*** sawblade6 has joined #openstack-ansible | 00:36 | |
prometheanfire | does this look right for cache prep network wise? http://logs.openstack.org/02/608102/7/check/openstack-ansible-functional-gentoo-17-0-systemd/ae79833/logs/host/lxc-cache-prep-commands.log.txt.gz | 00:37 |
---|---|---|
prometheanfire | I'm getting dns lookup errors | 00:37 |
*** sawblade6 has quit IRC | 00:40 | |
mnaser | cloudnull: that looks broken. is unbound running on the machine? | 00:57 |
mnaser | err prometheanfire ^ | 00:57 |
mnaser | cloudnull: i have a failing nspawn deployment, containers cant talk to host and vice versa | 00:58 |
mnaser | noprefixroute seems to be added which means it doesnt create a route.. i wonder if thats a cause? | 00:58 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-nspawn_hosts master: Avoid repeating CIDR twice https://review.openstack.org/608475 | 01:13 |
*** ansmith has quit IRC | 01:14 | |
prometheanfire | mnaser: is there a way I can have the gentoo stuff use nspawn only? or does it have to support lxc? | 01:14 |
mnaser | prometheanfire: nspawn is on its way out. i think i would be okay with gentoo being nspawn only | 01:14 |
mnaser | tho there are some fixes on the way | 01:14 |
mnaser | i just found another bug that would cause things to halt | 01:15 |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 01:15 |
prometheanfire | mnaser: ya, given that I'm am trying to use standard things to make deploying on gentoo work the same as other systemd hosts it'd be nice to use that instead | 01:15 |
prometheanfire | mnaser: but tests still use lxc, so I still have to support it there? | 01:16 |
mnaser | prometheanfire: not when im done with https://review.openstack.org/#/c/608433/ :) | 01:16 |
prometheanfire | watching it | 01:17 |
prometheanfire | and the host should be running unbound, infra tests check for it | 01:19 |
prometheanfire | the log I linked is prepping the container | 01:19 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-nspawn_hosts master: Revert "Modify the network to ensure their more restrictive" https://review.openstack.org/608476 | 01:55 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-nspawn_hosts master: Switch container create jobs to voting https://review.openstack.org/608477 | 01:55 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible master: Enable passing nspawn jobs https://review.openstack.org/608433 | 01:57 |
*** faizy98 has joined #openstack-ansible | 02:25 | |
*** faizy_ has quit IRC | 02:27 | |
*** dave-mccowan has quit IRC | 02:30 | |
*** idlemind has joined #openstack-ansible | 02:49 | |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 03:43 |
prometheanfire | mnaser: file (/sys/class/net/lxcbr0/bridge/bridge_id) is absent, cannot continue | 03:51 |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 03:52 |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 04:32 |
*** shyamb has joined #openstack-ansible | 04:59 | |
*** shyamb has quit IRC | 05:05 | |
openstackgerrit | 98k proposed openstack/ansible-hardening master: fix tox python3 overrides https://review.openstack.org/608571 | 05:17 |
*** shyamb has joined #openstack-ansible | 05:20 | |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 05:30 |
*** shyamb has quit IRC | 06:02 | |
*** shyamb has joined #openstack-ansible | 06:05 | |
prometheanfire | http://logs.openstack.org/02/608102/7/check/openstack-ansible-functional-gentoo-17-0-systemd/7069cd8/logs/ara-report/ | 06:33 |
prometheanfire | can someone confirm that my networkd looks correct? (just search for networkd) | 06:33 |
prometheanfire | if you search for 'debug info' you'll see that it is not bringing the interface up... | 06:34 |
*** fghaas has joined #openstack-ansible | 06:38 | |
*** mma has joined #openstack-ansible | 06:40 | |
*** ioni has joined #openstack-ansible | 07:00 | |
*** rgogunskiy has joined #openstack-ansible | 07:08 | |
*** tosky has joined #openstack-ansible | 07:13 | |
*** yetiszaf has joined #openstack-ansible | 07:19 | |
*** mma has quit IRC | 07:20 | |
*** electrofelix has joined #openstack-ansible | 07:20 | |
*** mma has joined #openstack-ansible | 07:21 | |
*** mma has quit IRC | 07:22 | |
*** mma has joined #openstack-ansible | 07:23 | |
*** chandankumar has joined #openstack-ansible | 07:40 | |
benkohl | I'm getting "The conditional check 'physical_host in hostvars['localhost']['unbound_physical_hosts']' failed. The error was: template error while templating string: unexpected '|'" and "The error appears to have been in '/opt/openstack-ansible/playbooks/common-tasks/unbound-clients.yml': line 18, column 3, [...]" in TASK [Apply resolver sorting]. | 07:40 |
benkohl | Any Idea what that means? | 07:40 |
*** mma has quit IRC | 07:45 | |
*** mma has joined #openstack-ansible | 07:45 | |
*** shyamb has quit IRC | 07:48 | |
*** mma has quit IRC | 07:51 | |
*** olivierb has joined #openstack-ansible | 08:06 | |
*** mma has joined #openstack-ansible | 08:11 | |
*** mma has quit IRC | 08:15 | |
*** mma has joined #openstack-ansible | 08:15 | |
*** mma has quit IRC | 08:17 | |
*** mma has joined #openstack-ansible | 08:17 | |
*** mma has quit IRC | 08:20 | |
*** mma has joined #openstack-ansible | 08:20 | |
*** mma has quit IRC | 08:23 | |
*** mma has joined #openstack-ansible | 08:24 | |
*** mma has quit IRC | 08:25 | |
*** mma has joined #openstack-ansible | 08:25 | |
*** DanyC has joined #openstack-ansible | 08:28 | |
*** mma has quit IRC | 08:30 | |
evrardjp | benkohl: wrong inventory? | 08:30 |
*** mma has joined #openstack-ansible | 08:30 | |
benkohl | evrardjp: Nothing changed... | 08:35 |
evrardjp | benkohl: wrong user variables? | 08:36 |
benkohl | evrardjp: I only changed one ceph variable (osd_scenario from lvm to collocated and back to lvm). | 08:38 |
*** shardy has joined #openstack-ansible | 08:39 | |
evrardjp | it's obviously in your user space that you should look at the issue, just saying ;) | 08:46 |
evrardjp | logan-: tempest testing was reduced to a very minimum smoke test, so we aren't that impacted anymore. We are still impacted, but with jesse bumping the timeouts, we can avoid nested virt for stability. | 08:49 |
evrardjp | imo | 08:49 |
benkohl | evrardjp: Hm I'm very new to ansible and osa (student) and I'm really confused getting errors without doing a breaking change. | 08:49 |
*** chandankumar has quit IRC | 08:51 | |
*** fghaas has quit IRC | 08:54 | |
*** shyamb has joined #openstack-ansible | 08:55 | |
openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible stable/rocky: Update all SHAs for final RC https://review.openstack.org/608042 | 09:06 |
evrardjp | benkohl: hey, no worries :) | 09:08 |
evrardjp | benkohl: I just don't have any data to help you with -- you don't provide your inventory nor your user variables | 09:08 |
evrardjp | it must be somewhere around there | 09:08 |
evrardjp | jrosser: are you there? | 09:09 |
evrardjp | or hwoarang | 09:09 |
jrosser | evrardjp: o/ hi | 09:09 |
evrardjp | we cannot release rocky unless 1) we release the previous rc (which I will do I think), 2) we merge this: https://review.openstack.org/#/c/603677/2 | 09:09 |
evrardjp | both needs to be done IMO | 09:09 |
evrardjp | so I'd be happy to have a review on the backport there | 09:10 |
jrosser | done | 09:11 |
evrardjp | for releases I discussed with mnaser (he trusts me, what an idea!), odyssey4me (the earliest the better), but I haven't discussed with you | 09:11 |
evrardjp | i'd rather stop waiting too, and then iterate like we usually do | 09:11 |
evrardjp | so we follow the same process as usual | 09:11 |
evrardjp | so I can then focus on simplifying the process further | 09:11 |
jrosser | the backport there looks sensible, i'm OK with that | 09:11 |
evrardjp | yup | 09:12 |
evrardjp | thanks | 09:12 |
evrardjp | good to know we now carry ceilometer everywhere | 09:12 |
benkohl | evrardjp: I'm sorry. I use OSA rocky with ceph. Here is my config: http://paste.openstack.org/show/731670/ | 09:12 |
evrardjp | love the back to the future references benkohl | 09:13 |
*** vnogin has joined #openstack-ansible | 09:14 | |
evrardjp | benkohl: interestingly enough, I don't see a problem in your files, which I might expect a failure from previous run, or a failure in the inventory | 09:16 |
evrardjp | benkohl: could you paste the full error message, and your openstack_inventory.json somewhere? | 09:16 |
*** fghaas has joined #openstack-ansible | 09:18 | |
benkohl | evrardjp: The inventory: http://paste.openstack.org/show/731671/ and the message: https://snag.gy/HuT6aS.jpg | 09:26 |
benkohl | evrardjp: And thank you very very much!!! | 09:27 |
*** hamzaachi has joined #openstack-ansible | 09:29 | |
yetiszaf | On first look: something is wrong with the behaviour of your shell on login | 09:37 |
benkohl | evrardjp: Sorry the inventory wasn't completly pasted: here the better version: https://gist.github.com/benkuly/49be4c3bc28e7178883be0ce0e8643a3 | 09:40 |
*** shyamb has quit IRC | 09:41 | |
benkohl | yetiszaf: yep that was my first thought, but I get the same message without linux-screen | 09:41 |
*** fghaas has quit IRC | 09:41 | |
*** shyamb has joined #openstack-ansible | 09:43 | |
yetiszaf | benkohl: What you're seeing is that the ansible-facts-lookup returns some sort of xterm-init-sequence instead of the hostvars. I have no idea why it would do that to be honest. | 09:43 |
*** faizy_ has joined #openstack-ansible | 09:44 | |
benkohl | yetiszaf: Maybe I could reset the ansible facts? | 09:48 |
*** faizy98 has quit IRC | 09:48 | |
odyssey4me | so, it seems like there are issues with opensuse on the packages front (bad deps, as it stood on friday) and then also nested virt issues | 09:51 |
odyssey4me | https://review.openstack.org/#/c/608427/ was done, but failed an opensuse job - interestingly, the metal job passed | 09:52 |
odyssey4me | the bionic issues appear to be specific to when containers are used | 09:52 |
odyssey4me | and jrosser it looks like it's when apt cacher is used | 09:52 |
odyssey4me | how about we disable the apt cacher in the gate (quite frankly, I'd be ok with removing it entirely) | 09:52 |
yetiszaf | benkohl: I think they're being generated as you go, but you can check if there's a facts-directory that contains the stuff and clean it | 09:53 |
jrosser | odyssey4me: this https://review.openstack.org/#/c/608445/ appears to show that bypassing apt-cacher-ng means apt-get doesnt get stuck | 09:54 |
*** noonedeadpunk has quit IRC | 09:54 | |
jrosser | however, the jobs still do ocassionaly run slowly to the point of timing out, but i think that is for some other reason | 09:54 |
jrosser | odyssey4me: i used a hacked up ansible to strace the apt installs here https://review.openstack.org/#/c/608423/, interested in some more eyes on that but it seems that when things get stuck, apt-get connect() returns EINPROGRESS and then there is a very long delay before the process continues | 09:56 |
odyssey4me | jrosser: ok, honestly I have no idea what that means - is bionic's apt-cacher-ng perhaps broken? | 09:57 |
jrosser | as far as i can see EINPROGRESS should only really be there to allow the call to return asyncronously whilst the TCP handshake takes place, certainly not for multi-minute delays | 09:57 |
benkohl | yetiszaf: This helped: rm /etc/openstack_deploy/ansible_facts/* . Thank you guys! You are great! | 09:57 |
jrosser | odyssey4me: that is quite possible - as mnaser pointed out it is a very very large increase in version string in bionic | 09:57 |
odyssey4me | jrosser: so perhaps for now we just set https://github.com/openstack/openstack-ansible-repo_server/blob/master/defaults/main.yml#L49 to false if it's bionic ? | 09:58 |
odyssey4me | it looks like opensuse was passing in those jobs - so is opensuse working again, or is it just because it's not using the cache? | 10:00 |
evrardjp | odyssey4me: I'd be okay to remove it | 10:01 |
jrosser | odyssey4me: it can't hurt - apt-get on bionic appears happier without it, will help uncovering what else is making bionic slower | 10:01 |
evrardjp | also it's an infrastructure part, we are expecting OSA to use existing infra (networking, mirrors...) not setting them up | 10:02 |
evrardjp | personal opinion ofc :p | 10:03 |
*** fghaas has joined #openstack-ansible | 10:04 | |
evrardjp | jrosser: whereis the log for the EINPROGRESS thing? I mean strace log | 10:05 |
jrosser | evrardjp: it's in here http://logs.openstack.org/23/608423/2/check/openstack-ansible-deploy-aio_lxc-ubuntu-bionic/d7f50f0/logs/ara-report/result/8cb3c352-b053-4d16-be55-8a7ce3f5389f/ | 10:08 |
jrosser | evrardjp: and heres what i did to ansible to generate it https://github.com/bbc/ansible/commit/777eac0bb67be667e26636d2fbc679276a7929c1 | 10:09 |
jrosser | if you can think of a more useful/better strace invocation then we can try it | 10:09 |
*** fghaas has quit IRC | 10:13 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 10:17 |
evrardjp | jrosser: yeah adding read and write on strace | 10:17 |
evrardjp | while it shouldn't be a problem in this case (as this is an ip, I am curious how the /var/run/nscd/socket failure would be impactful) | 10:18 |
jrosser | evrardjp: maybe select/poll too, i've been wondering if there is a bogus timeout value there | 10:18 |
evrardjp | well because it's async I am expecting the read happens, but doesn't return right away -- so yeah timeout value | 10:19 |
evrardjp | as odyssey4me said, if we get rid of the middleman, it would probably make our code easier and remove that kind of issues, at the expense of no cache | 10:19 |
evrardjp | which should be fine, because the cache is quite local :p | 10:20 |
evrardjp | (in infra at least) | 10:20 |
odyssey4me | the cache absolutely does no good for us in the gate at the very least, because all mirrors are local | 10:20 |
evrardjp | exactly odyssey4me | 10:20 |
evrardjp | I like the code reduction too | 10:20 |
evrardjp | this can go in ops should someone want it | 10:21 |
odyssey4me | I should perhaps add a cleanup bit to that patch? | 10:21 |
evrardjp | afk for food | 10:21 |
evrardjp | ttyl | 10:21 |
*** shyamb has quit IRC | 10:29 | |
*** noonedeadpunk has joined #openstack-ansible | 10:34 | |
mnaser | Hi all | 10:35 |
*** vrobert has joined #openstack-ansible | 10:38 | |
*** noonedeadpunk has quit IRC | 10:38 | |
vrobert | Hi, I realized my nova-api-os-compute (17.04.dev1) uwsgi processes are slowing down if I get the console logs from a VM in horizon for example or running tempest test: tempest.api.compute.admin.test_hypervisor.HypervisorAdminTestJSON.test_get_hypervisor_uptime. | 10:39 |
vrobert | ne process thread is blocked on read(): http://paste.openstack.org/show/731677/ I found that nova-compute process on my compute nodes just doing epoll: do_poll (eventlet/hubs/epolls.py:62). Is it a normal behavior? | 10:40 |
vrobert | One process thread is blocked on read(): http://paste.openstack.org/show/731677/ I found that nova-compute process on my compute nodes just doing epoll: do_poll (eventlet/hubs/epolls.py:62). Is it a normal behavior? | 10:40 |
vrobert | Is anybody expecting the same? | 10:40 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 10:42 |
*** hamzaachi has quit IRC | 10:45 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 10:46 |
*** hamzaachi has joined #openstack-ansible | 10:46 | |
*** dave-mccowan has joined #openstack-ansible | 10:47 | |
*** fghaas has joined #openstack-ansible | 10:49 | |
*** noonedeadpunk has joined #openstack-ansible | 10:49 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Remove apt-cacher-ng https://review.openstack.org/608636 | 10:54 |
odyssey4me | evrardjp jrosser mnaser Thoughts on https://review.openstack.org/#/q/topic:apt-cacher-ng-removal+(status:open+OR+status:merged) ? | 10:55 |
mnaser | odyssey4me: I’m making my way to settle down soon (just arrived to Sweden not long ago) | 10:56 |
*** aedc has joined #openstack-ansible | 10:57 | |
odyssey4me | oh wow, jmccrory strikes again: https://review.openstack.org/581166 | 10:59 |
odyssey4me | (see the comment) | 10:59 |
odyssey4me | mnaser: answered your comment in https://review.openstack.org/607522 | 11:01 |
*** fghaas has quit IRC | 11:03 | |
*** shyamb has joined #openstack-ansible | 11:05 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 11:08 |
*** faizy98 has joined #openstack-ansible | 11:22 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement more axes for an increased test matrix https://review.openstack.org/551791 | 11:23 |
*** faizy_ has quit IRC | 11:24 | |
yetiszaf | I don't know if this is the right place to ask this or not ... I've got a setup that was deployed using openstack-ansible and I can't get debugging to work in the neutron-agents-lxc-containers because there's apparently a bunch of libs for ovs missing. Can I just figure out where they're supposed to come from and install them or is that prone to breaking the setup? | 11:24 |
*** jamesdenton has quit IRC | 11:30 | |
odyssey4me | yetiszaf: I'm not at all familiar with using OVS, but it would help to indicate which branch/tag you're using | 11:33 |
odyssey4me | yetiszaf: also, are you running plays with limits or not? | 11:34 |
*** rpittau has joined #openstack-ansible | 11:35 | |
yetiszaf | odyssey4me: I'd have to check which tag exactly, it's some sort of pike. And I don't have the problems with the ansible-part, but I don't know if the lxc-containers are built for openstack-ansible or are generic. | 11:38 |
odyssey4me | yetiszaf: in pike the agents are in lxc containers by default, yes | 11:38 |
yetiszaf | odyssey4me: so I should ask the neutron-people directly? | 11:39 |
odyssey4me | yetiszaf: well, when in containers they inherit the host kernel and kernel modules - so the modules must be loaded on the host | 11:40 |
odyssey4me | I'm trying to find where that's done | 11:40 |
yetiszaf | odyssey4me: Ah, hang on. | 11:40 |
yetiszaf | odyssey4me: The containers are missing libraries. | 11:41 |
yetiszaf | as in "pip install will probably fix it" | 11:41 |
odyssey4me | maybe, but maybe not - it depends what exactly you're doing and how you're doing it | 11:41 |
odyssey4me | mnaser: uses OVS so perhaps he can help | 11:41 |
odyssey4me | he's just in transit right now | 11:42 |
yetiszaf | odyssey4me: Well, the real problem is that one set of qdhcp-netns does not have connectivity so I am trying to get the ovs-tcpdump to work and that complains that its includes are missing. | 11:43 |
yetiszaf | I'll try asking him later or whenever he pops up. | 11:43 |
yetiszaf | thanks. | 11:43 |
openstackgerrit | Nicolas Bock proposed openstack/openstack-ansible master: Pretty-print output during bootstrap process https://review.openstack.org/607526 | 11:47 |
*** strattao has joined #openstack-ansible | 11:47 | |
*** hamzaachi has quit IRC | 11:53 | |
*** hamzaachi_ has joined #openstack-ansible | 11:53 | |
*** vrobert has left #openstack-ansible | 11:53 | |
*** mma has quit IRC | 11:56 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement changes to allow test usage by roles https://review.openstack.org/551791 | 11:56 |
*** mma has joined #openstack-ansible | 11:56 | |
*** mma has quit IRC | 11:58 | |
benkohl | I get this error: https://snag.gy/kpmiMr.jpg I thought it could be this bug https://bugs.launchpad.net/openstack-ansible/+bug/1777254, but the workaround doesn't work. I use osa rocky with this config: http://paste.openstack.org/show/731670/ | 11:58 |
openstack | Launchpad bug 1777254 in openstack-ansible "using ceph backend with lvm strategy leads to error" [Undecided,Confirmed] - Assigned to Mohammed Naser (mnaser) | 11:58 |
*** mma has joined #openstack-ansible | 11:58 | |
*** hamzaachi_ has quit IRC | 11:58 | |
*** shyamb has quit IRC | 12:01 | |
*** shyamb has joined #openstack-ansible | 12:01 | |
*** vnogin has quit IRC | 12:09 | |
*** mma has quit IRC | 12:12 | |
*** mma has joined #openstack-ansible | 12:12 | |
yetiszaf | benkohl: Best guess: the ceph-volume command you're using is too old and does not support "batch" yet. | 12:15 |
*** mma has quit IRC | 12:17 | |
benkohl | yetiszaf: Okay, but actually ceph-ansible from osa uses this. Now I set the user_variable ceph_stable_release to mimic. Maybe that works. | 12:18 |
*** jamesdenton has joined #openstack-ansible | 12:20 | |
yetiszaf | benkohl: you should be able to check that by running it manually or looking into the man-page if that's installed. | 12:21 |
benkohl | yetiszaf: I found a bug report (https://github.com/ceph/ceph-ansible/issues/3146) and the problem seems to be that the version of ceph-ansible that osa rocky points to doesn't work with osd_scenario: lvm. | 12:26 |
yetiszaf | Oh. Yeah, fair enough, that sounds plausible. | 12:27 |
yetiszaf | looking at the report: Yes. | 12:27 |
*** mma has joined #openstack-ansible | 12:28 | |
*** mma has quit IRC | 12:34 | |
*** mma has joined #openstack-ansible | 12:37 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 12:40 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 12:42 |
jamesdenton | odyssey4me Did this opensuse error ever get resolved for you? http://logs.openstack.org/69/584069/11/check/openstack-ansible-deploy-aio_distro_lxc-opensuse-423/01d6240/logs/ara-report/result/a77bbf6a-8035-4ede-baa0-f8b22517652a/ | 12:44 |
odyssey4me | jamesdenton: I've seen some suse jobs passed today or over the weekend, but haven't determined anything concrete just yet | 12:49 |
jamesdenton | cool. I'll try again. Couldn't find a matching version in the repository > than 1.18.0 | 12:50 |
*** shyam89 has joined #openstack-ansible | 12:50 | |
*** shyamb has quit IRC | 12:51 | |
*** ansmith has joined #openstack-ansible | 12:51 | |
*** lbragstad has joined #openstack-ansible | 12:54 | |
*** mma has quit IRC | 12:57 | |
*** mma has joined #openstack-ansible | 12:58 | |
*** mma has quit IRC | 13:02 | |
*** shyam89 has quit IRC | 13:09 | |
*** shyam89 has joined #openstack-ansible | 13:11 | |
*** spatel has joined #openstack-ansible | 13:13 | |
*** mma has joined #openstack-ansible | 13:14 | |
*** ArchiFleKs has joined #openstack-ansible | 13:17 | |
*** lbragstad has quit IRC | 13:24 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement changes to allow test usage by roles https://review.openstack.org/551791 | 13:26 |
*** weezS has joined #openstack-ansible | 13:27 | |
*** shyam89 has quit IRC | 13:30 | |
*** fghaas has joined #openstack-ansible | 13:32 | |
*** lykki has joined #openstack-ansible | 13:32 | |
*** lbragstad has joined #openstack-ansible | 13:32 | |
*** lykki has left #openstack-ansible | 13:33 | |
*** vnogin has joined #openstack-ansible | 13:39 | |
*** spatel has quit IRC | 13:41 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-haproxy_server master: Add 'absent' service state https://review.openstack.org/603158 | 13:43 |
odyssey4me | logan-: ^ That seems to work for me. The idempotence test threw a bit of a spanner in the works. | 13:44 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-haproxy_server master: Add 'absent' service state https://review.openstack.org/603158 | 13:45 |
*** fghaas has quit IRC | 13:47 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 13:48 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Remove apt-cacher-ng https://review.openstack.org/608636 | 13:53 |
*** kukacz has quit IRC | 13:54 | |
*** kukacz has joined #openstack-ansible | 13:54 | |
*** defionscode has joined #openstack-ansible | 13:58 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement changes to allow test usage by roles https://review.openstack.org/551791 | 14:07 |
*** munimeha1 has joined #openstack-ansible | 14:07 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-haproxy_server master: Add 'absent' service state https://review.openstack.org/603158 | 14:11 |
arxcruz | mnaser: i think this requires a re-workflow https://review.openstack.org/#/c/606084/ | 14:12 |
odyssey4me | arxcruz: it would, except it's rebased on other patches which haven't merged yet, are failing, and likely to change again | 14:13 |
arxcruz | (╯°□°)╯︵ ┻━┻ | 14:13 |
odyssey4me | mnaser: on https://review.openstack.org/608427 - odd that the metal opensuse is passing, but the container builds are failing... perhaps the apt-cacher-ng issue is present for more than just bionic | 14:14 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Disable nested virt and set SUSE+Bionic jobs to non-voting https://review.openstack.org/608427 | 14:16 |
odyssey4me | mnaser: updated ^ to also set the aio_basekit job for opensuse to non-voting | 14:16 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Restore Bionic jobs to voting https://review.openstack.org/608428 | 14:17 |
jamesdenton | odyssey4me FWIW - it appears to be failing due to the python-neutron-lib dependency via nova-compute install, which is not being installed on metal | 14:20 |
*** olivierb has quit IRC | 14:21 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Disable nested virt and set SUSE+Bionic jobs to non-voting https://review.openstack.org/608427 | 14:21 |
*** noonedeadpunk has quit IRC | 14:22 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Restore OpenSUSE jobs https://review.openstack.org/608429 | 14:24 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Restore Bionic jobs to voting https://review.openstack.org/608428 | 14:24 |
*** skiedude has joined #openstack-ansible | 14:24 | |
*** weezS has quit IRC | 14:25 | |
*** noonedeadpunk has joined #openstack-ansible | 14:27 | |
*** weezS has joined #openstack-ansible | 14:28 | |
*** weezS has quit IRC | 14:28 | |
*** weezS has joined #openstack-ansible | 14:29 | |
*** faizy_ has joined #openstack-ansible | 14:33 | |
*** faizy98 has quit IRC | 14:36 | |
cloudnull | mornings | 14:43 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Remove apt-cacher-ng https://review.openstack.org/608636 | 14:44 |
*** Emine has joined #openstack-ansible | 14:49 | |
nicolasbock | o/ cloudnull | 14:50 |
logan- | nice work on https://review.openstack.org/#/c/603158 odyssey4me | 14:50 |
odyssey4me | logan-: it'd e nice if we could make the add/remove tasks cleaner by pre-preparing the lists so that there aren't skipped ones for those not meeting conditionals | 14:52 |
odyssey4me | you're better at map/extract than me though - that makes my head hurt | 14:52 |
logan- | i was just reading thru the conditional you changed on the state absent removal task though | 14:52 |
logan- | and with your changes to the test playbook, I don't know why the conditional would need to change from what it was before | 14:52 |
*** vnogin has quit IRC | 14:53 | |
odyssey4me | logan-: well, previously you had item.state instead of item.service.state | 14:53 |
logan- | oh gotcha | 14:53 |
odyssey4me | but also, it now will also remove the files for anything with no backends or if it's disabled | 14:53 |
logan- | ah nice | 14:53 |
nicolasbock | I have (potential) fix for the openSUSE gates in https://review.openstack.org/#/c/608370/ | 14:54 |
logan- | i figured there was something like that you were accommodating just wasn't sure :) | 14:54 |
odyssey4me | logan-: I wish it could be done more cleanly - that was just a dirty hack to get it working... I'm pretty sure it could be made cleaner. | 14:55 |
logan- | yeah | 14:55 |
odyssey4me | logan-: please think about it and modify - my main aim was to get the idempotence working | 14:55 |
logan- | cool will do | 14:55 |
*** faizy98 has joined #openstack-ansible | 14:56 | |
odyssey4me | nicolasbock: oh awesome, I wasn't aware suse was using the rpm-packaging project | 14:56 |
*** mma has quit IRC | 14:56 | |
*** fghaas has joined #openstack-ansible | 14:56 | |
nicolasbock | Yes, we are, sometimes ;) | 14:56 |
*** mma has joined #openstack-ansible | 14:56 | |
odyssey4me | haha | 14:57 |
odyssey4me | nicolasbock: does that mean that we could work around the issue for now by just adding another package? | 14:57 |
*** mma has quit IRC | 14:57 | |
skiedude | Question about galera. We're running a 2 controller (infra) setup. When the main galera node goes down, it takes down the backup with it. Poking around http://galeracluster.com/documentation-webpages/twonode.html shows some config options you can pass so that this does not happen (with some risk). Does anyone run a 2 node cluster and have suggestions on how to handle this? | 14:58 |
*** faizy_ has quit IRC | 14:58 | |
*** mma has joined #openstack-ansible | 14:58 | |
nicolasbock | We would need to build `python-neutron-lib-1.19.0` since that's the version that `python-neutron` expects. The current version in OBS is `1.18.0`. | 14:59 |
nicolasbock | But yes, if we could build that package and install it it would unblock `python-neutron` | 15:00 |
odyssey4me | skiedude: if that can be added to a .conf file, then config overrides can be used: https://docs.openstack.org/openstack-ansible/latest/reference/configuration/using-overrides.html#overriding-openstack-configuration-defaults | 15:00 |
odyssey4me | nicolasbock: ah ok, so regardless we need to wait for a rebuild | 15:00 |
skiedude | is there a suggested method when running 2 nodes instead of 3? In a 3 node, you could lose a controller and be fine, but in 2, its a bit different odyssey4me | 15:01 |
nicolasbock | I think so odyssey4me | 15:01 |
*** mma has quit IRC | 15:02 | |
odyssey4me | skiedude: we run 3 infra nodes, as do most reasonably sized builds... I've never seen a question from anyone running 2 - usually it's one or three | 15:04 |
skiedude | yeah I'm finding out that running a 2 node cluster, really negates most of the HA for some of this stuff like rmq/galera | 15:06 |
odyssey4me | yep - one option is to pick a single compute node to run containers for galera/rabbitmq only so that you retain ha | 15:06 |
skiedude | yeah we've considered that, but I feel that takes away the planned compute resources for VMs | 15:07 |
skiedude | something to ponder on | 15:07 |
skiedude | thanks | 15:07 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement changes to allow test usage by roles https://review.openstack.org/551791 | 15:08 |
*** fghaas has quit IRC | 15:08 | |
*** vnogin has joined #openstack-ansible | 15:12 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat master: Move heat domain setup into service setup tasks https://review.openstack.org/604362 | 15:13 |
*** vnogin has quit IRC | 15:15 | |
*** rgogunskiy has quit IRC | 15:20 | |
*** Emine has quit IRC | 15:21 | |
openstackgerrit | Nicolas Bock proposed openstack/openstack-ansible master: Pretty-print output during bootstrap process https://review.openstack.org/607526 | 15:26 |
*** aedc has quit IRC | 15:34 | |
noonedeadpunk | hi cloudnull. Can you please check this patch? https://review.openstack.org/#/c/605769/ You've already reviewed previous version of it, and I'd implemented requested changes. | 15:35 |
noonedeadpunk | thanks) | 15:38 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-ceph_client stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608708 | 15:41 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-rsyslog_client stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608710 | 15:42 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-pip_install stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608711 | 15:42 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-os_ceilometer stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608712 | 15:42 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-lxc_container_create stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608713 | 15:42 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-repo_build stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608714 | 15:42 |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-os_aodh stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608715 | 15:42 |
*** ohsnap has joined #openstack-ansible | 15:42 | |
openstackgerrit | jacky06 proposed openstack/openstack-ansible-haproxy_server stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/608716 | 15:43 |
*** vnogin has joined #openstack-ansible | 15:44 | |
*** vnogin has quit IRC | 15:48 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [WIP] Implement changes to allow test usage by roles https://review.openstack.org/551791 | 15:48 |
*** faizy_ has joined #openstack-ansible | 15:48 | |
*** faizy98 has quit IRC | 15:52 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Move the embedded ansible tools to its own subdir https://review.openstack.org/608718 | 15:52 |
skiedude | One of my final hurdles for the setup is the repo_container, specifically the apt-cacher-ng service. This service when installed creates a folder in /run/ named apt-cacher-ng where it puts it socket file. When the repo_container is rebooted this folder goes away because its a tmpfs. So when the container restarts, the service cannot start because that folder does not exist. This seems like an issue with the service, and not OSA di | 15:53 |
skiedude | rectly, but has anyone run into this? | 15:53 |
cloudnull | skiedude - ive not seen this recently . | 15:54 |
cloudnull | what release are you running ? | 15:54 |
skiedude | queens, 17.1.0. it happens every time I restart the repo_container | 15:54 |
* cloudnull goes to restart a repo server container . | 15:54 | |
odyssey4me | skiedude: I have not, but you can disable the package cache with 'repo_pkg_cache_enabled: no' in your user_variables.yml file | 15:54 |
skiedude | I assume its not a mission critical thing if its easily disabled | 15:54 |
skiedude | presumably just speeds up yum with the cache enabled? | 15:55 |
odyssey4me | yep | 15:55 |
skiedude | not a big deal | 15:55 |
odyssey4me | ah, so this is centos then? | 15:55 |
skiedude | yeah I am running centos which may be partly the issue, since I believe this is normally a ubuntu friendly service | 15:55 |
cloudnull | skiedude I'm not seeing the issue, however, my environment is ubuntu | 15:56 |
cloudnull | patel you around? | 15:56 |
odyssey4me | cloudnull: on that topic, what're your thoughts for https://review.openstack.org/#/q/topic:apt-cacher-ng-removal+(status:open+OR+status:merged) ? bionic is absolutely failing with the cache, and it's not essential and can rather be implemented in other ways by deployers | 15:56 |
cloudnull | skiedude mind checking the service unit . on ubuntu it looks like so | 15:57 |
cloudnull | http://paste.openstack.org/show/731701/ | 15:57 |
cloudnull | maybe there's something different there? | 15:57 |
skiedude | yeah, sec | 15:57 |
skiedude | http://paste.openstack.org/show/731702/ | 15:58 |
skiedude | pretty similar | 15:58 |
*** mma has joined #openstack-ansible | 16:00 | |
jrosser | odyssey4me: can env vars really be used to point to a mirror (just reading the reno for acng removal) | 16:00 |
odyssey4me | jrosser: to a mirror, no - to a proxy, yes | 16:00 |
cloudnull | odyssey4me I have apt-cacher-ng running in a bionic cloud right now . | 16:00 |
odyssey4me | afaik | 16:00 |
cloudnull | seems not broken here? | 16:00 |
odyssey4me | cloudnull: it's causing massive slowdowns: https://etherpad.openstack.org/p/bionic-timeouts | 16:01 |
cloudnull | itd really be a shame for that to go . on the clouds we run having a single point of ingress for packages is a huge timesaver | 16:01 |
odyssey4me | jrosser has been digging into it and confirmed that without it things work just fine | 16:01 |
odyssey4me | sure, nothing prevents a cache being used - but the cache can be outside of the OSA scope | 16:02 |
skiedude | its also easily fixed with playbook to mkdir/chown/chmod the dir it expects to be there, just not sure how much effort thats worth everytime the container is restarted | 16:02 |
cloudnull | skiedude do you have the file /usr/lib/tmpfiles.d/apt-cacher-ng.conf ? | 16:04 |
cloudnull | http://paste.openstack.org/show/731704/ | 16:04 |
odyssey4me | cloudnull: of course, another option could be to make it a service in its own container/host group - we could add something in the ops repo for it | 16:04 |
jrosser | cloudnull: i have a whole bunch of jobs testing the acng failures - please do take a look https://review.openstack.org/#/c/608423/ in particular finds the point that it gets stuck | 16:04 |
*** mma has quit IRC | 16:05 | |
skiedude | I do not | 16:05 |
skiedude | cloudnull: | 16:05 |
cloudnull | probably something we should add to that release to ensure centos is well covered. | 16:05 |
cloudnull | i suspect that temp file was part of the apt package and is not part of the cent one . | 16:06 |
cloudnull | **rpm one | 16:06 |
skiedude | I assume I could manually touch that file to test it out | 16:06 |
cloudnull | odyssey4me ++ I wont block it especially if its a point of failure. | 16:06 |
cloudnull | downstream I can figure something out if need be | 16:06 |
skiedude | something to note, is the folder 0700 on centos | 16:07 |
skiedude | I see in your file its 0755 | 16:07 |
odyssey4me | cloudnull: yeah, I'm thinking perhaps a user story for implementing caching would be good - with a playbook and config files in the ops repo | 16:07 |
cloudnull | skiedude ++ if you dont mind adding that file and restarting the container it'd be a great data-point. | 16:07 |
skiedude | k testing | 16:07 |
odyssey4me | let me work something up, and perhaps jrosser can add pointers for using an existing cache | 16:08 |
jrosser | i have a role for a mirror | 16:09 |
jrosser | we can bung that in the ops repo no problem, but it doesnt use acng | 16:09 |
odyssey4me | for a mirror, I wouldn't expect it to | 16:09 |
jrosser | and tbh - we could/should probably borrow from the infra mirror work if possible, becasue it's non trivial | 16:09 |
*** faizy_ has quit IRC | 16:10 | |
*** faizy_ has joined #openstack-ansible | 16:10 | |
odyssey4me | I was thinking of describing using a cache (squid/acng), with all the right env vars and such. | 16:10 |
odyssey4me | a mirror would be an alternative to that in the general 'limited connectivity' story | 16:11 |
skiedude | cloudnull: worked like a charm, this was the contents of the file, only change is the 0700 since thats what it was defaulting to on its own on first creation | 16:11 |
skiedude | http://paste.openstack.org/show/731706/ | 16:11 |
*** faizy_ has quit IRC | 16:15 | |
*** faizy98 has joined #openstack-ansible | 16:16 | |
*** fghaas has joined #openstack-ansible | 16:17 | |
*** faizy98 has quit IRC | 16:20 | |
*** fghaas has quit IRC | 16:21 | |
*** faizy98 has joined #openstack-ansible | 16:23 | |
*** rgogunskiy has joined #openstack-ansible | 16:24 | |
openstackgerrit | Merged openstack/openstack-ansible-os_cinder master: Added volume usage collection, when cinder_ceilometer_enabled https://review.openstack.org/605769 | 16:26 |
openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_cinder stable/rocky: Added volume usage collection, when cinder_ceilometer_enabled https://review.openstack.org/608726 | 16:28 |
*** rgogunskiy has quit IRC | 16:29 | |
*** zenirc369 has joined #openstack-ansible | 16:41 | |
*** kaiokmo has quit IRC | 16:43 | |
*** hamzaachi has joined #openstack-ansible | 16:44 | |
*** vnogin has joined #openstack-ansible | 16:46 | |
*** gyee has joined #openstack-ansible | 16:48 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Move the embedded ansible tools to its own subdir https://review.openstack.org/608718 | 16:51 |
*** kaiokmo has joined #openstack-ansible | 16:53 | |
*** weezS has quit IRC | 16:58 | |
*** weezS has joined #openstack-ansible | 16:59 | |
*** aedc has joined #openstack-ansible | 16:59 | |
odyssey4me | jrosser mnaser https://review.openstack.org/608636 works, except for suse due to the known package issue | 17:08 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat master: Move heat domain setup into service setup tasks https://review.openstack.org/604362 | 17:10 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Remove apt-cacher-ng https://review.openstack.org/608636 | 17:14 |
prometheanfire | cloudnull: you may know this, but have you been able to create a bridge in networkd with no devices attached? I can't create lxcbr0 | 17:14 |
prometheanfire | netdev is created but not brough up or address configured | 17:14 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Restore Bionic jobs to voting https://review.openstack.org/608428 | 17:14 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Build translated docs in tox https://review.openstack.org/604567 | 17:19 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Fix order for restarts https://review.openstack.org/593032 | 17:19 |
noonedeadpunk | Folks, can you please vote for this cherry-pick to rocky? https://review.openstack.org/#/c/608209/ | 17:19 |
*** faizy_ has joined #openstack-ansible | 17:20 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Update variable migration script for Stein https://review.openstack.org/606932 | 17:22 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Add stackviz directory to the list of find_log_files https://review.openstack.org/606084 | 17:22 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Fix for proper package name depending on base OS distribution https://review.openstack.org/606902 | 17:22 |
*** faizy98 has quit IRC | 17:23 | |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 17:25 |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-tests master: add Gentoo jobs as non-voting https://review.openstack.org/608102 | 17:26 |
*** shardy has quit IRC | 17:40 | |
*** vnogin has quit IRC | 17:42 | |
*** aedc has quit IRC | 18:00 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-nspawn_hosts master: Correct host only routes https://review.openstack.org/608735 | 18:05 |
*** electrofelix has quit IRC | 18:05 | |
cloudnull | prometheanfire yea the bridge has to have a device within it for networkd to bring it up. I've done some of that with a oneshot service unit. | 18:05 |
prometheanfire | cloudnull: actually, that wasn't it, I needed ConfigureWithoutCarrier=yes | 18:08 |
prometheanfire | I did test adding a tap | 18:08 |
prometheanfire | cloudnull: so, gentoo needs some bind mount type of things | 18:09 |
cloudnull | ++ cool ! | 18:09 |
prometheanfire | https://wiki.gentoo.org/wiki/Chroot#Init_scripts | 18:09 |
prometheanfire | I could add those as seperate tasks, but it'd be a ton and polute things | 18:09 |
prometheanfire | also templates/prep-scripts/_container_sys_setup.sh.j2 would need to be modified, maybe | 18:10 |
prometheanfire | easiest would be a list of chroot prep commands that are executed in series (and unprep commands) | 18:11 |
openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: add gentoo support https://review.openstack.org/608393 | 18:12 |
*** weezS has quit IRC | 18:17 | |
*** weezS has joined #openstack-ansible | 18:19 | |
openstackgerrit | Merged openstack/openstack-ansible-repo_server master: Remove apt-cacher-ng https://review.openstack.org/608631 | 18:22 |
*** skiedude has quit IRC | 18:23 | |
*** jonher has quit IRC | 18:23 | |
*** jonher has joined #openstack-ansible | 18:24 | |
*** aedc has joined #openstack-ansible | 18:25 | |
jrosser | hmm so we need to unstick opensuse | 18:34 |
openstackgerrit | Merged openstack/openstack-ansible-os_ceilometer stable/rocky: Fix of oslomsg undefined variables for services notifications https://review.openstack.org/608209 | 18:37 |
cloudnull | jrosser what seems to be stuck? | 18:37 |
cloudnull | something i can help with ? | 18:37 |
cloudnull | prometheanfire seems sensible to add it to the chroot prep tooling ? | 18:38 |
jrosser | cloudnull: looks repo related - this needs to go in next https://review.openstack.org/#/c/608636/ | 18:38 |
noonedeadpunk | Sorry for disturbing you, but 1 more cherry-pick is here..... https://review.openstack.org/#/c/608726/ | 18:38 |
jrosser | cloudnull: stupid deadlock time where this fails on bionic timout https://review.openstack.org/#/c/608427/ and fixing bionic fails on opensuse repos | 18:39 |
cloudnull | jrosser looks like the role change to pull the cacher has merged. | 18:40 |
jrosser | oh no not quite - centos threw in brokennes there just for good measure by the looks of it | 18:41 |
prometheanfire | cloudnull: ya, I have a workaround, /dev needs a symlink, simple | 18:41 |
prometheanfire | maybe more mknod not sure | 18:41 |
jrosser | cloudnull: yeah - i think that means the main repo part of acng removal *must* be the next merge | 18:43 |
cloudnull | ++ already voted we just need it to gate | 18:44 |
*** hamzaachi has quit IRC | 18:47 | |
spotz | jrosser: we're trying to set opensuse as non-voting for the time being to fix the gates | 18:49 |
*** faizy_ has quit IRC | 18:55 | |
*** cshen has joined #openstack-ansible | 18:56 | |
*** mmercer has joined #openstack-ansible | 18:57 | |
openstackgerrit | Merged openstack/openstack-ansible-ops master: Move the embedded ansible tools to its own subdir https://review.openstack.org/608718 | 18:59 |
*** cshen has quit IRC | 19:05 | |
*** skiedude has joined #openstack-ansible | 19:10 | |
*** cshen has joined #openstack-ansible | 19:10 | |
openstackgerrit | Gaudenz Steinlin proposed openstack/openstack-ansible-os_ceilometer master: Make polling namespaces configurable https://review.openstack.org/607202 | 19:13 |
* mnaser starts catching up on reviews and buffer | 19:13 | |
*** cshen has quit IRC | 19:15 | |
*** cshen has joined #openstack-ansible | 19:21 | |
mnaser | odyssey4me: so about removing the apt-cacher-ng stuff. are we going to backport it to stable/rocky to fix bionic? | 19:21 |
*** strattao has quit IRC | 19:22 | |
*** sawblade6 has joined #openstack-ansible | 19:29 | |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-nspawn_hosts master: Switch container create jobs to voting https://review.openstack.org/608477 | 19:30 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-nspawn_hosts master: Correct host only routes https://review.openstack.org/608735 | 19:30 |
*** sawblade6 has quit IRC | 19:30 | |
*** sawblade6 has joined #openstack-ansible | 19:30 | |
mnaser | cloudnull: odyssey4me https://review.openstack.org/#/c/608477 and https://review.openstack.org/#/c/608735/2 to get nspawn working again (and make sure we dont break it again!) | 19:30 |
*** sawblade6 has quit IRC | 19:33 | |
*** sawblade6 has joined #openstack-ansible | 19:33 | |
openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible master: Integrate Blazar with OpenStack Ansible https://review.openstack.org/549956 | 19:33 |
*** sawblade6 has quit IRC | 19:33 | |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible master: Enable passing nspawn jobs https://review.openstack.org/608433 | 19:34 |
*** sawblade6 has joined #openstack-ansible | 19:34 | |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_octavia stable/rocky: Only run OpenStack tasks once https://review.openstack.org/608757 | 19:35 |
*** Taseer has joined #openstack-ansible | 19:35 | |
*** sawblade6 has quit IRC | 19:35 | |
*** sawblade6 has joined #openstack-ansible | 19:37 | |
*** sawblade6 has quit IRC | 19:39 | |
*** hamzaachi has joined #openstack-ansible | 19:40 | |
Taseer | mnaser: here is the error => http://logs.openstack.org/56/549956/20/check/openstack-ansible-deploy-aio_lxc-ubuntu-bionic/4b4f0e8/job-output.txt.gz#_2018-10-06_07_41_03_596482 | 19:44 |
mnaser | click | 19:45 |
mnaser | hmm | 19:45 |
mnaser | let me see if i can help | 19:45 |
mnaser | though i'd usually defer repo related stuff to our resident expert odyssey4me for that stuff | 19:45 |
Taseer | mnaser: any kind of comments would be appreciated. need to get this merged before OSS | 19:46 |
Taseer | mnaser: no worries if you cant figure out the problem | 19:47 |
mnaser | Taseer: im checking | 19:47 |
Taseer | thanks! | 19:47 |
*** aedc has quit IRC | 19:48 | |
mnaser | Taseer: i put up review of a few initial things | 19:52 |
Taseer | mnaser: Thanks! will address the comments. Much appreciated! | 19:53 |
mnaser | Taseer: also not sure why it's pulling in blazar | 19:54 |
mnaser | it shouldn't | 19:54 |
mnaser | Taseer: i suggest maybe pinging odyssey4me tomororw about this, he might be much better | 19:54 |
Taseer | mnaser: okay sure. thanks again! | 19:55 |
mnaser | sorry i couldnt be much help | 19:56 |
mnaser | but we'll absolutely make sure it lands before then | 19:56 |
Taseer | mnaser: no problem! | 19:59 |
*** sawblade6 has joined #openstack-ansible | 19:59 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-nspawn_hosts master: Further simplify the networking stack by eliminating veths https://review.openstack.org/608759 | 20:00 |
*** sawblade6 has quit IRC | 20:01 | |
*** sawblade6 has joined #openstack-ansible | 20:02 | |
cjloader | anyone having bionic lxc-create failures? | 20:03 |
*** sawblade6 has quit IRC | 20:03 | |
*** sawblade6 has joined #openstack-ansible | 20:04 | |
cjloader | on rocky | 20:04 |
*** sawblade6 has quit IRC | 20:04 | |
*** sawblade6 has joined #openstack-ansible | 20:05 | |
*** skiedude has quit IRC | 20:05 | |
*** sawblade6 has quit IRC | 20:07 | |
*** sawblade6 has joined #openstack-ansible | 20:07 | |
*** Taseer has quit IRC | 20:08 | |
*** sawblade6 has quit IRC | 20:08 | |
openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible-os_horizon master: Turning off SSL is currently offline https://review.openstack.org/607969 | 20:18 |
*** DanyC_ has joined #openstack-ansible | 20:24 | |
*** DanyC has quit IRC | 20:27 | |
openstackgerrit | Merged openstack/openstack-ansible-os_cinder stable/rocky: Added volume usage collection, when cinder_ceilometer_enabled https://review.openstack.org/608726 | 20:27 |
cloudnull | cjloader what failures are you seeing? | 20:29 |
*** sawblade6 has joined #openstack-ansible | 20:30 | |
*** sawblade6 has quit IRC | 20:33 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-nspawn_container_create master: Add test for host_only networks and ensure network ordering https://review.openstack.org/608765 | 20:35 |
*** lbragstad has quit IRC | 20:37 | |
*** lbragstad has joined #openstack-ansible | 20:37 | |
*** ansmith has quit IRC | 20:47 | |
*** pcaruana has quit IRC | 20:49 | |
*** skiedude has joined #openstack-ansible | 20:53 | |
*** ohsnap has quit IRC | 20:55 | |
*** spatel has joined #openstack-ansible | 21:04 | |
*** weezS_ has joined #openstack-ansible | 21:16 | |
*** weezS has quit IRC | 21:16 | |
*** weezS_ is now known as weezS | 21:16 | |
*** hamzaachi has quit IRC | 21:24 | |
*** weezS has quit IRC | 21:33 | |
*** cshen has quit IRC | 21:34 | |
*** aedc has joined #openstack-ansible | 21:56 | |
*** skiedude has quit IRC | 22:05 | |
prometheanfire | so... we don't support python2 for the python lxc module | 22:16 |
prometheanfire | {PYTHON_TARGETS="python3_5 python3_6 python3_7"} | 22:16 |
prometheanfire | fatal: [infra1 -> localhost]: FAILED! => {"changed": false, "msg": "The `lxc` module is not importable. Check the requirements."} | 22:18 |
openstackgerrit | Andy Smith proposed openstack/openstack-ansible master: Add qdrouterd role for oslo.messaging backend deployment https://review.openstack.org/570261 | 22:21 |
prometheanfire | ah https://github.com/lxc/python2-lxc | 22:22 |
prometheanfire | cloudnull: is there a timeline for osa to switch to ansible py3? | 22:23 |
*** DanyC_ has quit IRC | 22:27 | |
spotz | prometheanfire cloudnull - we also may not support non-SSL testing in tempest:( Cause I know my latest patch worked when manually applied and just totally borked in the gates:( | 22:28 |
*** ansmith has joined #openstack-ansible | 22:35 | |
prometheanfire | ok, packaged lxc-python2 | 22:37 |
*** vnogin has joined #openstack-ansible | 22:43 | |
*** vnogin has quit IRC | 22:47 | |
*** dave-mccowan has quit IRC | 22:49 | |
*** munimeha1 has quit IRC | 22:50 | |
cloudnull | prometheanfire stien | 22:50 |
*** dave-mccowan has joined #openstack-ansible | 22:51 | |
cloudnull | spotz that might be, and would explain a but | 22:51 |
cloudnull | **bit | 22:51 |
*** tosky has quit IRC | 23:03 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-nspawn_container_create master: Add test for host_only networks and ensure network ordering https://review.openstack.org/608765 | 23:08 |
*** lbragstad has quit IRC | 23:16 | |
*** zenirc369 has quit IRC | 23:18 | |
prometheanfire | well, up to lxc_container_create | 23:26 |
*** cshen has joined #openstack-ansible | 23:29 | |
*** bradm has joined #openstack-ansible | 23:31 | |
*** cshen has quit IRC | 23:34 | |
*** gyee has quit IRC | 23:57 | |
prometheanfire | much of what I'm doing is converting to networkd it seems | 23:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!