| *** tosky has quit IRC | 00:13 | |
| jamesdenton | thanks cloudnull and jrosser! | 00:23 |
|---|---|---|
| *** weezS has quit IRC | 00:34 | |
| *** amab has quit IRC | 01:04 | |
| *** cshen has quit IRC | 01:37 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 01:56 |
| pabelanger | how do lxc containers get their IP addressing? | 03:00 |
| pabelanger | I'm trying to bootstrap a new controller, to confirm my deployment process works, but this time around, not of the containers can be SSH into | 03:00 |
| pabelanger | seem IP address haven't been setup | 03:00 |
| *** maddtux has joined #openstack-ansible | 03:01 | |
| pabelanger | https://object-storage-ca-ymq-1.vexxhost.net/v1/a0b4156a37f9453eb4ec7db5422272df/logs/e9/e9d0ce46f58e74abefc15558e6c70bff84e9bbe7/post/packet-ci-cloud-deploy/6975fbe/job-output.html#l3246 | 03:02 |
| pabelanger | show ansible unable to gather facts | 03:02 |
| *** chhagarw has joined #openstack-ansible | 03:59 | |
| *** ram5391 has joined #openstack-ansible | 04:53 | |
| *** udesale has joined #openstack-ansible | 04:54 | |
| *** pcaruana has joined #openstack-ansible | 05:09 | |
| *** mma has joined #openstack-ansible | 05:11 | |
| *** ram5391 has quit IRC | 05:11 | |
| *** mma has quit IRC | 05:15 | |
| *** lemouchon has quit IRC | 05:48 | |
| *** cshen has joined #openstack-ansible | 06:04 | |
| *** cshen has quit IRC | 06:09 | |
| *** gyee has quit IRC | 06:20 | |
| *** cshen has joined #openstack-ansible | 06:22 | |
| *** aedc has quit IRC | 06:25 | |
| *** cshen has quit IRC | 06:27 | |
| *** ahosam has joined #openstack-ansible | 06:29 | |
| *** udesale has quit IRC | 06:43 | |
| *** udesale has joined #openstack-ansible | 06:44 | |
| *** udesale has quit IRC | 06:45 | |
| *** udesale has joined #openstack-ansible | 06:46 | |
| *** udesale has quit IRC | 06:57 | |
| *** chkumar|away is now known as chkumar|ruck | 06:59 | |
| *** gkadam has joined #openstack-ansible | 07:00 | |
| *** radeks has joined #openstack-ansible | 07:07 | |
| jrosser | pabelanger: OSA does not SSH directly to the containers, there is a connection plugin which works out what physical_host is for each container, ssh to there then uses a local technique to execute the ansible against the correct container | 07:08 |
| jrosser | so "no route to host" is maybe more to do with the physical host in the case than the container | 07:09 |
| jrosser | pabelanger: i also see that mitogen is installed - is ansible configured to use that in your setup? | 07:11 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_nova master: Ensure that UEFI firmware is present https://review.openstack.org/620026 | 07:12 |
| *** rgogunskiy has quit IRC | 07:28 | |
| *** dcdamien has joined #openstack-ansible | 07:34 | |
| *** hamzaachi has joined #openstack-ansible | 07:45 | |
| *** DanyC has joined #openstack-ansible | 07:51 | |
| *** udesale has joined #openstack-ansible | 07:53 | |
| *** DanyC has quit IRC | 07:56 | |
| *** mma has joined #openstack-ansible | 07:57 | |
| *** udesale has quit IRC | 07:58 | |
| *** rpittau has joined #openstack-ansible | 08:01 | |
| jonher | Morning, could we get some +workflow on https://review.openstack.org/619092/ and https://review.openstack.org/581248/ Thank you :) | 08:04 |
| openstackgerrit | Merged openstack/openstack-ansible-haproxy_server stable/rocky: Add 'absent' service state https://review.openstack.org/620337 | 08:06 |
| *** aedc has joined #openstack-ansible | 08:08 | |
| Miouge- | Good morning everyone | 08:09 |
| *** jawad_axd has joined #openstack-ansible | 08:10 | |
| *** udesale has joined #openstack-ansible | 08:11 | |
| jrosser | Morning | 08:16 |
| *** hamzaachi has quit IRC | 08:19 | |
| *** rpittau_ has joined #openstack-ansible | 08:31 | |
| *** ahosam has quit IRC | 08:31 | |
| *** hamzaachi has joined #openstack-ansible | 08:31 | |
| *** ahosam has joined #openstack-ansible | 08:31 | |
| *** hamzaachi has quit IRC | 08:32 | |
| *** hamzaachi has joined #openstack-ansible | 08:32 | |
| *** hamzaachi has quit IRC | 08:33 | |
| *** rpittau has quit IRC | 08:33 | |
| *** hamzaachi has joined #openstack-ansible | 08:34 | |
| *** rpittau_ is now known as rpittau | 08:35 | |
| *** rpittau_ has joined #openstack-ansible | 08:40 | |
| *** rpittau has quit IRC | 08:40 | |
| *** rgogunskiy has joined #openstack-ansible | 08:41 | |
| *** rgogunskiy has quit IRC | 08:42 | |
| *** rgogunskiy has joined #openstack-ansible | 08:44 | |
| *** hamzaachi has quit IRC | 08:45 | |
| *** hamzaachi has joined #openstack-ansible | 08:46 | |
| *** rpittau_ is now known as rpittau | 08:49 | |
| *** tosky has joined #openstack-ansible | 08:51 | |
| *** aedc has quit IRC | 09:02 | |
| *** shardy has joined #openstack-ansible | 09:10 | |
| *** ahosam has quit IRC | 09:10 | |
| *** ahosam has joined #openstack-ansible | 09:10 | |
| *** rgogunskiy has quit IRC | 09:11 | |
| openstackgerrit | Andriy Shevchenko proposed openstack/openstack-ansible-os_barbican master: fix tox python3 overrides https://review.openstack.org/606825 | 09:14 |
| *** dcdamien has quit IRC | 09:20 | |
| evrardjp | morning everyone | 09:21 |
| jonher | morning \o | 09:21 |
| *** DanyC has joined #openstack-ansible | 09:30 | |
| *** DanyC has quit IRC | 09:30 | |
| *** DanyC has joined #openstack-ansible | 09:31 | |
| *** rpittau has quit IRC | 09:34 | |
| jrosser | would be good to get some reviews on this to improve centos gate reliability https://review.openstack.org/#/c/620313/ | 09:49 |
| *** ahosam has quit IRC | 09:49 | |
| Miouge- | evrardjp: I saw your comment on the LetsEncrypt patch (586774) “If someone wants to implement a different way that's not killing haproxy connections with a lua backend, up to him/her :)”. I am not familiar with the HAProxy Lua client, but can't we redirect the /.well-known/acme-challenge to the certbot standalone webserver in a new LXC container? (idea inspired from https://serversforhackers.com/c/letsencrypt-with-haproxy ) | 10:00 |
| *** cshen has joined #openstack-ansible | 10:05 | |
| *** dcdamien has joined #openstack-ansible | 10:05 | |
| *** hamzaachi has quit IRC | 10:08 | |
| *** hamzaachi has joined #openstack-ansible | 10:09 | |
| *** tobias-urdin has quit IRC | 10:10 | |
| *** cshen has quit IRC | 10:10 | |
| *** cshen has joined #openstack-ansible | 10:12 | |
| evrardjp | Miouge-: we can | 10:17 |
| evrardjp | Miouge-: But we can probably use haproxy's internal system to deal with that | 10:17 |
| evrardjp | assuming real passthrough is not used ofc | 10:18 |
| *** ahosam has joined #openstack-ansible | 10:19 | |
| noonedeadpunk | morning folks | 10:22 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Add ability to deploy Ceph https://review.openstack.org/607338 | 10:26 |
| evrardjp | morning noonedeadpunk | 10:26 |
| noonedeadpunk | I've got a bit offtopic question (not directly related to OSA). Does anybody knows/remeber how to make nova work with libvirt 4? It's kinda P->Q upgrade question.... | 10:27 |
| noonedeadpunk | It seems, that nova should generate config for VMs without serial, but it still includes it, which results in "'serial' is deprecated, please use the corresponding option of '-device' instead" | 10:30 |
| odyssey4me | noonedeadpunk I don't know if there's a solution - you might need to ask in #openstack-nova | 10:31 |
| odyssey4me | A workaround would be to create a fork of nova which does what's needed, then use that fork. We give you the tools to do that after all. ;) | 10:31 |
| noonedeadpunk | Yeah, I've just thought that's it's something, that you faced with (and probably fixed) during P->Q upgrade (as Ubuntu in P uses libvirt 3.6, and in Q it's already 4.0) | 10:32 |
| odyssey4me | noonedeadpunk maybe there's a setting somewhere? or maybe this is a nova bug, actually | 10:33 |
| Miouge- | evrardjp: it doesn’t look like the haproxy-lua thing works with several haproxy servers, it would need to share the challenge, keys and certificates with the other haproxy servers | 10:34 |
| *** hamzaachi has quit IRC | 10:35 | |
| evrardjp | Miouge-: you probably want to share the certificates anyway, as they are sharing the same endpoints | 10:36 |
| noonedeadpunk | odyssey4me may be:) anyway thanks, at least now I know, that it's not smth common, solution to which I just wasn't able to find | 10:36 |
| evrardjp | but yeah I'd be surprised on how to rotate certificates on sub nodes. | 10:37 |
| evrardjp | whether with a certbot in a container or not | 10:37 |
| Miouge- | evrardjp: Yep, it needs a mechanism to copy challenge/key/cert to other haproxy servers. AFAIK haproxy servers are not really aware of each other | 10:38 |
| evrardjp | lsyncd should be enough | 10:38 |
| *** electrofelix has joined #openstack-ansible | 10:39 | |
| evrardjp | Miouge-: I haven't digged deep into what it takes to reload a certificate from haproxy. That's why I left that comment | 10:40 |
| evrardjp | yeah the 'with a lua backend' should have been '(with a lua backend, for example)' | 10:40 |
| Miouge- | evrardjp: I think I start to see how this could work: ACME client (lua/certbot/whatever) + lsyncd + something to reload. A little similar to ingress-nginx in Kubernetes | 10:45 |
| *** Miouge- has left #openstack-ansible | 10:46 | |
| *** Miouge- has joined #openstack-ansible | 10:46 | |
| *** vnogin has joined #openstack-ansible | 10:47 | |
| *** ahosam has quit IRC | 10:57 | |
| *** hamzaachi has joined #openstack-ansible | 10:57 | |
| *** vakuznet has quit IRC | 11:19 | |
| *** vnogin has quit IRC | 11:21 | |
| openstackgerrit | Debo Zhang proposed openstack/openstack-ansible master: Fix Chinese quotes https://review.openstack.org/620566 | 11:22 |
| *** chason has quit IRC | 11:31 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Update playbook to newer syntax. https://review.openstack.org/620569 | 11:32 |
| odyssey4me | hwoarang evrardjp so ceph on stable/rocky for suse is broken, and we have https://review.openstack.org/619796 proposed to resolve that because the stable-3.2 branch has the right fixes... there was recentlya patch to move stable/rocky to use stable-3.1 because stable-3.2 is still a release candidate... so we need to decide - do we accept a release candidate for our rocky release, or do we make ceph/suse non voting while a fix is | 11:35 |
| odyssey4me | implemented upstream in stable-3.1 | 11:35 |
| odyssey4me | implemented upstream in stable-3.1 | 11:35 |
| odyssey4me | the revert to use stable-3.1 is here: https://review.openstack.org/616481 | 11:36 |
| *** mma has quit IRC | 11:40 | |
| *** dave-mccowan has joined #openstack-ansible | 11:44 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Ensure that LDAP config is deployed on all keystone hosts https://review.openstack.org/620574 | 11:45 |
| *** cshen has quit IRC | 11:47 | |
| *** mma has joined #openstack-ansible | 11:48 | |
| *** hamzaachi has quit IRC | 11:50 | |
| *** mma has quit IRC | 11:54 | |
| openstackgerrit | Vieri proposed openstack/openstack-ansible-os_cloudkitty master: fix tox python3 overrides https://review.openstack.org/606819 | 11:54 |
| *** cshen has joined #openstack-ansible | 12:14 | |
| *** hamzaachi has joined #openstack-ansible | 12:14 | |
| *** ansmith has quit IRC | 12:17 | |
| *** cshen has quit IRC | 12:18 | |
| *** cshen has joined #openstack-ansible | 12:20 | |
| jamesdenton | how does one rerun a gate job that failed? | 12:21 |
| kaiokmo | jamesdenton: recheck? | 12:24 |
| jamesdenton | is that sufficient for the +2 as well? | 12:25 |
| *** cshen has quit IRC | 12:25 | |
| *** cshen has joined #openstack-ansible | 12:28 | |
| kaiokmo | even if you recheck it will probably mantain +2cr and +1w IIRC, and it should be enough | 12:28 |
| kaiokmo | maintain* | 12:30 |
| *** ahosam has joined #openstack-ansible | 12:33 | |
| *** maddtux has quit IRC | 12:53 | |
| *** fghaas has joined #openstack-ansible | 12:56 | |
| *** hamzaachi has quit IRC | 13:06 | |
| *** ahosam has quit IRC | 13:13 | |
| mnaser | jamesdenton: its not possible to rerun gate jobs only | 13:17 |
| mnaser | a recheck will go through both check and gate | 13:17 |
| *** strattao has joined #openstack-ansible | 13:17 | |
| jamesdenton | bummer, but OK. thanks mnaser. and kaiokmo. | 13:17 |
| mnaser | jamesdenton: #makejobsstableagain | 13:18 |
| jamesdenton | no kidding | 13:18 |
| mnaser | anything you're struggling with jamesdenton ? | 13:18 |
| jamesdenton | not at the moment. stupid gate timeout | 13:18 |
| jamesdenton | today is the day it's gonna work, right? | 13:19 |
| jrosser | evrardjp: rechecking https://review.openstack.org/#/c/618822/ isnt going to work - we need to fix ceph/suse first? | 13:19 |
| evrardjp | I thought this was a temp faire | 13:19 |
| evrardjp | failure | 13:19 |
| jrosser | no, it is brok | 13:19 |
| evrardjp | as aio distro suse worked for this branch | 13:20 |
| evrardjp | I see | 13:20 |
| jrosser | see the comments earlier about https://review.openstack.org/619796 | 13:20 |
| evrardjp | http://logs.openstack.org/22/618822/1/check/openstack-ansible-deploy-distro_ceph-opensuse-423/cb90b16/job-output.txt.gz#_2018-11-26_17_28_49_952237 | 13:20 |
| evrardjp | ok | 13:21 |
| jrosser | yes that - ceph-ansible 3.1 does not set that var for suse at all | 13:21 |
| mnaser | ah centos was the culprit | 13:21 |
| mnaser | and opensuse-150 | 13:21 |
| jamesdenton | for me? yes | 13:22 |
| mnaser | http://logs.openstack.org/69/584069/14/gate/openstack-ansible-deploy-aio_lxc-centos-7/aa06a7e/logs/ara-report/ | 13:22 |
| mnaser | doesnt seem like anything that took much longer in there | 13:23 |
| evrardjp | jrosser: I must say I love this kind of comments: https://github.com/ceph/ceph-ansible/issues/3078 | 13:23 |
| mnaser | wait what | 13:24 |
| mnaser | ceph-ansible is not maintained? | 13:24 |
| evrardjp | taht's what I read. | 13:24 |
| *** hamzaachi has joined #openstack-ansible | 13:24 | |
| mnaser | i think maybe | 13:24 |
| mnaser | they meant to say | 13:24 |
| mnaser | not maintained for coreos | 13:24 |
| mnaser | ? | 13:24 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 13:25 |
| evrardjp | I sure hope so | 13:25 |
| evrardjp | I hope it's not moving away to be container only | 13:25 |
| kaiokmo | README does not says that it is not maintained | 13:25 |
| kaiokmo | and there is a lot of recent commits, so | 13:25 |
| mnaser | https://github.com/ceph/ceph-ansible/commit/d749dc9f365643f0be171efb7f934be7d763582f | 13:26 |
| mnaser | i think this is what they mean | 13:26 |
| evrardjp | kaiokmo: neither does the docs | 13:26 |
| kaiokmo | mnaser: yeah probably | 13:27 |
| evrardjp | mnaser: makes sense | 13:27 |
| evrardjp | still need the ceph_uid there, and we should probably make it -nv until it's resolved | 13:27 |
| evrardjp | I don't have the cycles right now | 13:27 |
| mnaser | im curious more on how | 13:27 |
| mnaser | it broke. | 13:27 |
| evrardjp | mnaser: we updated the shas | 13:27 |
| mnaser | so did ceph-ansible go from *not* having it to *having* it? | 13:28 |
| jrosser | We accidentally were using master | 13:28 |
| evrardjp | this ceph_uid var? yes | 13:28 |
| jrosser | Fixed that | 13:28 |
| mnaser | ah | 13:28 |
| mnaser | okay | 13:28 |
| jrosser | So it went away | 13:28 |
| evrardjp | ceph_uid var is a 'new var' for us, that is well handled in ubuntu/centos. It appeared during the change of sha (move to track 3.2 vs master) | 13:29 |
| jrosser | And when it got rolled back to 3.1 I don’t think 3.2 was a proper thing, still not quite sure about that | 13:29 |
| evrardjp | sorry for me using half sentences sometimes | 13:29 |
| evrardjp | oh it's maybe 3.1 I can't remember | 13:29 |
| evrardjp | anyway | 13:29 |
| *** fghaas has quit IRC | 13:30 | |
| pabelanger | jrosser: yah, after I feel a sleep I figured I should retest without mitogen and see if that was the problem. Looks like it is, I'll keep it disabled for now. But strange, I thought the containers did get an IP address. | 13:30 |
| *** fghaas has joined #openstack-ansible | 13:32 | |
| *** strattao has quit IRC | 13:33 | |
| jrosser | pabelanger: they do indeed get an IP on the mgmt network, but thats now how ansible connects to them in this case | 13:33 |
| *** chason has joined #openstack-ansible | 13:33 | |
| pabelanger | ack | 13:34 |
| jrosser | pabelanger: i have done tests with mitogen in an AIO and it sort of works, but nothing yet on multiple nodes - some thought would be needed about passing physical_host into mitogen_via | 13:34 |
| pabelanger | jrosser: yup, I think there is some weirdness going on too. I've just disabled it for now, as I was just testing how it worked | 13:36 |
| jrosser | delegation is broken really, and it will eventually blow up here https://github.com/dw/mitogen/issues/414 | 13:39 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Fix warning for guestfish workaround task https://review.openstack.org/620599 | 13:43 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: If there is no secondary data disk, use the root disk https://review.openstack.org/620600 | 13:43 |
| odyssey4me | mnaser evrardjp the job to make the suse job voting was created before the switch of the ceph-ansible sha, and it merged after the switch of the ceph-ansible sha | 13:45 |
| odyssey4me | zuul clearly didn't do a rebase in that case | 13:45 |
| odyssey4me | mnaser evrardjp jrosser so the question I posed earlier stands - are we prepared to use the not-yet-release stable-3.2 branch for rocky to ensure suse support works... or do we make suse support non-voting until an upstream patch is done and merged | 13:47 |
| mnaser | odyssey4me: i think its opensuse 150 which is broken right? not 423? | 13:48 |
| evrardjp | I don't have the time to fix that today, so I suggest we move to nv, add a bug, and maybe assign that to me | 13:48 |
| evrardjp | mnaser: 42.3 | 13:48 |
| evrardjp | is broken | 13:48 |
| evrardjp | only for ceph with distro packages | 13:48 |
| evrardjp | https://review.openstack.org/#/c/618822/ | 13:48 |
| mnaser | i would rather move one distro to nv rather than potentially risk the others | 13:49 |
| jrosser | agreed - it's also not very clear from the ceph-ansible repo is 3.2 is released or not | 13:49 |
| odyssey4me | ok, if you're happy with that, so am I | 13:49 |
| jrosser | theres a branch an a release note | 13:49 |
| odyssey4me | 3.2 has not been tagged - only RC tags | 13:49 |
| evrardjp | I'd rather stay on 3.1 too | 13:49 |
| jrosser | aahhhhh | 13:49 |
| odyssey4me | ok, I'll work up the patch to switch to non-voting then and do the bug for evrardjp - thanks evrardjp | 13:50 |
| evrardjp | well if I had more time I'd fix that now, but I have plenty on my plate today | 13:50 |
| odyssey4me | mnaser jrosser I'm still looking for a review for https://review.openstack.org/551791 | 13:51 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-os_neutron master: Modify OVS hostname value to match server's hostname https://review.openstack.org/620604 | 13:51 |
| mnaser | ah dang odyssey4me i got distracted | 13:52 |
| mnaser | going over it now | 13:52 |
| mnaser | odyssey4me: lgtm. :) | 13:56 |
| *** cshen_ has joined #openstack-ansible | 14:04 | |
| mgariepy | someone have better eyes than me? http://logs.openstack.org/40/620140/1/check/openstack-ansible-linters/0e3897a/job-output.txt.gz#_2018-11-27_14_45_39_254393 | 14:04 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Set SUSE ceph-distro job to non-voting https://review.openstack.org/620607 | 14:05 |
| *** cshen has quit IRC | 14:06 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Remove unnecessary octavia scenario AIO bootstrap https://review.openstack.org/619792 | 14:10 |
| *** vakuznet has joined #openstack-ansible | 14:10 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Ensure AIO container_tech/install_method vars are namespaced https://review.openstack.org/620278 | 14:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Implement documentation changes for translations https://review.openstack.org/620279 | 14:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: [docs] Clean up the AIO user story https://review.openstack.org/620280 | 14:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Update all SHAs for 18.1.1 https://review.openstack.org/618822 | 14:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Update playbook to newer syntax. https://review.openstack.org/620569 | 14:11 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Add extra volume types to AIO https://review.openstack.org/617053 | 14:12 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Move ARA install to end of bootstrap https://review.openstack.org/617785 | 14:12 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: haproxy: remove repo_cache service https://review.openstack.org/620336 | 14:12 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Add MariaDB infrastructure mirrors https://review.openstack.org/619714 | 14:12 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible stable/rocky: Add openstackclient bash completion https://review.openstack.org/619354 | 14:13 |
| odyssey4me | mgariepy when they're in the same repo - no need to use depends-on, better to just rebase on top of the patch it depends on | 14:13 |
| odyssey4me | depends-on is actually for cross-repo depenencies | 14:14 |
| odyssey4me | no matter - it still works :) | 14:14 |
| mgariepy | ok | 14:14 |
| mgariepy | next time :D | 14:15 |
| *** ansmith has joined #openstack-ansible | 14:16 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 14:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: haproxy: remove repo_cache service https://review.openstack.org/620336 | 14:18 |
| odyssey4me | jrosser could you cast your eyes on https://review.openstack.org/#/c/614194/ - it's been sitting a while | 14:21 |
| odyssey4me | and it needs your ceph expertise | 14:21 |
| odyssey4me | or logan- if you're available | 14:22 |
| *** udesale has quit IRC | 14:24 | |
| *** udesale has joined #openstack-ansible | 14:25 | |
| pabelanger | logan-: I seem to be having some routing issue getting to keystone from my nodepool node: https://object-storage-ca-ymq-1.vexxhost.net/v1/a0b4156a37f9453eb4ec7db5422272df/logs/47/474cf5544baa48eb04a6309395ba8c8c5194a083/post/packet-ci-cloud-deploy/f23881f/job-output.html#l6337 | 14:33 |
| pabelanger | logan-: curious if you had this issue or I missed something | 14:34 |
| *** strattao has joined #openstack-ansible | 14:36 | |
| mnaser | pabelanger: looks like you've deployed keystone on the internal ip | 14:40 |
| mnaser | which is unreachable | 14:40 |
| mnaser | that job is delegated to localhsot | 14:41 |
| mnaser | so services (aka keystone) must be reachable | 14:41 |
| pabelanger | mnaser: if I am reading right, localhost is trying to reach the network, and you are right that range is only exposed via SSH proxyhost | 14:41 |
| pabelanger | but, I think logan- is doing the same thing | 14:42 |
| *** fghaas has quit IRC | 14:42 | |
| logan- | im also running pike which does not have the new delegation setup yet afaik | 14:43 |
| *** vnogin has joined #openstack-ansible | 14:43 | |
| pabelanger | Oh | 14:43 |
| pabelanger | ah | 14:43 |
| logan- | but i think you could just delegate that task to the utility container instead | 14:43 |
| logan- | iirc there is a var to configure the delegation | 14:43 |
| pabelanger | Yah, I suspect I need to setup an HTTP proxy for that network | 14:44 |
| pabelanger | if the deployment host needs to do HTTP things with it | 14:44 |
| pabelanger | logan-: any ideas where I make that change? | 14:44 |
| logan- | it shouldn't need that ability, you should be able to delegate the task so the http request comes from the utility container | 14:44 |
| logan- | not sure, checking | 14:45 |
| pabelanger | okay, same looking to see how it works | 14:45 |
| logan- | keystone_service_setup_host | 14:47 |
| *** cshen_ has quit IRC | 14:47 | |
| *** cshen has joined #openstack-ansible | 14:47 | |
| logan- | so user_variables.yml, set keystone_service_setup_host: "{{ groups['utility_all'][0] }}" and see if that works | 14:47 |
| logan- | actually youll hit this problem with all services, so set openstack_service_setup_host instead | 14:48 |
| logan- | keystone_service_setup_host inherits that var | 14:48 |
| logan- | heh even found a reno :) https://git.openstack.org/cgit/openstack/openstack-ansible-os_keystone/tree/releasenotes/notes/keystone-service-setup-host-cd3ee3346af823e6.yaml | 14:48 |
| pabelanger | okay, creating PR now | 14:49 |
| pabelanger | logan-: is keystone the only service that does this now, or will I have to do it for every openstack service | 14:50 |
| pabelanger | sorry, not familiar with the change | 14:50 |
| logan- | if you set openstack_service_setup_host it should fix it for all openstack services | 14:51 |
| pabelanger | kk | 14:51 |
| pabelanger | https://git.openstack.org/cgit/openstack/openstack-ansible/tree/releasenotes/notes/openstack-service-setup-host-f38d655eed285f57.yaml | 14:53 |
| antonym | odyssey4me: so this sshd version tag fun... do we have a process for backport a fix to eol versions? it's breaking the leapfrog code because it bootstraps each version as it goes, i guess the other alternative would be just to fork all of the branches to another repo, fix them and point leapfrog to that | 14:53 |
| odyssey4me | antonym pike is done, as it ocata - anything else is EOL and will require a fork | 14:54 |
| antonym | ok, so i'll put up something to the leap code to make the osa repo location a variable then | 14:55 |
| odyssey4me | antonym yeah, not so much fun - but c'est la vie | 14:56 |
| antonym | yeah, my main concern it would break other users trying the leap code and that don't have the fixed repo | 14:56 |
| antonym | will probably have to include the forked repo with the fixes in the notes then | 14:56 |
| odyssey4me | antonym another option could be to implement something in the leap code to modify the a-r-r after cloning and before using it to bootstrap | 14:57 |
| odyssey4me | at least from ocata onwards we have the extended maintenance policy to keep the branches alive longer | 14:58 |
| antonym | yeah, i thought about that too but it just seems a bit messier, especially if these type of repos break over time | 14:58 |
| odyssey4me | yeah, unfortunately we can't bring those branches back as far as I know - so perhaps for those which are eol the mess is necessary | 15:00 |
| pabelanger | logan-: okay, PR merged, post job running | 15:01 |
| pabelanger | once this works, the next step is to drop the need for jumphost to br-mgmt network | 15:02 |
| pabelanger | and just use the public IPs directly | 15:02 |
| *** vnogin has quit IRC | 15:12 | |
| *** chkumar|ruck is now known as chkumar|away | 15:17 | |
| *** mgariepy has quit IRC | 15:18 | |
| *** mgariepy has joined #openstack-ansible | 15:19 | |
| Miouge- | I have a 10 nodes, single region OpenStack Fuel deployment (Mitaka I think) that I would love to convert to OSA. What is the general procedure? Any ideas where I can find info about this kind of thing? any talk on this out there? | 15:21 |
| *** ivve has joined #openstack-ansible | 15:21 | |
| odyssey4me | Miouge- some folks have done it, but I'm not aware of any notes anywhere | 15:21 |
| odyssey4me | I know mnaser and his crew have done it, as well as the city cloud folks. | 15:22 |
| *** vakuznet has quit IRC | 15:22 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Ensure that a consistent mirror is used for RDO https://review.openstack.org/620313 | 15:23 |
| mgariepy | i did switch from fuel to osa a while ago, but didn't keep the vms. | 15:25 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Ensure that a consistent mirror is used for RDO https://review.openstack.org/620629 | 15:26 |
| mgariepy | i kept the volumes and network (-ish) ovs > linux bridge, gre > vxlan | 15:26 |
| Miouge- | mgariepy: Ceph for Cinder? | 15:26 |
| mgariepy | yep | 15:26 |
| jrosser | odyssey4me: you going to edit that for rocky? | 15:26 |
| Miouge- | mgariepy: ceph-ansible “took over” the Ceph cluster? maybe that could be a first step | 15:27 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Ensure that a consistent mirror is used for RDO https://review.openstack.org/620629 | 15:27 |
| jrosser | cool | 15:28 |
| mgariepy | my ceph cluster is only 8 nodes, i didn't bother with it, only moved the mons from controllers to the ceph cluster nodes. | 15:28 |
| mgariepy | but yeah, probably move out the ceph cluster, from fuel, then rabbit and galera i guess. | 15:29 |
| mgariepy | maybe it has changed but in the old days, fuel wasn't really managing stuff, it was only a run once thing | 15:30 |
| *** jawad_ax_ has joined #openstack-ansible | 15:31 | |
| pabelanger | Hmm | 15:31 |
| pabelanger | shade is required for this module: https://object-storage-ca-ymq-1.vexxhost.net/v1/a0b4156a37f9453eb4ec7db5422272df/logs/b2/b2cc94a1fd0e2c169ece8d343d39c8f50d2c941f/post/packet-ci-cloud-deploy/12a69c7/job-output.html#l6017 | 15:32 |
| Miouge- | mgariepy: I don’t follow, did you make a cresh Ceph cluster or added nodes to the existing one and used the autoheal to migrate data? | 15:32 |
| Miouge- | s/cresh/fresh | 15:32 |
| mgariepy | nop, i just moved the ceph mons on other nodes. | 15:32 |
| mgariepy | the mons where on the controllers. | 15:33 |
| *** jawad_axd has quit IRC | 15:33 | |
| *** jawad_a__ has joined #openstack-ansible | 15:33 | |
| mgariepy | and moved them on the ceph nodes. | 15:33 |
| mgariepy | do you need to keep the vms running ? | 15:34 |
| Miouge- | I can reboot them or turn off for a week end or so | 15:35 |
| *** gkadam has quit IRC | 15:35 | |
| pabelanger | logan-: odyssey4me: looks like a bug with https://review.openstack.org/568146/ seems the utility container doesn't have shade installed | 15:36 |
| pabelanger | see log above | 15:36 |
| *** jawad_ax_ has quit IRC | 15:36 | |
| logan- | ahh makes sense | 15:36 |
| mgariepy | what os is on the fuel nodes ? | 15:37 |
| mgariepy | is it xenial ? | 15:37 |
| pabelanger | not sure how best to solve this, delegate shade dependencies to op openstack_service_setup_host too? | 15:37 |
| Miouge- | mgariepy: Fuel node is Centos VM but all OpenStack nodes are Ubuntu 14.04 I think | 15:38 |
| pabelanger | how is shade installed on to deployment host? I am guessing from pip? | 15:41 |
| pabelanger | if so, I think I can just add shade to utility_pip_packages | 15:42 |
| *** udesale has quit IRC | 15:43 | |
| *** udesale has joined #openstack-ansible | 15:43 | |
| odyssey4me | pabelanger it's already there, but in a venv - just chaging that var won't help because it assumes whatever the service setup host is has the right packages in the system | 15:45 |
| odyssey4me | we may have to try and work out something nicer to make that easier | 15:45 |
| odyssey4me | installing python packages into the system is something we're actively trying to avoid | 15:45 |
| odyssey4me | (via pip I mean) | 15:45 |
| pabelanger | odyssey4me: okay, will defer to you on how best to fix then :) | 15:46 |
| mgariepy | Miouge-, start by moving ceph-mon and rgw somewhere else | 15:46 |
| mgariepy | Miouge-, watch for : https://bugs.launchpad.net/nova/+bug/1452641 | 15:46 |
| openstack | Launchpad bug 1452641 in nova (Ubuntu) "Static Ceph mon IP addresses in connection_info can prevent VM startup" [Medium,In progress] - Assigned to Corey Bryant (corey.bryant) | 15:46 |
| pabelanger | odyssey4me: when you say system, you mean outside of the venv? | 15:47 |
| odyssey4me | pabelanger we may have to include shade (rocky)/openstacksdk (stein) in the utility_pip_packages for just this purpose | 15:48 |
| mnaser | odyssey4me: i had a huge upgrade fire yesterday caused by the galera upgrade playbooks | 15:48 |
| mnaser | we somehow restart all of them at some point? one by one? without ever checking that they have fully sync'd up.. | 15:49 |
| odyssey4me | pabelanger yes, unfortunately I don't see an easy way around that - thankfully it's a small number of packages | 15:50 |
| mnaser | and also, somehow, haproxy was routing things to one of the galera servers that was *not* ready to serve traffic yet | 15:50 |
| mnaser | resulting in a bunch of WSREP not responding | 15:50 |
| pabelanger | odyssey4me: if it was pulled in via apt / yum, is that the better way to deal with the package? | 15:50 |
| odyssey4me | mnaser run-upgrade does a serialised upgrade IIRC - and the playbooks to deploy do so too... I think we should actually do it in the normal deploy playbook, not have a different process for both | 15:50 |
| mnaser | odyssey4me: but we call "Restart mysql (All)" for some reason | 15:51 |
| mnaser | at least the log show that | 15:51 |
| odyssey4me | pabelanger yeah, that's a decent option | 15:51 |
| odyssey4me | mnaser it's a bit of a context switch for me right now, and I don't remember the details off-hand | 15:51 |
| mnaser | sorry | 15:52 |
| mnaser | i'll dig into it more | 15:52 |
| odyssey4me | pabelanger let me take a peek through what's going on there to see what'll work best | 15:52 |
| mnaser | but yeah, it was pretty tragic for the cloud, i will dig more | 15:52 |
| pabelanger | odyssey4me: okay, thanks. will hold off on utility_pip_packages for the moment | 15:52 |
| mnaser | also to figure out why haproxy was sending traffic to a non-ready server | 15:52 |
| *** cshen has quit IRC | 15:53 | |
| mnaser | ton of: 2018-11-26 21:40:28.906 109358 ERROR sqlalchemy.pool.QueuePool InternalError: (1047, u'WSREP has not yet prepared node for application use') | 15:53 |
| odyssey4me | pabelanger ok, as I thought - https://github.com/openstack/openstack-ansible/blob/master/playbooks/utility-install.yml#L103-L112 installs utility_pip_packages into the venv | 15:53 |
| pabelanger | odyssey4me: and that is bad, in this case? | 15:54 |
| *** cshen has joined #openstack-ansible | 15:54 | |
| *** vakuznet has joined #openstack-ansible | 15:54 | |
| odyssey4me | pabelanger no, it means that we'll ned to try and use a distro package - so add shade to https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/utility_all.yml#L29-L32 | 15:54 |
| odyssey4me | except that the package name is different across distributions, so this may be a little fun :) | 15:54 |
| pabelanger | odyssey4me: yah, okay for now I can add it to my user_variables file and see if that fixes the issue | 15:55 |
| odyssey4me | an option is to allow the resulting python executable in each role to be overridden, but default back to the system | 15:56 |
| pabelanger | odyssey4me: yah, that is what I do with my zuul / nodepool roles | 15:56 |
| pabelanger | I think that is a good option also | 15:56 |
| odyssey4me | so in https://github.com/openstack/openstack-ansible-os_keystone/blob/master/tasks/keystone_service_update.yml#L30 change ansible_python['executable'] to be something like keystone_service_setup_host_python_executable and have that be set in defaults as keystone_service_setup_host_python_executable: "{{ openstack_service_setup_host_python_executable ¬ default(ansible_python['executable']) }}" | 15:58 |
| Miouge- | mgariepy: I am thinking to do: fuel Mitaka => OSA Mitaka => OSA Newton => ... => OSA Rocky | 15:58 |
| odyssey4me | then in the utility group_vars we can set that var by default so that everything works like magic | 15:59 |
| odyssey4me | assuming that works, I think that's the best option | 15:59 |
| pabelanger | odyssey4me: actually, we already install some openstack client libraries using utility_distro_openstack_clients_packages maybe just see about adding shade there by default too? | 15:59 |
| *** weezS has joined #openstack-ansible | 15:59 | |
| pabelanger | https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/utility_all.yml#L34 | 15:59 |
| odyssey4me | pabelanger that list is for distro (not source) builds | 16:00 |
| mgariepy | Miouge-, so, fuel Mitaka => OSA Mitaka > osa newton 14.04 > osa newton 16.04 > osa o p q r < osa R 1804 | 16:00 |
| mgariepy | Miouge-, lots of works ahead :) | 16:01 |
| pabelanger | odyssey4me: okay and we prefer source build for shade? | 16:01 |
| odyssey4me | pabelanger distro/source is a setting for the build as a whole | 16:02 |
| odyssey4me | the distro install of openstack is new, and only really fully working for suse right now | 16:02 |
| pabelanger | ah, okay | 16:02 |
| odyssey4me | I'll do a couple of patches now which you can try out. | 16:02 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Enable overriding the service setup host python interpreter https://review.openstack.org/620644 | 16:09 |
| odyssey4me | bother, now it gets complicated :/ | 16:20 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the utility container service setup interprter automatically https://review.openstack.org/620651 | 16:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the utility container service setup interpreter automatically https://review.openstack.org/620651 | 16:26 |
| *** gyee has joined #openstack-ansible | 16:27 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the utility container service setup interpreter automatically https://review.openstack.org/620651 | 16:27 |
| odyssey4me | pabelanger something like that - the patch done for keystone would have to be propogated to other roles too | 16:27 |
| *** hamzaachi has quit IRC | 16:27 | |
| Miouge- | mgariepy: yes, it won’t happen overnight. I need to do this in stages | 16:29 |
| *** udesale has quit IRC | 16:34 | |
| *** jawad_a__ has quit IRC | 16:35 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Update all SHAs for 16.0.23 https://review.openstack.org/618840 | 16:35 |
| *** jawad_axd has joined #openstack-ansible | 16:36 | |
| mgariepy | Miouge-, when you get there, here are the notes for newton 1404 > 1604 upgrades: https://etherpad.openstack.org/p/osa-newton-xenial-upgrade | 16:37 |
| *** jawad_axd has quit IRC | 16:41 | |
| cloudnull | mornings all | 16:42 |
| Miouge- | mgariepy: Thanks! That’s the kind of thing I am looking for | 16:43 |
| *** hamerins has joined #openstack-ansible | 16:43 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: Remove non-existant 'functional' test https://review.openstack.org/620655 | 16:44 |
| *** cshen has quit IRC | 16:46 | |
| *** hamerins has quit IRC | 16:47 | |
| *** hamerins has joined #openstack-ansible | 16:48 | |
| *** fghaas has joined #openstack-ansible | 16:49 | |
| pabelanger | odyssey4me: oh noes: To utilize this module, the installed version ofthe shade library MUST be >=1.8.0 | 16:56 |
| pabelanger | looking to see what version was installed vi apip | 16:57 |
| pabelanger | that is from utility_distro_packages | 16:58 |
| *** lbragstad has quit IRC | 17:02 | |
| *** pcaruana has quit IRC | 17:03 | |
| *** lbragstad has joined #openstack-ansible | 17:05 | |
| *** KeithMnemonic has quit IRC | 17:14 | |
| *** strattao has quit IRC | 17:20 | |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible-ops master: Allow the option to specify OSA git repo URL https://review.openstack.org/620662 | 17:24 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 17:26 |
| *** hamerins has quit IRC | 17:28 | |
| *** jackivanov has quit IRC | 17:28 | |
| *** hamerins has joined #openstack-ansible | 17:29 | |
| odyssey4me | pabelanger yep, using a distro package sucks for this where we need more precise control - which is why the venv is better | 17:35 |
| * cloudnull is a general hater of disto packaging | 17:39 | |
| odyssey4me | pabelanger if you could try out those two patches to verify the concept then we can get that propogated | 17:40 |
| cloudnull | odyssey4me -re: https://review.openstack.org/#/c/607808 - im totally confused. | 17:40 |
| pabelanger | odyssey4me: sure, let me see how to apply them to my OSA deployment host | 17:40 |
| cloudnull | wouldnt that loop iterate over all nodes, which would include the first one ? | 17:41 |
| odyssey4me | cloudnull the issue I'm trying to get resolved is that the sshd must be enabled as soon as possible | 17:41 |
| odyssey4me | it's all nice and dandy if they all eventually get sshd enabled, but it's much better if it's enabled right from the get-go | 17:41 |
| cloudnull | so should I order the list? | 17:41 |
| odyssey4me | cloudnull nope, just make the task execute when the ansible_host is the first ansible_play_host | 17:42 |
| cloudnull | like sorted by playbook host and inventory hostname | 17:42 |
| odyssey4me | cloudnull nope, like https://github.com/openstack/openstack-ansible-os_keystone/blob/master/tasks/main.yml#L131 | 17:43 |
| cloudnull | then wouldn't it run on only the one host ? | 17:43 |
| odyssey4me | it only need to, because it loops through delegation to all the others, right? | 17:43 |
| odyssey4me | so run on host 1, but loop through the group and delegate to them all - exactly like https://github.com/openstack/openstack-ansible-os_keystone/commit/aefb58080337633e6d4c10bb2709c8d73ab9564d | 17:44 |
| cloudnull | it needs to run on all the hosts that are responsible for the ring. | 17:44 |
| odyssey4me | it's doing run_once now, right? | 17:45 |
| cloudnull | oh so you want run_once to be changed to that when ? | 17:45 |
| odyssey4me | yes, exactly | 17:45 |
| odyssey4me | so the host it runs on is deterministically the first in the play every time | 17:45 |
| cloudnull | OK. | 17:46 |
| * cloudnull was being stupid | 17:46 | |
| odyssey4me | doesn't that make sense? | 17:46 |
| *** fghaas has quit IRC | 17:48 | |
| *** hamzaachi has joined #openstack-ansible | 17:48 | |
| cloudnull | yea! | 17:50 |
| cloudnull | I was thinking we were debating on what needs to run and where. however we're just making sure where we start the iteration is deterministic | 17:51 |
| cloudnull | which makes total sense | 17:51 |
| * cloudnull updating that change | 17:51 | |
| jamesdenton | Can I get some eyes on https://review.openstack.org/#/c/620604/ when someone has a sec? Simple change. | 17:52 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_swift master: Add variable for the ssh service and ensure its enabled https://review.openstack.org/607808 | 17:52 |
| cloudnull | I went with the list intersection version of that as that should skip the task entirely should a limit be used that doesn't include the swift_proxy group | 17:56 |
| cloudnull | anyone around able to review https://review.openstack.org/#/c/620399 | 17:58 |
| *** hamerins has quit IRC | 18:01 | |
| *** hamerins has joined #openstack-ansible | 18:02 | |
| cloudnull | also https://review.openstack.org/#/c/620386/ would be a great cherr-pick to get into rocky | 18:03 |
| *** dcdamien has quit IRC | 18:05 | |
| cloudnull | also https://review.openstack.org/#/c/607814 - should i ever get cent to stop timing out | 18:05 |
| *** shardy has quit IRC | 18:11 | |
| *** hamerins has quit IRC | 18:12 | |
| *** hamerins has joined #openstack-ansible | 18:12 | |
| pabelanger | odyssey4me: comment on https://review.openstack.org/620651 | 18:22 |
| openstackgerrit | Merged openstack/ansible-role-systemd_service stable/rocky: Allow user/group to be overriden in unit section https://review.openstack.org/620386 | 18:25 |
| pabelanger | odyssey4me: and a 2nd comment, to fix syntax error | 18:25 |
| *** cshen has joined #openstack-ansible | 18:26 | |
| pabelanger | odyssey4me: those 2 changes seemed to have made os_keystone work | 18:29 |
| pabelanger | trying to see why glance failed now | 18:29 |
| *** vollman has joined #openstack-ansible | 18:31 | |
| *** chhagarw has quit IRC | 18:41 | |
| pabelanger | odyssey4me: ah, we'll need to same patch against all OS roles, since they too run on the utility container | 18:43 |
| *** electrofelix has quit IRC | 18:45 | |
| *** DanyC_ has joined #openstack-ansible | 18:52 | |
| *** DanyC has quit IRC | 18:56 | |
| *** DanyC_ has quit IRC | 18:57 | |
| cloudnull | hey all - https://bugs.launchpad.net/openstack-ansible/+bug/1805239 - regarding the issue, seems we have some gaps with rsysog and centos (RHEL). so the question is, do we fill in the logrotate gaps or do we refactor some of that code and try and not override the distro specific config files? | 18:58 |
| openstack | Launchpad bug 1805239 in openstack-ansible "missing logrotate of the syslog, daemon.log " [Medium,Confirmed] - Assigned to Vadim Kuznetsov (vakuznet) | 18:58 |
| *** hamerins has quit IRC | 19:00 | |
| *** hamerins has joined #openstack-ansible | 19:01 | |
| cloudnull | seems like an important issue to get resolved, especially in stable for folks running centos | 19:01 |
| *** preece has joined #openstack-ansible | 19:03 | |
| *** gisak has joined #openstack-ansible | 19:05 | |
| gisak | hello guys | 19:06 |
| gisak | is it possible to put glance on controller node? | 19:06 |
| gisak | if yes, I guess in this case I need a br-storage for controller as well, right ? | 19:07 |
| *** lbragstad has quit IRC | 19:08 | |
| jrosser | gisak: I think that is the default layout, you want br-storage on your controllers anyway I think | 19:10 |
| *** lbragstad has joined #openstack-ansible | 19:12 | |
| gisak | ok, thanks ) | 19:12 |
| *** lbragstad has quit IRC | 19:12 | |
| *** lbragstad has joined #openstack-ansible | 19:14 | |
| *** lbragstad has quit IRC | 19:14 | |
| *** lbragstad has joined #openstack-ansible | 19:16 | |
| *** lbragstad has quit IRC | 19:16 | |
| *** vakuznet has quit IRC | 19:18 | |
| jrosser | cloudnull: a quick skim of the rsyslog client role the code looks quite blunt | 19:21 |
| jrosser | At the very least it requires a ton of comments adding :/ | 19:21 |
| *** fghaas has joined #openstack-ansible | 19:23 | |
| *** radeks_ has joined #openstack-ansible | 19:25 | |
| *** gisak has quit IRC | 19:25 | |
| *** radeks has quit IRC | 19:26 | |
| *** jawad_axd has joined #openstack-ansible | 19:34 | |
| *** vakuznet has joined #openstack-ansible | 19:35 | |
| *** strattao has joined #openstack-ansible | 19:37 | |
| cloudnull | jrosser totally agree. its largely been unchanged since it was originally created. | 19:41 |
| jrosser | I can pretty much not understand the intent there without studying it much harder - perhaps that points to a refactor? | 19:42 |
| *** cshen has quit IRC | 19:45 | |
| *** mrhillsman is now known as mrhillsman|lunch | 19:46 | |
| *** strattao has quit IRC | 19:47 | |
| cloudnull | ++ its due for sure. | 19:47 |
| cloudnull | that said, theres been work to decommission our use of rsyslog in favor of journald | 19:48 |
| cloudnull | so it may just be something of benefit for stable | 19:48 |
| cloudnull | which then begs the question, is it worth doing? | 19:49 |
| *** jawad_axd has quit IRC | 19:49 | |
| *** jawad_axd has joined #openstack-ansible | 19:52 | |
| *** cshen has joined #openstack-ansible | 19:52 | |
| *** radeks has joined #openstack-ansible | 19:53 | |
| *** radeks_ has quit IRC | 19:55 | |
| *** jawad_axd has quit IRC | 20:15 | |
| *** cshen has quit IRC | 20:16 | |
| *** cshen has joined #openstack-ansible | 20:28 | |
| *** lbragstad has joined #openstack-ansible | 20:30 | |
| *** hamzaachi has quit IRC | 20:31 | |
| *** mrhillsman|lunch is now known as mrhillsman | 20:35 | |
| *** openstackgerrit has quit IRC | 20:36 | |
| *** dcdamien has joined #openstack-ansible | 20:39 | |
| *** ahosam has joined #openstack-ansible | 20:48 | |
| *** openstackgerrit has joined #openstack-ansible | 21:13 | |
| openstackgerrit | Merged openstack/openstack-ansible-ceph_client master: SUSE: Fix ceph packages for Leap 15 https://review.openstack.org/615535 | 21:13 |
| *** fghaas has quit IRC | 21:16 | |
| *** klamath has quit IRC | 21:26 | |
| *** klamath has joined #openstack-ansible | 21:26 | |
| *** vollman has quit IRC | 21:27 | |
| *** radeks has quit IRC | 21:32 | |
| *** lbragstad has quit IRC | 21:37 | |
| *** lbragstad has joined #openstack-ansible | 21:37 | |
| *** ansmith has quit IRC | 21:59 | |
| jamesdenton | noonedeadpunk Just FYI (open resolver issue): https://review.openstack.org/#/c/333829/ | 22:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Include Swift AUTH_%(tenant_id)s suffix in rgw Keystone endpoint https://review.openstack.org/614194 | 22:27 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 22:29 |
| *** strattao has joined #openstack-ansible | 22:37 | |
| *** strattao has quit IRC | 22:39 | |
| *** ahosam has quit IRC | 22:51 | |
| *** ktims has joined #openstack-ansible | 23:23 | |
| ktims | hi folks i wonder if someone can help me with openstack+ceph deployment with openstack-ansible. currently i'm stuck at lxc_hosts cache preparation. it appears that the container is getting neither ca-certificates nor the APT key for the ceph repository and i can't find any ansible variables that apply to this. | 23:25 |
| *** cshen has quit IRC | 23:28 | |
| *** xjra has joined #openstack-ansible | 23:41 | |
| xjra | We find ourselves with a requirement to run multiple Horizon instances for clients in different security contexts; is it possible to have OSA deploy multiple Horizons with different configurations? | 23:43 |
| xjra | http://paste.openstack.org/show/736340/ | 23:46 |
| xjra | Oops, unrelated paste | 23:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!