:52 | |
| andymccr | ahh oanson: add this line to the neutron-overrides-dragonflow.yml: https://github.com/openstack/openstack-ansible-os_neutron/blob/master/tests/neutron-overrides-ovs.yml#L18 | 18:54 |
| andymccr | that should fix that one | 18:54 |
| mgariepy | hmmm let's see the result. https://review.openstack.org/#/c/407690/23/tests/bootstrap-aio.yml | 18:54 |
| oanson | andymccr, tried that. Came out as 'localhost; | 18:54 |
| oanson | 'localhost'* | 18:54 |
| andymccr | hmm | 18:54 |
| *** marc_ab has quit IRC | 18:54 | |
| oanson | Also tried inventory_hostname. also localhost. Now trying hostvars[inventory_hostname]['ansible_host'] as per https://github.com/ansible/ansible/issues/16139 | 18:55 |
| andymccr | hmm tahts odd i'll take a look later and see if i can figure that out then | 18:55 |
| oanson | Thanks | 18:55 |
| *** sdake has joined #openstack-ansible | 18:56 | |
| *** TxGirlGeek has quit IRC | 18:57 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible master: Update reno for stable/ocata https://review.openstack.org/428862 | 18:57 |
| *** rob0t_ has quit IRC | 18:57 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-apt_package_pinning master: Update reno for stable/ocata https://review.openstack.org/428865 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-pip_install master: Update reno for stable/ocata https://review.openstack.org/428868 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-galera_client master: Update reno for stable/ocata https://review.openstack.org/428871 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-galera_server master: Update reno for stable/ocata https://review.openstack.org/428874 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-ceph_client master: Update reno for stable/ocata https://review.openstack.org/428877 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-haproxy_server master: Update reno for stable/ocata https://review.openstack.org/428880 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-lxc_container_create master: Update reno for stable/ocata https://review.openstack.org/428883 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-lxc_hosts master: Update reno for stable/ocata https://review.openstack.org/428886 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-memcached_server master: Update reno for stable/ocata https://review.openstack.org/428889 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-security master: Update reno for stable/ocata https://review.openstack.org/428892 | 18:57 |
| *** poopcat has joined #openstack-ansible | 18:57 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-openstack_hosts master: Update reno for stable/ocata https://review.openstack.org/428895 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_keystone master: Update reno for stable/ocata https://review.openstack.org/428898 | 18:57 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-openstack_openrc master: Update reno for stable/ocata https://review.openstack.org/428901 | 18:57 |
| *** klamath has quit IRC | 18:58 | |
| *** TxGirlGeek has joined #openstack-ansible | 18:58 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_aodh master: Update reno for stable/ocata https://review.openstack.org/428904 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_barbican master: Update reno for stable/ocata https://review.openstack.org/428907 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_ceilometer master: Update reno for stable/ocata https://review.openstack.org/428910 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_cinder master: Update reno for stable/ocata https://review.openstack.org/428913 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_glance master: Update reno for stable/ocata https://review.openstack.org/428916 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_gnocchi master: Update reno for stable/ocata https://review.openstack.org/428919 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_heat master: Update reno for stable/ocata https://review.openstack.org/428922 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_horizon master: Update reno for stable/ocata https://review.openstack.org/428925 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_ironic master: Update reno for stable/ocata https://review.openstack.org/428928 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_magnum master: Update reno for stable/ocata https://review.openstack.org/428931 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_trove master: Update reno for stable/ocata https://review.openstack.org/428934 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_neutron master: Update reno for stable/ocata https://review.openstack.org/428937 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_nova master: Update reno for stable/ocata https://review.openstack.org/428940 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_rally master: Update reno for stable/ocata https://review.openstack.org/428943 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_sahara master: Update reno for stable/ocata https://review.openstack.org/428946 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_swift master: Update reno for stable/ocata https://review.openstack.org/428949 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_tempest master: Update reno for stable/ocata https://review.openstack.org/428952 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-plugins master: Update reno for stable/ocata https://review.openstack.org/428955 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-rabbitmq_server master: Update reno for stable/ocata https://review.openstack.org/428958 | 18:58 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-repo_build master: Update reno for stable/ocata https://review.openstack.org/428961 | 18:59 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-repo_server master: Update reno for stable/ocata https://review.openstack.org/428964 | 18:59 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-rsyslog_client master: Update reno for stable/ocata https://review.openstack.org/428967 | 18:59 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-rsyslog_server master: Update reno for stable/ocata https://review.openstack.org/428970 | 18:59 |
| openstackgerrit | OpenStack Release Bot proposed openstack/openstack-ansible-os_designate master: Update reno for stable/ocata https://review.openstack.org/428973 | 18:59 |
| *** stuartgr has left #openstack-ansible | 19:01 | |
| *** rob0t_ has joined #openstack-ansible | 19:02 | |
| oanson | My next test will be container_networks.management_address..address | 19:02 |
| *** dalees has quit IRC | 19:02 | |
| openstackgerrit | German Eichberger proposed openstack/openstack-ansible master: Additions to install OpenStack Octavia https://review.openstack.org/422062 | 19:04 |
| *** TxGirlGeek has quit IRC | 19:06 | |
| openstackgerrit | German Eichberger proposed openstack/openstack-ansible-os_neutron master: [WIP] Adds Octavia to Neutron https://review.openstack.org/428979 | 19:06 |
| *** rob0t_ has quit IRC | 19:06 | |
| *** TxGirlGeek has joined #openstack-ansible | 19:07 | |
| openstackgerrit | German Eichberger proposed openstack/openstack-ansible-os_octavia master: [WIP] Octavia role hacking https://review.openstack.org/417210 | 19:08 |
| *** whiteveil has quit IRC | 19:09 | |
| *** rob0t_ has joined #openstack-ansible | 19:12 | |
| *** lucas_ has quit IRC | 19:12 | |
| *** lucas_ has joined #openstack-ansible | 19:12 | |
| *** cmart has quit IRC | 19:14 | |
| *** lucas_ has quit IRC | 19:14 | |
| *** lucas_ has joined #openstack-ansible | 19:15 | |
| *** crushil has quit IRC | 19:15 | |
| *** rob0t_ has quit IRC | 19:17 | |
| *** whiteveil has joined #openstack-ansible | 19:19 | |
| *** TxGirlGeek has quit IRC | 19:20 | |
| *** admin0_ has quit IRC | 19:21 | |
| *** TxGirlGeek has joined #openstack-ansible | 19:21 | |
| *** rob0t_ has joined #openstack-ansible | 19:22 | |
| *** nbante has quit IRC | 19:23 | |
| *** rob0t_ has quit IRC | 19:27 | |
| *** admin0_ has joined #openstack-ansible | 19:27 | |
| *** lucas_ has quit IRC | 19:27 | |
| *** lucas_ has joined #openstack-ansible | 19:27 | |
| *** jwitko has quit IRC | 19:27 | |
| *** admin0_ has quit IRC | 19:29 | |
| *** cmart has joined #openstack-ansible | 19:31 | |
| *** admin0_ has joined #openstack-ansible | 19:31 | |
| *** cmart has quit IRC | 19:35 | |
| *** rob0t_ has joined #openstack-ansible | 19:39 | |
| *** Jack_Iv has joined #openstack-ansible | 19:39 | |
| *** rob0t_ has quit IRC | 19:43 | |
| mgariepy | mhayden, my last build is looking good on centos ;) | 19:45 |
| mgariepy | cinder is done after 44 minutes, was like 67 minutes last time. | 19:46 |
| *** skuda has quit IRC | 19:53 | |
| *** rob0t_ has joined #openstack-ansible | 19:55 | |
| *** adrian_otto1 has quit IRC | 19:55 | |
| *** lucas_ has quit IRC | 19:55 | |
| *** McMurlock1 has quit IRC | 19:56 | |
| *** adrian_otto has joined #openstack-ansible | 19:57 | |
| *** rob0t_ has quit IRC | 19:59 | |
| *** zao has quit IRC | 20:04 | |
| *** lucas_ has joined #openstack-ansible | 20:04 | |
| *** rob0t_ has joined #openstack-ansible | 20:07 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible master: [WIP] Containerize HAProxy https://review.openstack.org/429019 | 20:07 |
| *** lucas_ has quit IRC | 20:08 | |
| cloudnull | logan-: nice! | 20:08 |
| logan- | yes.. but will it blend.. :) | 20:08 |
| *** woodard has quit IRC | 20:09 | |
| *** woodard has joined #openstack-ansible | 20:10 | |
| palendae | Will it proxy | 20:11 |
| *** rob0t_ has quit IRC | 20:11 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible master: [WIP] Add CentOS support for AIO setup https://review.openstack.org/407690 | 20:15 |
| *** zao has joined #openstack-ansible | 20:16 | |
| cloudnull | logan-: I'm a firm believe it will all blend | 20:21 |
| cloudnull | http://www.willitblend.com/ | 20:21 |
| *** rob0t_ has joined #openstack-ansible | 20:22 | |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron master: Implement Dragonflow deployment https://review.openstack.org/391524 | 20:22 |
| *** pramodrj07 has joined #openstack-ansible | 20:24 | |
| *** pramodrj07 has quit IRC | 20:24 | |
| *** MasterOfBugs has joined #openstack-ansible | 20:25 | |
| cloudnull | this is one of my favorites https://www.youtube.com/watch?v=DORyTWjqg7U&list=PL5B2372870FCE5A72&index=9 | 20:25 |
| *** jheroux has quit IRC | 20:26 | |
| *** lucas_ has joined #openstack-ansible | 20:26 | |
| *** rob0t_ has quit IRC | 20:26 | |
| *** rob0t_ has joined #openstack-ansible | 20:32 | |
| *** lucas_ has quit IRC | 20:33 | |
| logan- | haha | 20:33 |
| *** chris_hultin|AWA is now known as chris_hultin | 20:34 | |
| *** lucas_ has joined #openstack-ansible | 20:34 | |
| mgariepy | any reason for this ? https://github.com/openstack/openstack-ansible-openstack_openrc/blob/master/templates/openrc.j2#L2 | 20:36 |
| *** rob0t_ has quit IRC | 20:36 | |
| *** lucas_ has quit IRC | 20:39 | |
| *** adrian_otto has quit IRC | 20:40 | |
| openstackgerrit | Tim Simmons proposed openstack/openstack-ansible-os_designate master: Enable current Designate services https://review.openstack.org/427810 | 20:42 |
| *** rob0t_ has joined #openstack-ansible | 20:42 | |
| *** karimb has quit IRC | 20:42 | |
| *** karimb has joined #openstack-ansible | 20:43 | |
| *** lucas_ has joined #openstack-ansible | 20:44 | |
| *** rob0t_ has quit IRC | 20:47 | |
| *** skuda has joined #openstack-ansible | 20:47 | |
| *** karimb has quit IRC | 20:48 | |
| *** rob0t_ has joined #openstack-ansible | 20:52 | |
| mgariepy | have a nice weekend everyone | 20:55 |
| *** lucas_ has quit IRC | 20:56 | |
| *** gouthamr has quit IRC | 20:59 | |
| *** rob0t_ has quit IRC | 21:01 | |
| csmart | mgariepy: thanks, you too! | 21:03 |
| *** gouthamr has joined #openstack-ansible | 21:05 | |
| *** sdake has quit IRC | 21:06 | |
| *** lucas_ has joined #openstack-ansible | 21:11 | |
| *** Jack_Iv has quit IRC | 21:12 | |
| *** Jack_Iv has joined #openstack-ansible | 21:13 | |
| *** pjh03 has joined #openstack-ansible | 21:14 | |
| *** Jack_Iv_ has joined #openstack-ansible | 21:16 | |
| *** askb has joined #openstack-ansible | 21:18 | |
| *** pjh03 has quit IRC | 21:18 | |
| *** Jack_Iv has quit IRC | 21:19 | |
| *** askb has quit IRC | 21:22 | |
| *** admin0_ has quit IRC | 21:23 | |
| *** smatzek has quit IRC | 21:23 | |
| *** admin0_ has joined #openstack-ansible | 21:23 | |
| *** admin0_ has quit IRC | 21:24 | |
| *** crushil has joined #openstack-ansible | 21:26 | |
| *** skuda has quit IRC | 21:26 | |
| *** Jack_Iv_ has quit IRC | 21:32 | |
| *** Jack_Iv has joined #openstack-ansible | 21:32 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
| *** crushil has quit IRC | 21:35 | |
| *** Jeffrey4l has joined #openstack-ansible | 21:35 | |
| *** whiteveil has quit IRC | 21:45 | |
| *** chris_hultin is now known as chris_hultin|AWA | 21:47 | |
| *** whiteveil has joined #openstack-ansible | 21:48 | |
| *** topol_ is now known as topol | 21:54 | |
| *** lucas_ has quit IRC | 21:54 | |
| agrebennikov__ | qq to whoever is still there... | 21:56 |
| agrebennikov__ | glance api v1 has been deprecated in newton | 21:56 |
| agrebennikov__ | while defaults everywhere until master say glance_enable_v1_api=true | 21:56 |
| agrebennikov__ | do I need to fix it? | 21:56 |
| *** adrian_otto has joined #openstack-ansible | 21:57 | |
| *** gouthamr has quit IRC | 21:57 | |
| agrebennikov__ | cloudnull or evrardjp | 21:57 |
| agrebennikov__ | please | 21:57 |
| *** Jack_Iv_ has joined #openstack-ansible | 21:58 | |
| *** Jack_Iv has quit IRC | 22:01 | |
| *** whiteveil has quit IRC | 22:05 | |
| agrebennikov__ | or odyssey4me maybe... | 22:07 |
| *** karimb has joined #openstack-ansible | 22:14 | |
| *** Andrew_jedi has quit IRC | 22:15 | |
| *** skuda has joined #openstack-ansible | 22:16 | |
| *** karimb has quit IRC | 22:17 | |
| *** Andrew_jedi has joined #openstack-ansible | 22:17 | |
| *** whiteveil has joined #openstack-ansible | 22:21 | |
| *** smatzek has joined #openstack-ansible | 22:24 | |
| *** thorst_ has quit IRC | 22:25 | |
| *** Jack_Iv_ has quit IRC | 22:27 | |
| *** smatzek has quit IRC | 22:29 | |
| *** strattao_ has left #openstack-ansible | 22:31 | |
| cloudnull | agrebennikov__: I believe that the glance api v1 is still in newton and will be removed in ocata? | 22:32 |
| *** ssmith_ has quit IRC | 22:32 | |
| *** klamath has joined #openstack-ansible | 22:32 | |
| cloudnull | are you seeing that option causing issues ? | 22:33 |
| *** sdake has joined #openstack-ansible | 22:33 | |
| *** calebb has quit IRC | 22:34 | |
| *** whiteveil has quit IRC | 22:37 | |
| *** Adri2000 has quit IRC | 22:37 | |
| *** klamath has quit IRC | 22:41 | |
| *** Adri2000 has joined #openstack-ansible | 22:43 | |
| *** gouthamr has joined #openstack-ansible | 22:45 | |
| agrebennikov__ | cloudnull, kind of... at least on one of the newton envs we are deploying glance --os-image-api-version 1 image-list fails | 22:46 |
| *** weezS has quit IRC | 22:47 | |
| agrebennikov__ | and it fails because of the broken communication between glance-registry and glance-api somehow | 22:47 |
| *** calebb has joined #openstack-ansible | 22:47 | |
| agrebennikov__ | because of this reason rally fails on some scenarios | 22:47 |
| agrebennikov__ | if you don't specifically mention you want v2 | 22:48 |
| *** karimb has joined #openstack-ansible | 22:52 | |
| *** chris_hultin|AWA is now known as chris_hultin | 23:02 | |
| *** chris_hultin is now known as chris_hultin|AWA | 23:03 | |
| *** acormier has quit IRC | 23:05 | |
| *** adrian_otto has quit IRC | 23:05 | |
| *** esberglu_ has quit IRC | 23:06 | |
| *** nadeem has joined #openstack-ansible | 23:06 | |
| *** Adri2000 has quit IRC | 23:07 | |
| *** sdake has quit IRC | 23:11 | |
| *** Adri2000 has joined #openstack-ansible | 23:14 | |
| *** acormier has joined #openstack-ansible | 23:17 | |
| *** pmannidi has joined #openstack-ansible | 23:20 | |
| stevelle | agrebennikov__: glance v1 can continue to be enabled in newton and ocata | 23:21 |
| stevelle | it is being given an extended deprecation period | 23:21 |
| agrebennikov__ | stevelle, this is fine, but it just doesn't work seems so :D | 23:22 |
| agrebennikov__ | I'll definitely dig deeper | 23:22 |
| stevelle | agrebennikov__: glance still has bugs open in master (pike) for the api -> registry version issues | 23:22 |
| agrebennikov__ | awsome :/ | 23:23 |
| agrebennikov__ | so why keeping v1 in stable branch if it doesn't work correctly then? | 23:23 |
| stevelle | agrebennikov__: also, heat still working on their conversion | 23:23 |
| *** acormier has quit IRC | 23:23 | |
| stevelle | agrebennikov__: https://review.openstack.org/#/c/416766/ | 23:24 |
| agrebennikov__ | wowww..... | 23:25 |
| agrebennikov__ | this is super cool :D | 23:25 |
| agrebennikov__ | really lol | 23:25 |
| *** pmannidi has quit IRC | 23:26 | |
| stevelle | agrebennikov__: v1 does work, you have to configure stuff correctly | 23:26 |
| stevelle | it shouldn't be that hard though | 23:26 |
| agrebennikov__ | agree. but in this case in our stable it should have deep logic of configuring all these 4 things by default | 23:27 |
| agrebennikov__ | because by default all 4 are true | 23:27 |
| *** TxGirlGeek has quit IRC | 23:28 | |
| *** pmannidi has joined #openstack-ansible | 23:29 | |
| *** adrian_otto has joined #openstack-ansible | 23:30 | |
| agrebennikov__ | :q | 23:33 |
| agrebennikov__ | Жй | 23:33 |
| *** skuda has quit IRC | 23:35 | |
| *** karimb has quit IRC | 23:45 | |
| stevelle | agrebennikov__: you lose me | 23:49 |
| stevelle | in our table, deep logic, all 4 things. XD | 23:50 |
| agrebennikov__ | it is stevelle, but guess what happens if you go with default? | 23:52 |
| agrebennikov__ | all them become true | 23:52 |
| agrebennikov__ | and lead to the broken env | 23:53 |
| stevelle | devstack only tests one permutation by default and glance apparently has insufficient functional testing for permutations. | 23:54 |
| *** rollniak has joined #openstack-ansible | 23:55 | |
| agrebennikov__ | any advise? | 23:56 |
| agrebennikov__ | maybe I just disable v1 in defaults then? ;) | 23:57 |
| agrebennikov__ | nah, probably we can't.... since cinder sets up v1 if there is no ceph :/ | 23:58 |
| * agrebennikov__ confused | 23:58 | |
| stevelle | I thought cinder was supposed to support v2 | 23:59 |
| stevelle | heat would have problems with v2 though | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!