*** yifei has joined #openstack-ansible | 00:18 | |
bndzor | takes for ever to run :S | 00:20 |
---|---|---|
*** woodard has joined #openstack-ansible | 00:21 | |
*** woodard has quit IRC | 00:26 | |
bndzor | logan-: pmed you the url | 00:31 |
bndzor | oh, that didnt work | 00:31 |
bndzor | pmed new url | 00:34 |
bndzor | so, i manually installed the rdb python part, now im getting | 01:17 |
bndzor | Error updating resources for node appserver1.openstack.local.: ObjectNotFound: error calling conf_read_file | 01:17 |
bndzor | this is very frustrating | 01:18 |
*** woodard has joined #openstack-ansible | 01:21 | |
*** woodard has quit IRC | 01:25 | |
logan- | the appserver1/2 are your compute_hosts right? | 01:38 |
logan- | down at the bottom of the output you can see the ceph_client role skipping on appserver1/2 | 01:39 |
bndzor | correct | 01:39 |
bndzor | il hjave a check | 01:39 |
logan- | https://github.com/openstack/openstack-ansible-os_nova/blob/c6c05a79a750c921251c16269a9849b606587c35/tasks/main.yml#L112-L123 | 01:39 |
logan- | for some reason the condition there is not being satisfied | 01:39 |
bndzor | hmm | 01:40 |
bndzor | well they are defined | 01:41 |
*** lbragstad has joined #openstack-ansible | 01:45 | |
bndzor | hmm, any hints ? :S | 01:49 |
*** hw_wutianwei has joined #openstack-ansible | 01:55 | |
savvas | logan-: any thoughts on my communication issue between glance / cinder and ceph? | 02:46 |
savvas | Not sure if you caught that earlier | 02:46 |
logan- | savvas: do you see the ceph_client role running during your glance and cinder playbook runs | 02:52 |
savvas | to be honest I haven't noticed, but I can tail the logs if you have a keyword or two for me | 02:52 |
savvas | I'll just run it again, ceph_client produces a lot | 02:55 |
savvas | [root@foreman playbooks]# tail -f /openstack/log/ansible-logging/ansible.log |grep ceph | 02:59 |
savvas | 2017-11-19 21:57:30,590 p=7526 u=root | TASK [ceph_client : include] ******************************************************************************************************************************************************************************************************************************************** | 02:59 |
savvas | 2017-11-19 21:59:12,493 p=7526 u=root | TASK [ceph_client : include] ******************************************************************************************************************************************************************************************************************************************** | 02:59 |
savvas | on glance playbook | 02:59 |
savvas | looks like it didn't execute anything logan- | 03:03 |
savvas | scrolled back in the playbook output and it passed by the task but didn't execute anything | 03:03 |
*** woodard has joined #openstack-ansible | 03:18 | |
*** woodard has quit IRC | 03:22 | |
*** udesale has joined #openstack-ansible | 03:23 | |
savvas | any thoughts logan-? | 03:38 |
*** dave-mccowan has quit IRC | 03:39 | |
savvas | Assuming I am looking at the correct role, the main.yml in /etc/ansible/roles/os_glance/tasks/main.yml checks on existence of the setting for glance_default_store to rbd, which I've set in user_variables.yml | 03:39 |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server master: Tune-up the galera role for efficiency https://review.openstack.org/466827 | 04:07 |
*** nshetty has joined #openstack-ansible | 04:10 | |
*** woodard has joined #openstack-ansible | 04:16 | |
*** woodard has quit IRC | 04:20 | |
logan- | savvas: for testing you could just rip the when condition off the ceph_client include_role task | 04:47 |
savvas | I already did that logan- | 04:54 |
savvas | skips /passes it the same way | 04:54 |
logan- | https://github.com/openstack/openstack-ansible-ceph_client/blob/022373becf228b861f5cbfaa0a33a8e3ece83e86/tasks/main.yml#L16-L22 | 04:55 |
logan- | i wonder if ceph_mons is not populated correctly | 04:55 |
logan- | ah I see | 04:57 |
logan- | https://review.openstack.org/#/c/520588/ | 04:57 |
logan- | this commit is not in pike yet | 04:57 |
logan- | you can set ceph_mons directly in user_variables like this | 04:57 |
logan- | copy/paste these 3 lines into user_variables: https://github.com/openstack/openstack-ansible/blob/0749a6c8b3db60692947d028fb292bc79e856a33/group_vars/all/ceph.yml#L36-L38 | 04:57 |
savvas | checking | 05:00 |
openstackgerrit | Logan V proposed openstack/openstack-ansible master: [WIP] Implement iptables management https://review.openstack.org/512099 | 05:00 |
savvas | logan-: that's the second time I am being hustled by that ;) | 05:01 |
savvas | that did the trick for Glance | 05:09 |
savvas | image created now | 05:09 |
savvas | let me try the same magic for cinder and nova, see if we can get this fully up and runnign now | 05:11 |
*** woodard has joined #openstack-ansible | 05:13 | |
savvas | for cinder it still seems to skip it | 05:16 |
*** woodard has quit IRC | 05:17 | |
savvas | had to comment out the when in main.yml, running now | 05:21 |
*** zerick has joined #openstack-ansible | 05:23 | |
*** hybridpollo has quit IRC | 05:28 | |
*** nshetty is now known as nshetty|afk | 05:48 | |
*** nshetty|afk is now known as nshetty | 05:58 | |
*** threestrands has quit IRC | 06:03 | |
*** gouthamr has quit IRC | 06:20 | |
*** woodard has joined #openstack-ansible | 06:22 | |
*** woodard has quit IRC | 06:26 | |
*** rodolof has joined #openstack-ansible | 06:27 | |
*** jbadiapa has joined #openstack-ansible | 06:35 | |
*** gouthamr has joined #openstack-ansible | 06:54 | |
*** woodard has joined #openstack-ansible | 07:14 | |
*** cshen has joined #openstack-ansible | 07:17 | |
*** woodard has quit IRC | 07:19 | |
*** nshetty is now known as nshetty|lunch | 07:25 | |
*** rodolof has quit IRC | 07:27 | |
*** pmannidi has quit IRC | 07:27 | |
*** rodolof has joined #openstack-ansible | 07:27 | |
evrardjp | morning people! | 07:28 |
*** savvas has quit IRC | 07:29 | |
*** savvas has joined #openstack-ansible | 07:29 | |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible stable/pike: Update all SHAs for 16.0.4 https://review.openstack.org/518829 | 07:38 |
*** mbuil has joined #openstack-ansible | 07:44 | |
*** shardy has joined #openstack-ansible | 07:46 | |
*** eumel8 has joined #openstack-ansible | 07:58 | |
*** sxc731 has joined #openstack-ansible | 07:58 | |
*** ccha has joined #openstack-ansible | 08:00 | |
*** pcaruana has joined #openstack-ansible | 08:04 | |
*** jvidal has joined #openstack-ansible | 08:07 | |
*** SerenaFeng has joined #openstack-ansible | 08:10 | |
*** SerenaFeng has quit IRC | 08:10 | |
*** mardim has joined #openstack-ansible | 08:11 | |
*** woodard has joined #openstack-ansible | 08:16 | |
*** woodard has quit IRC | 08:20 | |
*** nshetty|lunch is now known as nshetty | 08:23 | |
*** gouthamr has quit IRC | 08:25 | |
*** gkadam has joined #openstack-ansible | 08:26 | |
openstackgerrit | Dimitrios Markou proposed openstack/openstack-ansible-os_neutron master: Fix OVS manager configuration inconsistencies which are related to ODL https://review.openstack.org/520367 | 08:29 |
openstackgerrit | Dimitrios Markou proposed openstack/openstack-ansible-os_neutron master: Add hosts to ODL test inventory https://review.openstack.org/518964 | 08:29 |
openstackgerrit | Dimitrios Markou proposed openstack/openstack-ansible-os_neutron master: [DO_NOT_MERGE] Add OvS-NSH support https://review.openstack.org/517259 | 08:29 |
*** m|y|k has joined #openstack-ansible | 08:51 | |
*** michelv has joined #openstack-ansible | 09:08 | |
*** woodard has joined #openstack-ansible | 09:21 | |
*** woodard has quit IRC | 09:25 | |
*** yifei has quit IRC | 09:39 | |
*** tnogisto has joined #openstack-ansible | 09:41 | |
*** michelv has quit IRC | 09:48 | |
*** electrofelix has joined #openstack-ansible | 09:55 | |
evrardjp | is there someone that can handle the bug triage tomorrow? | 09:58 |
*** hw_wutianwei has quit IRC | 10:01 | |
bndzor | anyone having an idea how i can get past the RuntimeError: rbd python libraries not found ? | 10:13 |
*** woodard has joined #openstack-ansible | 10:14 | |
evrardjp | odyssey4me: are you working on https://review.openstack.org/#/c/514289/ right now? I am working on it if not | 10:15 |
evrardjp | bndzor: which version? | 10:15 |
evrardjp | it looks like you don't have the libs in your venv | 10:15 |
bndzor | pike/stable | 10:15 |
bndzor | well im using the ceph example | 10:16 |
evrardjp | https://github.com/openstack/openstack-ansible-ceph_client/blob/stable/pike/tasks/ceph_install_python_libs.yml should have run | 10:16 |
evrardjp | could you check on one of your logs to see if that properly ran? | 10:16 |
bndzor | send you the log output in a pm | 10:17 |
bndzor | logan said some condition was not met | 10:17 |
evrardjp | That doesn't sound right | 10:18 |
evrardjp | looks like it got skipped | 10:18 |
*** woodard has quit IRC | 10:18 | |
evrardjp | yeah you should trust logan- on the ceph bits :p | 10:18 |
evrardjp | which condition? | 10:18 |
evrardjp | https://github.com/openstack/openstack-ansible-os_nova/blob/master/tasks/main.yml#L119-L121 | 10:19 |
evrardjp | https://github.com/openstack/openstack-ansible-os_nova/blob/stable/pike/tasks/main.yml#L119-L121 | 10:19 |
evrardjp | to be precise | 10:19 |
bndzor | ye | 10:19 |
evrardjp | it looks like one of those conditions aren't done properly on your nodes. | 10:19 |
bndzor | on the compute nodes, right | 10:19 |
evrardjp | You could manually add a debug to see what the variables are. | 10:19 |
evrardjp | but it's probably something you should fix | 10:20 |
bndzor | but what could it possibly be, im using the example config as for a test | 10:20 |
bndzor | should that not be enough ? | 10:20 |
evrardjp | I don't know yet, maybe the doc is incomplete? Could you check what's missing? | 10:21 |
bndzor | i wish i knew where to begin :s | 10:22 |
evrardjp | (by using the debug as I said earlier) | 10:22 |
bndzor | How do i enable the debug ? | 10:22 |
evrardjp | edit your /etc/ansible/roles/os_nova/tasks/main.yml add a task to debug all those variables. It will show up in your run, and this way you know why it wasn't included | 10:22 |
evrardjp | but my little finger says it's maybe not your fault, and it could be an ansible bug instead. | 10:23 |
*** flaviosr has quit IRC | 10:23 | |
evrardjp | but I need evidence. | 10:23 |
*** flaviosr has joined #openstack-ansible | 10:25 | |
bndzor | oki | 10:25 |
openstackgerrit | Dimitrios Markou proposed openstack/openstack-ansible-os_neutron master: [DO_NOT_MERGE] Add OvS-NSH support https://review.openstack.org/517259 | 10:26 |
evrardjp | bndzor: is that enough help? | 10:28 |
evrardjp | you should post your debug results here, this way I can know what's wrong | 10:28 |
evrardjp | and know the scale of the bug :) | 10:29 |
*** stuartgr has joined #openstack-ansible | 10:30 | |
bndzor | trying to figure out how to do the debugging first, im very new to ansible | 10:32 |
bndzor | but yeah, i guess :) | 10:32 |
odyssey4me | evrardjp nope, any remaining zuul v3 job setups are free to be worked on | 10:42 |
evrardjp | ok | 10:42 |
odyssey4me | evrardjp You'll be happy to know that I've got a working implementation of https://github.com/openstack/openstack-ansible-specs/blob/master/specs/queens/python-build-install-simplification.rst#L160-L171 setup in a test environment and will be pushing up a review in the next day or two. | 10:44 |
odyssey4me | I've also been liaising with the glance team to get rolling upgrade tests into place, hence my patches to os_glance. | 10:45 |
* odyssey4me got quite a bit done this weekend. :) | 10:45 | |
evrardjp | great! | 10:45 |
evrardjp | I talked to Neutron and Cinder PTLs during the summit and they are not opposed to doing things like that too. | 10:46 |
evrardjp | I will review this soon | 10:48 |
evrardjp | bndzor: if you need help, just ping | 10:50 |
bndzor | evrardjp: ping heh, i dont quite understand how to add it in or what to look for | 10:51 |
evrardjp | bndzor: http://docs.ansible.com/ansible/latest/debug_module.html | 10:51 |
evrardjp | add a task like that in the file I mentioned you, the message should show the variables you need to debug | 10:51 |
mardim | odyssey4me, hwoarang when you find some spare time can you please review this so I can have some feedback https://review.openstack.org/#/c/517259/ | 10:52 |
mardim | odyssey4me, hwoarang Ignore the comment out lines | 10:53 |
mardim | odyssey4me, hwoarang thanks :) | 10:53 |
evrardjp | mardim: how are things going on? | 10:53 |
mardim | evrardjp, I think ok I have some problems with the gates | 10:53 |
mardim | evrardjp, I have two patches which are getting the +2 | 10:54 |
mardim | evrardjp, But I cannot pass the zuul gates for the merging | 10:54 |
odyssey4me | evrardjp you might notice we have a bunch of new lint test failures going on in master - that's due to https://review.openstack.org/520177 doing lint tests on the tests repo playbooks now too, so a few new issues are coming up and some repositories might need some attention. | 10:54 |
mardim | evrardjp, https://review.openstack.org/#/c/518964/ https://review.openstack.org/#/c/520367/ | 10:55 |
odyssey4me | https://review.openstack.org/521357 is an example of a fix set that's needed | 10:55 |
odyssey4me | oh and I saw some apt cache related failures popping up a bit too often, which https://review.openstack.org/521257 and https://review.openstack.org/521252 address | 10:55 |
odyssey4me | also https://review.openstack.org/521247 will help get pike moving again as we're getting opensuse failures for pike without it | 10:56 |
odyssey4me | and I think you'll be happy to see that https://review.openstack.org/517103 is now passing :) | 10:56 |
mardim | evrardjp, Also when you find some spare time can you give me a quick feedback on this https://review.openstack.org/#/c/517259/ thanks :) | 10:56 |
evrardjp | I am glad to be back, I see all those red things and feedbacks to give! | 10:57 |
hwoarang | i'm sure you missed it :) | 10:58 |
evrardjp | lol | 10:59 |
evrardjp | I am heading to lunch now. I will think about all of that when lamenting over a pizza | 11:00 |
*** szaher has quit IRC | 11:02 | |
*** yifei has joined #openstack-ansible | 11:13 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Allow deployment without glance-registry https://review.openstack.org/521251 | 11:16 |
*** m|y|k has quit IRC | 11:18 | |
*** m|y|k has joined #openstack-ansible | 11:20 | |
*** woodard has joined #openstack-ansible | 11:20 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Allow deployment without glance-registry https://review.openstack.org/521251 | 11:22 |
*** woodard has quit IRC | 11:25 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Disable glance-registry and the v1 API by default https://review.openstack.org/521264 | 11:26 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Disable glance-registry and the v1 API by default https://review.openstack.org/521264 | 11:27 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Disable glance-registry and the v1 API by default https://review.openstack.org/521264 | 11:32 |
*** openstackgerrit has quit IRC | 11:32 | |
*** openstackgerrit has joined #openstack-ansible | 11:33 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Implement testing using tempest https://review.openstack.org/521268 | 11:33 |
odyssey4me | evrardjp heh, I knew that would come up - see my response in https://review.openstack.org/#/c/521268/4 | 11:35 |
odyssey4me | bother, I meant https://review.openstack.org/#/c/521357/3/tests/test.yml | 11:35 |
*** szaher has joined #openstack-ansible | 11:51 | |
*** udesale has quit IRC | 11:57 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts master: Make the cache prep timeout configurable https://review.openstack.org/521313 | 12:01 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins master: Resolve ansible-lint test failures https://review.openstack.org/521503 | 12:09 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins master: Add missing domain parameter to _get_group_roles https://review.openstack.org/521283 | 12:09 |
*** yifei has quit IRC | 12:10 | |
*** yifei has joined #openstack-ansible | 12:11 | |
*** woodard has joined #openstack-ansible | 12:20 | |
*** nshetty is now known as nshetty|afk | 12:23 | |
*** woodard has quit IRC | 12:24 | |
evrardjp | odyssey4me: see my answer on https://review.openstack.org/#/c/521268/4 | 12:24 |
odyssey4me | evrardjp the os_image module is used in the os_tempest role | 12:25 |
evrardjp | I'd like to uniformize where it makes sense | 12:25 |
evrardjp | glance -> os_image | 12:25 |
evrardjp | designate -> os_zone or things like that | 12:25 |
odyssey4me | so in other words, it is still tested - just in a different place | 12:26 |
evrardjp | which is less obvious | 12:26 |
odyssey4me | the purpose of that role's tests is not to test the os_image module, so I don't really see why you feel it's reducing test coverage | 12:27 |
evrardjp | nevermind, you don't have the full picture. I'll deal with it later then. | 12:28 |
evrardjp | I'm fine with making things uniform :) | 12:29 |
*** udesale has joined #openstack-ansible | 12:29 | |
odyssey4me | well, having looked into glance specifically it appears that they do not use tempest tests for their gating, but it is used for the integrated gating | 12:30 |
odyssey4me | so I'm trying to understand more from them to figure out our best strategy, but for now at least this is more standard and does increase our coverage | 12:30 |
evrardjp | yeah I am fine with that | 12:31 |
odyssey4me | hwoarang it looks like opensuse is still having a bad day :( http://logs.openstack.org/03/521503/1/check/openstack-ansible-functional-opensuse-423/56bd0a8/host/lxc-cache-prep-commands.log.txt.gz | 12:32 |
odyssey4me | the cache prep is failing quite a bit | 12:32 |
hwoarang | odyssey4me: this is the old problem that is manifesting again because we are trying to talk to external mirrors again | 12:32 |
odyssey4me | ah I see | 12:32 |
odyssey4me | have the issues with the infra mirrors been fixed yet? | 12:33 |
hwoarang | i am trying to source an alternative rsync mirror so i can use that to sync the infra ones | 12:33 |
odyssey4me | or was it the dib that was the issue? | 12:33 |
hwoarang | unfortunately not yet :( | 12:33 |
hwoarang | the main rsync mirror is still down | 12:33 |
odyssey4me | wow, that's bad | 12:34 |
hwoarang | i have asked again but didn't get any replies. i will try after lunch time for EU people | 12:36 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Retry apt cache update https://review.openstack.org/521252 | 12:37 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Retry apt cache update https://review.openstack.org/521257 | 12:38 |
evrardjp | odyssey4me: thanks for the uniformization of the | failed or | changed | 12:43 |
evrardjp | that way the sed/replace will be easier for 2.4 :p | 12:43 |
*** SmearedBeard has joined #openstack-ansible | 12:45 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Allow deployment without glance-registry https://review.openstack.org/521251 | 12:48 |
odyssey4me | evrardjp we try, a little at a time :) | 12:48 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Disable glance-registry and the v1 API by default https://review.openstack.org/521264 | 12:48 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: Implement testing using tempest https://review.openstack.org/521268 | 12:49 |
evrardjp | could you deal with the bug triage tomorrow? | 12:49 |
odyssey4me | sorry, I won't be able to - I'm behind on other work so I won't be able to prep for it adequately | 12:50 |
evrardjp | ok | 12:50 |
evrardjp | anyone else? | 12:50 |
odyssey4me | you'll likely have to ping individually - not everyone has this window in focus ;) | 12:51 |
evrardjp | odyssey4me: @here sometimes IS good. | 12:52 |
*** nshetty|afk is now known as nshetty | 12:52 | |
evrardjp | bndzor: did you manage to make it work? | 12:56 |
evrardjp | logan-: could you ping me when you're back, it's about a ceph role issue, and I am thinking of using Mjolnir to fix it. | 12:57 |
*** sxc731 has quit IRC | 13:03 | |
*** tnogisto has quit IRC | 13:03 | |
*** nshetty has quit IRC | 13:11 | |
*** woodard has joined #openstack-ansible | 13:17 | |
*** tnogisto has joined #openstack-ansible | 13:22 | |
*** woodard has quit IRC | 13:22 | |
*** markvoelker has quit IRC | 13:23 | |
*** markvoelker has joined #openstack-ansible | 13:24 | |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-ceph_client master: Remove one layer of includes https://review.openstack.org/521532 | 13:32 |
evrardjp | logan-: ^ this would need a vote to backport back to Ocata or even N. | 13:35 |
odyssey4me | evrardjp that layer was added there to make the task skip the tasks dynamically | 13:36 |
evrardjp | yeah it breaks stuff. | 13:36 |
*** dave-mccowan has joined #openstack-ansible | 13:36 | |
evrardjp | no choice, without that patch, it just breaks. This is ugly anyway. I don't think we should do it that way, but that's another discussion, and I don't have the cycles to improve it. | 13:37 |
evrardjp | definitely low prio | 13:37 |
evrardjp | while a breakage is higher prio | 13:38 |
*** stuartgr has quit IRC | 13:38 | |
odyssey4me | evrardjp sure, but not doing dynamic includes caused long, long deploy times due to all the skipping | 13:38 |
odyssey4me | so if we can keep it by making all those includes dynamic, then we should be fine | 13:38 |
odyssey4me | I've suggested the changes in review | 13:38 |
*** udesale has quit IRC | 13:42 | |
*** dave-mcc_ has joined #openstack-ansible | 13:46 | |
*** dave-mccowan has quit IRC | 13:48 | |
*** Taseer has quit IRC | 13:56 | |
*** Taseer has joined #openstack-ansible | 13:56 | |
*** rodolof has quit IRC | 14:01 | |
openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Set the nova-lxd test to voting https://review.openstack.org/517103 | 14:02 |
*** taseer1 has joined #openstack-ansible | 14:03 | |
*** sxc731 has joined #openstack-ansible | 14:03 | |
*** Taseer has quit IRC | 14:04 | |
*** taseer2 has joined #openstack-ansible | 14:04 | |
logan- | o/ | 14:04 |
*** stuartgr has joined #openstack-ansible | 14:05 | |
*** woodard has joined #openstack-ansible | 14:05 | |
*** woodard has quit IRC | 14:05 | |
*** woodard has joined #openstack-ansible | 14:06 | |
evrardjp | answered in the review. | 14:08 |
*** taseer1 has quit IRC | 14:08 | |
evrardjp | will do a PoC on my machine based on different versions to see if I am wrong | 14:09 |
logan- | is there any more reading i could do to understand how that patch fixes the recursion loop issue in the log you linked to? | 14:09 |
logan- | it looks like an ansible bug https://github.com/ansible/ansible/issues/23609 | 14:09 |
evrardjp | 2.3.0 agrees with what I thought | 14:09 |
evrardjp | yes it's an ansible bug | 14:09 |
evrardjp | we don't fix the regression | 14:09 |
evrardjp | recursion issue* | 14:09 |
evrardjp | by reducing the level of inclusion you don't trigger that exception | 14:10 |
evrardjp | tested locally, works (R) (TM) | 14:10 |
evrardjp | logan-: this bug has been there for a while, there are other issues that should point to the same kind of things. I remember another issue filed at least. | 14:12 |
evrardjp | if you have any alternative feel free to propose it :) | 14:13 |
evrardjp | TIL: includes with static: yes and automatically static include are behaving differently on the serialized include task, but behaves the same on the task level under ansible 2.2. | 14:21 |
evrardjp | My theory still applies though :) | 14:22 |
logan- | yea I guess if this gets us out of the loop its a harmless fix, but it seems really fragile... hopefully ansible will fix the bug itself :/ | 14:30 |
*** gouthamr has joined #openstack-ansible | 14:32 | |
evrardjp | well I am pretty sure it's just a workaround for now, and we'll see that issue appear again | 14:33 |
evrardjp | but for me that was just enough for ocata ironic | 14:33 |
evrardjp | the best way is to remove the block level | 14:33 |
evrardjp | wow that was not clear. | 14:33 |
*** yifei has quit IRC | 14:33 | |
evrardjp | I meant: The best way is not get a higher block level inside the ansible strategy, which could be obtained by refactoring this to not have an include_role. | 14:34 |
evrardjp | Or alternatively, putting the recursion level to a higher value | 14:34 |
evrardjp | but that's core ansible and I'd prefer refactoring to avoid the include_role instead of messing up things. | 14:35 |
*** jamesdenton has joined #openstack-ansible | 14:39 | |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-ceph_client master: Remove one layer of includes https://review.openstack.org/521532 | 14:42 |
*** sxc731 has quit IRC | 14:42 | |
*** hachi_ has joined #openstack-ansible | 14:45 | |
*** taseer2 is now known as Taseer | 14:46 | |
*** woodard has quit IRC | 14:50 | |
*** jwitko has joined #openstack-ansible | 14:50 | |
*** hw_wutianwei has joined #openstack-ansible | 14:56 | |
*** udesale has joined #openstack-ansible | 15:00 | |
*** sxc731 has joined #openstack-ansible | 15:05 | |
*** dave-mcc_ is now known as dave-mccowan | 15:06 | |
*** Brew has joined #openstack-ansible | 15:15 | |
*** weezS has joined #openstack-ansible | 15:18 | |
*** udesale has quit IRC | 15:18 | |
*** marst has joined #openstack-ansible | 15:20 | |
*** SmearedBeard has quit IRC | 15:24 | |
openstackgerrit | Merged openstack/openstack-ansible-ceph_client master: Remove one layer of includes https://review.openstack.org/521532 | 15:25 |
*** woodard has joined #openstack-ansible | 15:26 | |
*** SmearedBeard has joined #openstack-ansible | 15:27 | |
cloudnull | mornings | 15:29 |
*** Smeared_Beard has joined #openstack-ansible | 15:30 | |
spotz | hey cloudnull | 15:31 |
*** SmearedBeard has quit IRC | 15:31 | |
*** hachi_ has quit IRC | 15:39 | |
cloudnull | o/ spotz | 15:41 |
cloudnull | hwoarang: any word on the suse repo saga - or is this http://logs.openstack.org/27/466827/23/check/openstack-ansible-functional-opensuse-423/c36b929/job-output.txt.gz#_2017-11-20_04_32_04_139840 something we're now not doing correctly ? | 15:42 |
odyssey4me | cloudnull evrardjp hwoarang should we perhaps flip all opensuse jobs to non-voting until the repo issue is resolved? | 15:43 |
hwoarang | cloudnull: sed etc is fine we are just hitting the external mirrors far too often now and we get timeouts from time to time. The main mirror is still down so i have proposed https://review.openstack.org/521508 | 15:43 |
odyssey4me | we can do that with one change to the tests repo | 15:43 |
cloudnull | hwoarang: ok | 15:44 |
*** jvidal has quit IRC | 15:44 | |
spotz | cloudnull: https://review.openstack.org/#/c/498841/2 | 15:44 |
evrardjp | odyssey4me: fine for me | 15:44 |
hwoarang | odyssey4me: you can make them non-voting if that helps. make sure you revert the previous workaround | 15:44 |
hwoarang | (and also the line about the 42.1 special handling :) ) | 15:44 |
cloudnull | ++ | 15:45 |
hwoarang | i will ping infra people when US people start waking up | 15:45 |
hwoarang | but even if that change goes in it will take 6h max to sync the infra mirrors | 15:45 |
hwoarang | i was told the sync 4 times per day | 15:45 |
hwoarang | *they | 15:46 |
odyssey4me | hwoarang so we need to revert https://review.openstack.org/520755 right? | 15:46 |
hwoarang | yep | 15:48 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Revert opensuse workaround and make opensuse non-voting https://review.openstack.org/521565 | 15:51 |
odyssey4me | evrardjp so you gonna backport https://review.openstack.org/521532 ? | 15:52 |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-ceph_client stable/pike: Remove one layer of includes https://review.openstack.org/521566 | 15:54 |
evrardjp | odyssey4me: yes | 15:55 |
*** jvidal has joined #openstack-ansible | 15:57 | |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-ceph_client stable/ocata: Remove one layer of includes https://review.openstack.org/521568 | 15:59 |
*** taseer1 has joined #openstack-ansible | 16:00 | |
*** Taseer has quit IRC | 16:01 | |
*** taseer2 has joined #openstack-ansible | 16:01 | |
*** taseer2 is now known as Taseer | 16:01 | |
evrardjp | I'm done for today. Cya in 2 days! | 16:03 |
*** chyka has joined #openstack-ansible | 16:04 | |
*** taseer1 has quit IRC | 16:05 | |
*** sxc731 has quit IRC | 16:13 | |
*** chyka has quit IRC | 16:14 | |
*** chyka has joined #openstack-ansible | 16:15 | |
*** pcaruana has quit IRC | 16:35 | |
*** sxc731 has joined #openstack-ansible | 16:44 | |
Smeared_Beard | hey all.. I am running the setup-infrastructure.yml part and it's getting to ' Create OpenStack-Ansible requirement wheels' and after 15 is minutes it just quits, and drops me off SSH.. Im going to rerun with more verbosity, but has anyone seen this before? | 16:49 |
*** jvidal has quit IRC | 16:49 | |
odyssey4me | nope | 16:52 |
odyssey4me | Smeared_Beard I haven't, personally. | 16:53 |
Smeared_Beard | it's strange, I've done it twice, same result both times | 16:53 |
odyssey4me | that task execution is very heavy on the bandwidth though | 16:53 |
cloudnull | maybe something to do with the sec role ? | 16:55 |
openstackgerrit | Logan V proposed openstack/openstack-ansible master: [WIP] Implement iptables management https://review.openstack.org/512099 | 16:55 |
cloudnull | I know it has a log out after some amount of inactivity? | 16:55 |
cloudnull | mhayden: ^ | 16:55 |
Smeared_Beard | maybe... Im rerunning with more verbosity now | 16:56 |
Smeared_Beard | aaand it's gone past that stage o.O typical | 17:08 |
bndzor | evrardjp: added and rerunning the playbook now | 17:15 |
*** weezS has quit IRC | 17:15 | |
*** gkadam has quit IRC | 17:17 | |
bndzor | for debugging 2 vars, would it be correct to add 2 'var: whatever' in '- debug:' ? | 17:20 |
bndzor | or should i create two '- debug' | 17:20 |
bndzor | or even var: x y | 17:20 |
*** electrofelix has quit IRC | 17:21 | |
openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts master: Fix ansible-lint errors https://review.openstack.org/521357 | 17:23 |
*** weezS has joined #openstack-ansible | 17:26 | |
errr | cloudnull: can you help do the needful to this? https://review.openstack.org/#/c/521283/ | 17:31 |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-tests master: Remove the serialization in tests https://review.openstack.org/521598 | 17:31 |
errr | thanks | 17:32 |
openstackgerrit | Merged openstack/openstack-ansible-tests master: Revert opensuse workaround and make opensuse non-voting https://review.openstack.org/521565 | 17:32 |
openstackgerrit | michael rice proposed openstack/openstack-ansible-plugins stable/ocata: Add missing domain parameter to _get_group_roles https://review.openstack.org/521600 | 17:34 |
openstackgerrit | michael rice proposed openstack/openstack-ansible-plugins stable/newton: Add missing domain parameter to _get_group_roles https://review.openstack.org/521601 | 17:34 |
*** mbuil has quit IRC | 17:42 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/pike: Temporarily make opensuse jobs non-voting https://review.openstack.org/521606 | 17:44 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/pike: Temporarily make opensuse jobs non-voting https://review.openstack.org/521607 | 17:45 |
odyssey4me | hwoarang cloudnull evrardjp https://review.openstack.org/521607 should unblock the pike merges which are blocked due to opensuse build issues | 17:46 |
bndzor | Ah well, i found the hard way :) | 17:50 |
*** pcaruana has joined #openstack-ansible | 18:00 | |
*** rodolof has joined #openstack-ansible | 18:03 | |
*** m|y|k has quit IRC | 18:10 | |
*** cjloader has joined #openstack-ansible | 18:11 | |
*** cjloader has quit IRC | 18:11 | |
errr | odyssey4me: cloudnull could yall help this into newton for us? https://review.openstack.org/#/c/521601/ | 18:12 |
bndzor | logan-: sent you the new output in a pm, hope maybe this helps ? | 18:16 |
bndzor | evrardjp: sent you also the new output in a pm with the debug | 18:17 |
*** weezS has quit IRC | 18:17 | |
*** hachi_ has joined #openstack-ansible | 18:26 | |
*** hw_wutianwei has quit IRC | 18:27 | |
openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts master: Make the cache prep timeout configurable https://review.openstack.org/521313 | 18:27 |
openstackgerrit | Logan V proposed openstack/openstack-ansible stable/pike: Add ceph_mons to ceph example https://review.openstack.org/521626 | 18:28 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts stable/pike: Make the cache prep timeout configurable https://review.openstack.org/521628 | 18:37 |
bndzor | odyssey4me: by default, the 60 retries and 600 timeout is to low, atleast using mechanical drives | 18:40 |
bndzor | odyssey4me: 900 async and 200 retries does the trick on "new" hardware | 18:41 |
odyssey4me | bndzor so you need 15 mins? | 18:41 |
bndzor | correct | 18:41 |
odyssey4me | that task doesn't do all that much, so that sounds pretty bad | 18:41 |
bndzor | well | 18:42 |
bndzor | it downloads half of internetz | 18:42 |
odyssey4me | well, with that patch at least it's configurable ;) | 18:42 |
bndzor | no idea what but i saw x libs and other crap | 18:42 |
bndzor | also ran into a bug, where if it times out and aborts | 18:42 |
bndzor | rerunning it once or twice, in the end there will be a error saying containers are exceeding quotas | 18:42 |
bndzor | a reboot fixed that | 18:43 |
odyssey4me | ja, I think that's an issue with the btrfs volume filling up or something when you retry | 18:44 |
odyssey4me | cloudnull would have to comment | 18:44 |
*** RandomTech has joined #openstack-ansible | 18:48 | |
RandomTech | hey, quick question, were working on editing the user_config file and were wondering how to exclude a range of ip. The example was also a bit unclear on what all needed excluded | 18:49 |
odyssey4me | RandomTech https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L73-L95 | 18:49 |
RandomTech | oaky i missed that line in the explanation | 18:50 |
*** rodolof has quit IRC | 18:57 | |
*** weezS has joined #openstack-ansible | 19:01 | |
*** stuartgr has quit IRC | 19:01 | |
RandomTech | hey will ansible do tasks out of order? | 19:06 |
*** woodard has quit IRC | 19:22 | |
errr | RandomTech: Plays, like tasks, run in the order specified in the playbook: top to bottom. see -> http://docs.ansible.com/ansible/latest/playbooks_intro.html | 19:29 |
cloudnull | odyssey4me: bndzor: is this an aio? | 19:29 |
cloudnull | if you're using the aio setup w/ containers and machinctl the deployment will be on a loop back file. | 19:30 |
RandomTech | hey cloudnull is a physical load balancer required? we were configuring our file and hit a snag with the load balancers virtual ip | 19:30 |
cloudnull | however if this is real deployment and you still want to use machinctl you can make /var/lib/machines a partition w/ btrfs which greatly improves the general performance. | 19:31 |
cloudnull | RandomTech: no, you can use haproxy | 19:31 |
RandomTech | cloudnull: would the internall virtual adress then be the management addresses of whichever noced we choose to be hproxy | 19:34 |
*** woodard has joined #openstack-ansible | 19:34 | |
cloudnull | yes. | 19:35 |
cloudnull | the internal and external addresses just have to be different | 19:35 |
cloudnull | http://paste.openstack.org/show/626848/ | 19:36 |
cloudnull | ^ from my openstack_user_config.yml | 19:36 |
RandomTech | what if we have multiple hosts. for example makeing all 3 of our controller nodes hproxy as well | 19:37 |
cloudnull | http://paste.openstack.org/show/626850/ | 19:37 |
cloudnull | ^ the interface file on my lb node | 19:38 |
RandomTech | do you have a seperate node for load balancing then? or is it an adittional node? | 19:38 |
cloudnull | if you want to have haproxy on multiple nodes you'd simply have to setup the keepalived bits | 19:38 |
RandomTech | were new to setting this up cloudnull sorry | 19:39 |
cloudnull | https://docs.openstack.org/openstack-ansible-haproxy_server/latest/configure-haproxy.html#making-haproxy-highly-available | 19:39 |
odyssey4me | RandomTech check the appendices for examples - there are a few | 19:39 |
odyssey4me | https://docs.openstack.org/project-deploy-guide/openstack-ansible/pike/app-config-prod.html | 19:39 |
*** sxc731 has quit IRC | 19:39 | |
cloudnull | so w/ multiple nodes you'll need to have the right keepalived config setup for your environment and then have your multiple nodes in the openstack_user_config.yml | 19:40 |
RandomTech | were confused on what we put for the "internal_lb_vip_address:" line if we have multiple nodes. | 19:45 |
odyssey4me | an address in the appropriate CIDR for keepalived to bind | 19:46 |
RandomTech | so one seperate from all of our hosts? | 19:47 |
cloudnull | yes. | 19:52 |
RandomTech | so if for example if our load balancing hosts have the ips 10.0.0.1,10.0.0.2,10.0.0.3 then we would want to use something like 10.0.0.4 | 19:52 |
cloudnull | yes. | 19:53 |
cloudnull | and keepalived would set the 10.0.0.4 address on the various lb hosts as needed | 19:53 |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server master: Implement a proper WSREP check for galera https://review.openstack.org/520665 | 19:56 |
RandomTech | cloudnull: mind if i show you what our networking is looking like? | 19:58 |
*** hybridpollo has joined #openstack-ansible | 20:00 | |
*** hybridpollo has quit IRC | 20:06 | |
*** hybridpollo has joined #openstack-ansible | 20:07 | |
*** sxc731 has joined #openstack-ansible | 20:15 | |
*** hybridpollo has quit IRC | 20:15 | |
*** hybridpollo has joined #openstack-ansible | 20:15 | |
*** hachi_ has quit IRC | 20:29 | |
*** savvas_ has joined #openstack-ansible | 20:34 | |
*** Brew has quit IRC | 20:35 | |
*** pcaruana has quit IRC | 20:39 | |
*** rodolof has joined #openstack-ansible | 20:50 | |
cloudnull | RandomTech: sure | 21:01 |
cloudnull | pastebin anything you want me to look at | 21:01 |
*** RandomTech has quit IRC | 21:03 | |
*** savvas_ has quit IRC | 21:04 | |
openstackgerrit | Merged openstack/openstack-ansible-tests stable/pike: Temporarily make opensuse jobs non-voting https://review.openstack.org/521607 | 21:17 |
*** dave-mccowan has quit IRC | 21:18 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-lxc_container_create stable/pike: Updated from OpenStack Ansible Tests https://review.openstack.org/521660 | 21:21 |
*** eumel8 has quit IRC | 21:25 | |
*** sxc731 has quit IRC | 21:26 | |
*** woodard has quit IRC | 21:32 | |
*** mpranjic has quit IRC | 22:00 | |
*** shardy has quit IRC | 22:06 | |
*** rodolof has quit IRC | 22:13 | |
*** woodard has joined #openstack-ansible | 22:18 | |
*** jwitko__ has joined #openstack-ansible | 22:22 | |
*** weezS has quit IRC | 22:24 | |
*** jwitko has quit IRC | 22:26 | |
*** weezS has joined #openstack-ansible | 22:27 | |
*** galstrom_zzz has quit IRC | 22:31 | |
*** jwitko__ has quit IRC | 22:32 | |
*** galstrom_zzz has joined #openstack-ansible | 22:33 | |
Smeared_Beard | hey, Im hitting an issue with barbican install it says a file does not exist, but it does.. Ive put the snipper here http://paste.openstack.org/show/626875/ | 22:44 |
Smeared_Beard | this is on CentOS as well | 22:45 |
*** lbragstad has quit IRC | 22:51 | |
Smeared_Beard | ah. the interpreter on barbican-manage is looking for /tmp/openstack-venv-builder/venvs/barbican/bin/python | 23:03 |
*** jwitko has joined #openstack-ansible | 23:03 | |
*** lbragstad has joined #openstack-ansible | 23:04 | |
*** pmannidi has joined #openstack-ansible | 23:05 | |
*** marst has quit IRC | 23:21 | |
*** woodard has quit IRC | 23:24 | |
*** savvas_ has joined #openstack-ansible | 23:30 | |
*** savvas_ has quit IRC | 23:34 | |
*** savvas_ has joined #openstack-ansible | 23:47 | |
*** Smeared_Beard has quit IRC | 23:52 | |
*** weezS has quit IRC | 23:52 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!