*** schwicht has joined #openstack-ansible | 00:11 | |
*** schwicht has quit IRC | 00:16 | |
*** schwicht has joined #openstack-ansible | 00:18 | |
*** schwicht has quit IRC | 00:26 | |
*** macz has quit IRC | 00:32 | |
*** schwicht has joined #openstack-ansible | 00:54 | |
*** schwicht has quit IRC | 01:27 | |
*** eat_those_lemons has quit IRC | 01:47 | |
*** hamzy_ has joined #openstack-ansible | 02:15 | |
*** macz has joined #openstack-ansible | 02:30 | |
*** macz has quit IRC | 02:35 | |
guilhermesp | cores, ready for review https://review.opendev.org/#/c/702464/1 | 02:47 |
---|---|---|
*** dave-mccowan has quit IRC | 03:29 | |
*** dave-mccowan has joined #openstack-ansible | 03:52 | |
*** gyee has quit IRC | 03:56 | |
*** rh-jelabarre has quit IRC | 04:11 | |
*** macz has joined #openstack-ansible | 04:30 | |
*** macz has quit IRC | 04:34 | |
*** dave-mccowan has quit IRC | 04:42 | |
*** evrardjp has quit IRC | 05:34 | |
*** evrardjp has joined #openstack-ansible | 05:34 | |
*** shyamb has joined #openstack-ansible | 06:06 | |
*** mugsie has quit IRC | 06:11 | |
*** mugsie has joined #openstack-ansible | 06:14 | |
*** elenalindq has joined #openstack-ansible | 06:22 | |
*** dmsimard has quit IRC | 06:50 | |
*** dmsimard has joined #openstack-ansible | 06:54 | |
*** KeithMnemonic1 has joined #openstack-ansible | 07:15 | |
*** KeithMnemonic has quit IRC | 07:19 | |
*** bhyrted has joined #openstack-ansible | 07:52 | |
bhyrted | hello, is letsencrypt on multinode haproxy (haproxy + 3 servers) supported on openstack ansible - train? | 07:52 |
jrosser | bhyrted: we have code for LE so it *should* be ok | 07:57 |
jrosser | are you able to try it out? | 07:58 |
bhyrted | yeah found it on https://docs.openstack.org/openstack-ansible-haproxy_server/latest/configure-haproxy.html | 07:58 |
bhyrted | thanks ;-) | 07:58 |
bhyrted | oh well "There is no certificate distribution implementation at this time, so this will only work for a single haproxy-server environment. The renewal is automatically handled via CRON and currently will shut down haproxy briefly during the certificate renewal. The haproxy shutdown/restart will result in a brief service interruption." | 07:59 |
jrosser | yes I just saw that | 07:59 |
jrosser | someone else asked about this last week and I had some ideas about how to fix that | 08:00 |
jrosser | but it would need some work | 08:00 |
bhyrted | ok will take a look at history, as far as I remember there is one | 08:01 |
jrosser | the issue is with multiple haproxy fronted by a virtual ip the existing code can only refresh the currently active haproxy instance | 08:04 |
jrosser | feels like the cleanest approach is for each haproxy to independently renew its own cert, not using the vip | 08:06 |
jrosser | and add an alternate name for the the vip to each cert | 08:07 |
bhyrted | it appears so | 08:08 |
jrosser | you’d need a public ip and dns entry for each haproxy instance | 08:08 |
jrosser | and a quite minor adjustment to the way certbot is called | 08:08 |
bhyrted | im considering dns challenge since i cannot get any more public ips | 08:09 |
jrosser | sadly I can’t use LE at all right now, so any help with figuring this out would be great | 08:10 |
bhyrted | if I have api to dns zone o could inject txt into DNS zone | 08:10 |
bhyrted | and make letsencrypt verification happy | 08:10 |
jrosser | you’d still need to update the ansible to renew once and distribute the cert across all the haproxy | 08:10 |
jrosser | but that’s probably a fairly small change | 08:11 |
bhyrted | I think so | 08:11 |
*** cshen has joined #openstack-ansible | 08:11 | |
jrosser | or in the case of a dns challenge you could do the renewal on the deploy host, then push it to all the haproxy | 08:11 |
jrosser | lots of options | 08:12 |
*** tosky has joined #openstack-ansible | 08:19 | |
*** arxcruz|off is now known as arxcruz | 08:20 | |
*** rpittau|afk is now known as rpittau | 08:22 | |
*** shyamb has quit IRC | 08:32 | |
*** sshnaidm|afk is now known as sshnaidm | 08:38 | |
*** DanyC has joined #openstack-ansible | 09:01 | |
*** DanyC has quit IRC | 09:03 | |
*** aedc has quit IRC | 09:07 | |
*** jcosmao has joined #openstack-ansible | 09:10 | |
*** cshen has quit IRC | 09:11 | |
*** DanyC has joined #openstack-ansible | 09:19 | |
*** cshen has joined #openstack-ansible | 09:33 | |
*** pcaruana has joined #openstack-ansible | 09:42 | |
CeeMac | morning | 09:42 |
jrosser | morning | 09:55 |
*** gshippey has joined #openstack-ansible | 09:56 | |
*** shyamb has joined #openstack-ansible | 10:01 | |
*** miloa has joined #openstack-ansible | 10:02 | |
openstackgerrit | Chenjun Shen proposed openstack/openstack-ansible-os_octavia master: Fix typo. https://review.opendev.org/702638 | 10:30 |
*** ahosam has joined #openstack-ansible | 10:38 | |
*** shyamb has quit IRC | 10:40 | |
*** bhyrted has quit IRC | 10:54 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 10:56 |
openstackgerrit | Chenjun Shen proposed openstack/openstack-ansible-os_octavia master: before Ansible 2.9, os_flavor_facts os_networks_facts os_image_facts return ansible_facts. https://review.opendev.org/702645 | 11:05 |
*** melwitt has quit IRC | 11:09 | |
*** pcaruana has quit IRC | 11:12 | |
noonedeadpunk | mornings | 11:22 |
*** dave-mccowan has joined #openstack-ansible | 11:48 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 11:56 |
*** ryan3 has quit IRC | 12:01 | |
*** sshnaidm is now known as sshnaidm|afk | 12:09 | |
*** lkoranda has joined #openstack-ansible | 12:10 | |
*** lkoranda has quit IRC | 12:16 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 12:24 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Return calico to voting https://review.opendev.org/702657 | 12:31 |
*** lkoranda_ has joined #openstack-ansible | 12:33 | |
*** schwicht has joined #openstack-ansible | 12:34 | |
*** lkorand__ has joined #openstack-ansible | 12:37 | |
*** lkoranda_ has quit IRC | 12:40 | |
openstackgerrit | Merged openstack/openstack-ansible-os_octavia master: Fix typo. https://review.opendev.org/702638 | 12:43 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-ceph_client master: Use print function https://review.opendev.org/702659 | 12:44 |
*** udesale has joined #openstack-ansible | 12:45 | |
noonedeadpunk | I'm wondering how ppl with ceph and ubuntu feel on train.... | 12:45 |
jrosser | noonedeadpunk: how do you mean? | 12:47 |
noonedeadpunk | but yeah, it's more about cephfs I think... | 12:47 |
*** ansmith has quit IRC | 12:47 | |
noonedeadpunk | I mean that symlinking of libraries is broken now :( https://review.opendev.org/702659 | 12:47 |
noonedeadpunk | But just maybe these are not so popular/used ones... | 12:48 |
*** pcaruana has joined #openstack-ansible | 12:49 | |
jrosser | oh interesting, so deployments on train with ceph will have the wrong/broken links | 12:49 |
jrosser | we did one this week so if you want anything double checking it's a reasonable time to take a look | 12:49 |
noonedeadpunk | jrosser: they actually just won't have some of the libs, because rc is not 0 (because SyntaxError is raised) | 12:51 |
*** schwicht has quit IRC | 12:51 | |
ChosSimbaOne_ | Hi everyone. Can anyone help with keepalived? Must the infra nodes has assigned ip on the haproxy_keepalived_external_interface? for keepalived to work? | 12:57 |
ChosSimbaOne_ | I think that is what i can gather out of the error in syslog, but would like some confirmation. | 12:58 |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 12:59 |
*** rh-jelabarre has joined #openstack-ansible | 13:00 | |
noonedeadpunk | Hm, not sure about requirement, but I never tried to use interface without static IP for VIP. I think keepalived just use interface name for assigining ip, but not sure it will add routing... | 13:01 |
noonedeadpunk | Maybe evrardjp is a better expert in keepalived :p | 13:02 |
evrardjp | hey | 13:03 |
evrardjp | what's the question? | 13:03 |
noonedeadpunk | ChosSimbaOne_ question was: "Must the infra nodes has assigned ip on the haproxy_keepalived_external_interface? for keepalived to work?" | 13:04 |
ChosSimbaOne_ | evrardjp: must the external vip have a static ip, for the floating assinged | 13:04 |
evrardjp | thanks noonedeadpunk | 13:04 |
ChosSimbaOne_ | so that the master has both it own static from /etc/network/interfaces and a secondary from keepalived. | 13:04 |
ChosSimbaOne_ | my syslog says | 13:05 |
ChosSimbaOne_ | Jan 15 13:42:58 infra2 Keepalived_vrrp[24596]: (external): track script pingable_check_script not found, ignoring... | 13:05 |
ChosSimbaOne_ | Jan 15 13:42:58 infra2 Keepalived_vrrp[24596]: (external): Cannot find an IP address to use for interface bond0.12 | 13:05 |
ChosSimbaOne_ | Jan 15 13:42:59 infra2 Keepalived_vrrp[24596]: Stopped | 13:05 |
ChosSimbaOne_ | Jan 15 13:42:59 infra2 Keepalived_healthcheckers[24594]: Stopped | 13:05 |
evrardjp | you should define the VIP you'll use indeed. It's not really "static", as the IP will move across who is the master of the VRRP group in question | 13:05 |
ChosSimbaOne_ | when i try to start without a static ip. | 13:05 |
evrardjp | yeah nodes need to still communicate with each other, so all the nodes need to be in the vrrp group | 13:05 |
ChosSimbaOne_ | yeah, so lets say my vip for external is 10.0.92.2 and i want that to "float" between infra 1,2,3 | 13:05 |
ChosSimbaOne_ | so 10.0.92.2 is only configure in /etc/keepalived/keepalived.conf right? | 13:06 |
ChosSimbaOne_ | not in /etc/network/interfaces | 13:06 |
noonedeadpunk | oh, indeed... | 13:06 |
evrardjp | cirrect | 13:06 |
evrardjp | correct* | 13:06 |
ChosSimbaOne_ | but infra1,2,3 must have 10.0.92.11,12,13 in /etc/network/interfaces | 13:06 |
ChosSimbaOne_ | for keepalived to work | 13:06 |
evrardjp | that's recommended | 13:07 |
evrardjp | there are other ways, but you probably should do that | 13:07 |
ChosSimbaOne_ | well, I like to go by recommendations :-) | 13:07 |
ChosSimbaOne_ | super, thank you for the input, I will update my network/interfaces and add this to my documentation :-) | 13:08 |
evrardjp | it is documented in the keepalived role in fact | 13:08 |
*** lkorand__ has quit IRC | 13:08 | |
evrardjp | in the expected input of the role, you'll see the "vips" and "interface" which matter to you | 13:09 |
evrardjp | technically it is possible to use that to have a vip on a different interface than the vrrp interface | 13:09 |
noonedeadpunk | sorry for summonning you evrardjp, just clean forgot that vrrp tcp/ip for communication... | 13:09 |
noonedeadpunk | *uses tcp/ip | 13:10 |
evrardjp | I think it's udp iirc | 13:10 |
evrardjp | but yeah | 13:10 |
evrardjp | oh it is not | 13:11 |
evrardjp | anyway, you have an interface for keepalived, and then a vip on which you assign to | 13:11 |
evrardjp | the fact it is on the same interface you put your vip and you put your keepalived VRRP is just a convenience. | 13:12 |
evrardjp | technically you could even have one interface for vrrp traffic, and spin up/down all the vips (internal/external) at the same time. | 13:13 |
evrardjp | on different interfaces | 13:13 |
evrardjp | ChosSimbaOne_: it totally depends on what you want to do :) | 13:13 |
evrardjp | I prefer to have the ability to debug things per interface, but also have some kind of way to balance differently internal and external | 13:14 |
evrardjp | hope that helps | 13:14 |
evrardjp | noonedeadpunk: no worries, don't hesitate as usual :) | 13:14 |
noonedeadpunk | oh, so you can kinda use storage network for vrrp and just spawn VIP on mgmt/public? I was thinking that it might be possible but wasn't sure though | 13:15 |
evrardjp | yup it is possible | 13:15 |
noonedeadpunk | thanks for your help as ususal:) | 13:15 |
evrardjp | in the way the keepalived-role is configured, you can override vip: to pass <ip> dev <nic device>, and that's a list. | 13:16 |
evrardjp | OSA just automatically wires some things for keepalived | 13:16 |
evrardjp | so that people don't have to do it | 13:16 |
evrardjp | ChosSimbaOne_: noonedeadpunk https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/haproxy/keepalived.yml#L64 | 13:17 |
evrardjp | that's the automatic wiring for convenience. | 13:17 |
evrardjp | you could even think of overriding the whole thing if it doesn't suit you :) | 13:19 |
ChosSimbaOne_ | wow that is a lot of keepalived stuff I do not understand :-) will have to dig into that that a bit deeper. | 13:19 |
noonedeadpunk | thanks for refreshing this - I will probably need this really soon to deploy another keepalived instance with ansible for manila and ganesha-nfs... | 13:19 |
noonedeadpunk | * with osa | 13:19 |
ChosSimbaOne_ | he he, super. But i do not see where the documentation says your haproxy_keepalived_external_interface must have a static ip configured. | 13:19 |
evrardjp | ChosSimbaOne_: it doesn't _need_ one | 13:20 |
ChosSimbaOne_ | but i cannot start keepalived without? | 13:20 |
evrardjp | well that's because if you don't have one, you have to trick it :p | 13:20 |
ChosSimbaOne_ | ahh, okay. | 13:20 |
evrardjp | it's just so convenient to have one | 13:20 |
evrardjp | basically having a static ip for the network (like external), you already have all the routing and ip configuration in place | 13:21 |
evrardjp | so it means the only thing you have is an extra ip | 13:21 |
ChosSimbaOne_ | yeah, and i guess there should not be a security consern with having the infra node on a static public ip directly instead of just a single floating. | 13:21 |
evrardjp | it's also the way people are generally doing things | 13:22 |
noonedeadpunk | ChosSimbaOne_: anyway you'll have a public ip on one of the infra nodes. Otherwise you can do some proxy and use just private networks | 13:23 |
ChosSimbaOne_ | Yeah, it make total sense. My goal is just to have it working and when i tested keepalived, forcing it to failover, everything stopped, and i located the error in syslog. | 13:23 |
evrardjp | I would expect you only give the public ip of the vip | 13:23 |
evrardjp | to your users | 13:23 |
ChosSimbaOne_ | Yeah, you are right. | 13:23 |
evrardjp | however for debugging it's probably useful to be able to quickly test each server | 13:23 |
ChosSimbaOne_ | yeah, well that would be a dns record anyway. | 13:23 |
evrardjp | for example for black box monitoring | 13:23 |
ChosSimbaOne_ | yeah, good point. | 13:24 |
evrardjp | ChosSimbaOne_: correct | 13:24 |
noonedeadpunk | evrardjp: and use private network in the same vlan with public one? | 13:24 |
evrardjp | nope I wouldn't do that : p | 13:24 |
ChosSimbaOne_ | Well thank you, will will setup static ips on the infra nodes, and test keepalived again :-) | 13:24 |
evrardjp | I would keep those well separate :p | 13:24 |
noonedeadpunk | ah, you mean to give IP to _users_ :)) | 13:24 |
evrardjp | yeah | 13:24 |
evrardjp | sorry | 13:25 |
ChosSimbaOne_ | Well thank you both for the inputs and insigth to how keepalived works, I will stick to what is recommended for now :-) | 13:25 |
*** aedc has joined #openstack-ansible | 13:25 | |
*** miloa has quit IRC | 13:25 | |
evrardjp | don't hesitate to check the manpage of keepalived | 13:26 |
evrardjp | keepalived is a simple but very powerful tool. | 13:26 |
evrardjp | I don't like using it for ipvs, but some contributors of the ansible-keepalived do use it for it | 13:27 |
ChosSimbaOne_ | will do :-) Yeah it seemd nice, before i asked i setup a simple test nginx cluster of two vm's to get a better understanding :-) | 13:27 |
evrardjp | yeah OSA is by default the dead simple, not super powerful version. But you can go crazy | 13:28 |
noonedeadpunk | comparing to pacemaker it's extremely nice tool :p | 13:28 |
*** schwicht has joined #openstack-ansible | 13:28 | |
evrardjp | far simpler imo | 13:28 |
evrardjp | to operate I mean | 13:28 |
noonedeadpunk | Maybe I just don't have much experience with pacemaker, but it didn't look to simple when I'm investigating masakari issues... | 13:30 |
*** ansmith has joined #openstack-ansible | 13:33 | |
mgariepy | +1 for keepalived and -10 for pacemaker IMO. | 13:33 |
*** dave-mccowan has quit IRC | 13:33 | |
mgariepy | but i guess it depends on the need. | 13:33 |
*** dave-mccowan has joined #openstack-ansible | 13:37 | |
*** miloa has joined #openstack-ansible | 13:39 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 13:43 |
*** mnaser has quit IRC | 13:54 | |
*** mnaser has joined #openstack-ansible | 13:55 | |
*** schwicht has quit IRC | 14:04 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible-tests master: DNM - POC - Testing collect logs https://review.opendev.org/702644 | 14:10 |
*** miloa has quit IRC | 14:10 | |
jrosser | arxcruz: fwiw the huge majority of CI with OSA is not using openstack-ansible-tests any more | 14:19 |
arxcruz | jrosser: what are being used? | 14:21 |
jrosser | the openstack-ansible repo itself is used to test all of the other various service roles | 14:21 |
jrosser | so for example a patch to openstack-ansible does a minimum config with keytone/cinder/glance/db etc | 14:22 |
arxcruz | jrosser: actually, i'm trying to incorporate openstack-role-collect-logs on osa jobs | 14:23 |
jrosser | but a review in os_magnum will actually use the openstack-ansible repo with magnum enabled as its CI job | 14:23 |
jrosser | i figured | 14:23 |
arxcruz | but being honest, i don't know how to add in osa | 14:23 |
arxcruz | jrosser: any help is appreciated :D | 14:23 |
arxcruz | if you know where to start | 14:23 |
jrosser | which is why i'm suggesting that the tests repo might not actually get you many things using it | 14:24 |
arxcruz | jrosser: cool, moving to openstack-ansible | 14:24 |
jrosser | the 'entry point' for a gate job is here https://github.com/openstack/openstack-ansible/blob/master/scripts/gate-check-commit.sh | 14:24 |
jrosser | and you'll see very very similar log collection stuff in the openstack-ansible repo as there is in the tests repo, a bunch of shell basically | 14:25 |
*** miloa has joined #openstack-ansible | 14:25 | |
arxcruz | jrosser: thanks, i'll take a look | 14:26 |
noonedeadpunk | Test env is mostly prepared with https://opendev.org/openstack/openstack-ansible/src/branch/master/tests/roles/bootstrap-host | 14:27 |
arxcruz | i need to add openstack-role-collect-logs in osa ansible role path | 14:28 |
arxcruz | so, in theory, the post.yaml would just call role: collect-logs | 14:28 |
*** DanyC has quit IRC | 14:31 | |
*** schwicht has joined #openstack-ansible | 14:34 | |
jrosser | arxcruz: there are inner and outer ansibles here | 14:37 |
jrosser | post.yml will be run by the zuul outer ansible | 14:37 |
jrosser | (i think?) | 14:37 |
*** DanyC has joined #openstack-ansible | 14:38 | |
*** sshnaidm|afk is now known as sshnaidm | 14:39 | |
*** lkoranda has joined #openstack-ansible | 14:40 | |
arxcruz | jrosser: yeah, I just don't know how to make osa recognize the collect-logs role | 14:41 |
arxcruz | yet | 14:41 |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible master: DNM - POC - Collect logs https://review.opendev.org/702676 | 14:46 |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible master: DNM - POC - Collect logs https://review.opendev.org/702676 | 14:47 |
*** electrofelix has joined #openstack-ansible | 14:49 | |
*** elenalindq has quit IRC | 14:52 | |
*** aedc has quit IRC | 14:52 | |
evrardjp | mgariepy: tbh pacemaker is good and very powerful. It's just doesn't let room for failure in operations. | 14:55 |
evrardjp | and ... hello, happy new year! | 14:55 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Add manila tempest testing https://review.opendev.org/674653 | 14:56 |
*** macz has joined #openstack-ansible | 14:57 | |
noonedeadpunk | evrardjp: like with pacemaker+corosync I'm facing split brain pretty regulary - when one node reports that it's the only cluster participant, while all other nodes claims that only this exact node is not available... | 14:58 |
noonedeadpunk | * claims that only _active_ participant (all other members are down) | 14:59 |
noonedeadpunk | maybe there's some hint regarding that... | 14:59 |
noonedeadpunk | and masakari in such situation just marks everything as down lol | 15:00 |
openstackgerrit | Chenjun Shen proposed openstack/openstack-ansible-os_octavia master: The variables exist in ansible_facts https://review.opendev.org/702645 | 15:11 |
cshen | jrosser: see above, I just polished my patch. | 15:13 |
cshen | If you need more clarifications, I could explain here. I feel it's a bug in the master branch. | 15:14 |
*** l3d has quit IRC | 15:14 | |
jrosser | noonedeadpunk: regarding log collection: | 15:15 |
jrosser | fungi> jrosser: there was a necessary change in handling of compressed log files, jobs should not compress logs on their own now if you want them to be viewable in a browser (this was to solve a problem where tarballs were getting silently decompressed on download without changing their filenames to drop the .gz extension) | 15:15 |
jrosser | a bunch of our logs are broken in the zuul UI now because they render as .gz raw | 15:15 |
noonedeadpunk | :( | 15:16 |
noonedeadpunk | I read through ML but hoped we're ok | 15:16 |
noonedeadpunk | but yeah, I see now we are not... | 15:17 |
jrosser | i think its fairly easy as our log collect scripts have a quite obvious place that the compression is done | 15:18 |
noonedeadpunk | Anyway I will probably have no time this week for looking into this :( | 15:18 |
noonedeadpunk | you mean https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/log-collect.sh#L159-L166 ? | 15:18 |
jrosser | yes i think it might be as simple as just getting rid of that | 15:19 |
jrosser | but i've not looked at the detail yet | 15:19 |
noonedeadpunk | Sounds good and reasonable | 15:19 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Remove log compression before upload https://review.opendev.org/702686 | 15:47 |
jrosser | cshen: i don't fully understand your patch, do you think the code is actually broken? | 15:52 |
jrosser | I have an octavia deployment here which seemed to go ok | 15:53 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] ceph-mds setup https://review.opendev.org/702424 | 15:54 |
*** udesale_ has joined #openstack-ansible | 15:54 | |
openstackgerrit | Arx Cruz proposed openstack/openstack-ansible master: DNM - POC - Collect logs https://review.opendev.org/702676 | 15:55 |
cshen | jrosser: which Ansible version do you use? | 15:55 |
jrosser | train branch, so 2.8.x | 15:56 |
cshen | ok, code here https://opendev.org/openstack/openstack-ansible-os_octavia/src/branch/master/tasks/octavia_post_install.yml#L50-L66 | 15:56 |
noonedeadpunk | do we have tempest testing of octavia or it's another role without proper ci now? | 15:57 |
cshen | you're supposed to only get the flavors of octavia service project. | 15:57 |
*** udesale has quit IRC | 15:57 | |
cshen | but variable openstack_flavors will be populated with all flavors. | 15:57 |
noonedeadpunk | ok, we do not test it.... https://8120cec50dbf22b0ed97-a15027182aab035aa882f99410b51a23.ssl.cf2.rackcdn.com/702645/1/check/openstack-ansible-deploy-aio_metal-ubuntu-bionic/f199b3e/logs/openstack/aio1-utility/stestr_results.html | 15:57 |
cshen | only ansible_facts.openstack_flavors will be populated with correct flavors. | 15:58 |
cshen | or get_flavor_facts.ansible_facts.openstack_flavors. | 15:58 |
cshen | you coudl DEDUG the variable openstack_flavors while the installation. | 15:59 |
*** DanyC has quit IRC | 15:59 | |
cshen | at least, in our installation, i watched the wrong behavior. | 15:59 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] ceph-mds setup https://review.opendev.org/702424 | 16:00 |
*** melwitt has joined #openstack-ansible | 16:00 | |
*** gyee has joined #openstack-ansible | 16:03 | |
*** DanyC has joined #openstack-ansible | 16:05 | |
mgariepy | evrardjp, you too happy new year! | 16:08 |
cshen | jrosser: sorry, I need go home now. We could talk tomorrow. | 16:11 |
*** cshen has quit IRC | 16:18 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] ceph-mds setup https://review.opendev.org/702424 | 16:20 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Add manila tempest testing https://review.opendev.org/674653 | 16:31 |
jrosser | cshen: heres a debug of openstack_flavors from here http://paste.openstack.org/show/788420/ | 16:36 |
jrosser | it seems to be showing be only one? | 16:36 |
jrosser | oh, he's gone for the day | 16:38 |
noonedeadpunk | yeah | 16:39 |
jrosser | i'm a bit puzzled because we did have to fix a whole ton of stuff when we did octavia | 16:40 |
jrosser | but not this | 16:40 |
noonedeadpunk | Yeah, I thought that octavia should be working now... | 16:40 |
noonedeadpunk | Actually I think I don;'t have any working deployment where I'd feel safe to check that... | 16:41 |
noonedeadpunk | (in terms of cotavia) | 16:41 |
noonedeadpunk | *octavia | 16:41 |
*** openstackgerrit has quit IRC | 16:43 | |
*** pcaruana has quit IRC | 16:45 | |
*** udesale_ has quit IRC | 16:45 | |
*** openstackgerrit has joined #openstack-ansible | 16:45 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-ceph_client master: Use print function https://review.opendev.org/702659 | 16:45 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_manila master: Add manila CI https://review.opendev.org/675934 | 16:47 |
jrosser | noonedeadpunk: i was a bit hesitating running the playbook too so i cut out everything except the post_install tasks | 16:57 |
*** miloa has quit IRC | 16:58 | |
openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Use appropriate vpnaas drivers https://review.opendev.org/702256 | 17:05 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron stable/train: Use appropriate vpnaas drivers https://review.opendev.org/702705 | 17:06 |
jrosser | anyone here using keystone federation? | 17:08 |
*** rpittau is now known as rpittau|afk | 17:15 | |
noonedeadpunk | cshen so with ansible 2.7.5 on stein I got the same result as jrosser did: http://paste.openstack.org/show/788424/ | 17:21 |
noonedeadpunk | do you deploy train or master? | 17:21 |
*** DanyC has quit IRC | 17:29 | |
*** arxcruz is now known as arxcruz|off | 17:29 | |
*** guilhermesp has quit IRC | 17:31 | |
*** guilhermesp has joined #openstack-ansible | 17:32 | |
*** evrardjp has quit IRC | 17:34 | |
*** evrardjp has joined #openstack-ansible | 17:34 | |
openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/train: Make use of horizon_git_track_branch https://review.opendev.org/702464 | 17:59 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible stable/train: Bump SHAs for stable/train https://review.opendev.org/702135 | 18:03 |
noonedeadpunk | guilhermesp: can you check this thing? https://review.opendev.org/#/c/702686/ | 18:05 |
*** jcosmao has left #openstack-ansible | 18:06 | |
*** electrofelix has quit IRC | 18:07 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Add manila tempest testing https://review.opendev.org/674653 | 18:12 |
openstackgerrit | Georgina Shippey proposed openstack/openstack-ansible-os_keystone master: Openstacksdk is not available in all installs https://review.opendev.org/702714 | 18:17 |
openstackgerrit | Georgina Shippey proposed openstack/openstack-ansible-os_keystone master: Openstacksdk is not available in all installs https://review.opendev.org/702714 | 18:18 |
*** mwhahaha has quit IRC | 18:30 | |
*** rpittau|afk has quit IRC | 18:30 | |
*** cyberpear has quit IRC | 18:30 | |
*** donnyd has quit IRC | 18:30 | |
*** soren has quit IRC | 18:30 | |
*** mnaser has quit IRC | 18:31 | |
*** donnyd has joined #openstack-ansible | 18:31 | |
*** jonmasters has quit IRC | 18:31 | |
*** cyberpear has joined #openstack-ansible | 18:32 | |
*** openstackstatus has quit IRC | 18:32 | |
*** mwhahaha has joined #openstack-ansible | 18:32 | |
*** Nick_A has quit IRC | 18:33 | |
*** soren has joined #openstack-ansible | 18:33 | |
*** johnsom has quit IRC | 18:33 | |
*** tosky has quit IRC | 18:33 | |
*** mnaser has joined #openstack-ansible | 18:33 | |
*** johnsom has joined #openstack-ansible | 18:35 | |
*** sshnaidm is now known as sshnaidm|afk | 18:38 | |
*** donnyd has quit IRC | 18:40 | |
*** donnyd has joined #openstack-ansible | 18:42 | |
*** cyberpear has quit IRC | 18:43 | |
*** oligau has quit IRC | 18:45 | |
*** oligau has joined #openstack-ansible | 18:48 | |
*** cyberpear has joined #openstack-ansible | 19:00 | |
*** sshnaidm|afk is now known as sshnaidm | 19:09 | |
*** openstackstatus has joined #openstack-ansible | 19:21 | |
*** ChanServ sets mode: +v openstackstatus | 19:21 | |
*** schwicht has quit IRC | 19:43 | |
*** schwicht has joined #openstack-ansible | 19:44 | |
*** gshippey has quit IRC | 20:35 | |
*** ryan3 has joined #openstack-ansible | 20:43 | |
*** aedc has joined #openstack-ansible | 20:49 | |
*** aedc has quit IRC | 20:54 | |
*** aedc has joined #openstack-ansible | 20:54 | |
*** DanyC has joined #openstack-ansible | 21:01 | |
*** ansmith has quit IRC | 21:39 | |
*** CeeMac has quit IRC | 21:40 | |
*** ansmith has joined #openstack-ansible | 22:54 | |
*** schwicht has quit IRC | 23:29 | |
*** schwicht has joined #openstack-ansible | 23:55 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!