| *** samcat116 has joined #openstack-kolla | 00:15 | |
| *** samcat116 has quit IRC | 00:30 | |
| *** dking has quit IRC | 00:46 | |
| *** LinPeiWen has quit IRC | 01:43 | |
| *** k-s-dean has quit IRC | 01:48 | |
| *** k-s-dean has joined #openstack-kolla | 01:48 | |
| *** zzzeek has quit IRC | 02:21 | |
| *** zzzeek has joined #openstack-kolla | 02:23 | |
| openstackgerrit | wu.chunyang proposed openstack/kolla-ansible master: [doc] introduce octavia tenant management network https://review.opendev.org/c/openstack/kolla-ansible/+/783876 | 02:33 |
|---|---|---|
| *** LinPeiWen has joined #openstack-kolla | 02:50 | |
| *** LinPeiWen has quit IRC | 02:56 | |
| *** vishalmanchanda has joined #openstack-kolla | 03:17 | |
| *** LinPeiWen has joined #openstack-kolla | 04:08 | |
| *** zzzeek has quit IRC | 04:59 | |
| *** zzzeek has joined #openstack-kolla | 05:00 | |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/victoria: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785922 | 05:10 |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/ussuri: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785923 | 05:10 |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/train: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785924 | 05:10 |
| *** Luzi has joined #openstack-kolla | 05:50 | |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/train: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785924 | 06:00 |
| *** ivan_lin has quit IRC | 06:01 | |
| *** ivan_lin has joined #openstack-kolla | 06:02 | |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/train: Fix installation errors with Python 2 https://review.opendev.org/c/openstack/kayobe/+/785624 | 06:03 |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/train: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785924 | 06:03 |
| *** kevko has joined #openstack-kolla | 06:05 | |
| *** ivan_lin has quit IRC | 06:40 | |
| *** ivan_lin has joined #openstack-kolla | 06:41 | |
| *** andrewbonney has joined #openstack-kolla | 07:14 | |
| *** bengates has joined #openstack-kolla | 07:28 | |
| *** bengates has quit IRC | 07:29 | |
| *** bengates has joined #openstack-kolla | 07:29 | |
| *** luksky has joined #openstack-kolla | 07:42 | |
| *** rpittau|afk is now known as rpittau | 07:43 | |
| *** dougsz has joined #openstack-kolla | 07:45 | |
| *** bengates_ has joined #openstack-kolla | 07:46 | |
| *** bengates has quit IRC | 07:50 | |
| *** ivan_lin has quit IRC | 07:51 | |
| *** ivan_lin has joined #openstack-kolla | 07:52 | |
| *** e0ne has joined #openstack-kolla | 07:54 | |
| hrw | morning | 07:57 |
| hrw | mnasiadka: any progress with mariadb? | 07:59 |
| mnasiadka | hrw: a bit, I know what is the problem, now trying to fix it ;) | 08:00 |
| hrw | cool, thanks | 08:01 |
| hrw | basically every distro we support has some issue with binary images | 08:01 |
| hrw | Debian needs update to newer release (and mariadb fix) | 08:02 |
| mnasiadka | well, maybe we need to dump binary :D | 08:02 |
| hrw | Ubuntu needs update to newer UCA (and upgrade CI job check) | 08:02 |
| hrw | CentOS looks stable but there is work behind scenes with adding needed repos to 8-stream | 08:03 |
| hrw | mnasiadka: next week we will discuss it ;D | 08:03 |
| *** gfidente|afk is now known as gfidente | 08:06 | |
| *** dougsz has quit IRC | 08:43 | |
| *** hrw has quit IRC | 08:47 | |
| *** Luzi has quit IRC | 08:47 | |
| *** Luzi has joined #openstack-kolla | 08:48 | |
| *** wuchunyang has joined #openstack-kolla | 08:48 | |
| *** hrw has joined #openstack-kolla | 08:49 | |
| *** k_mouza has joined #openstack-kolla | 09:01 | |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe master: Exclude shared directories when creating environments https://review.opendev.org/c/openstack/kayobe/+/785673 | 09:06 |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe master: Add release note for multiple environments https://review.opendev.org/c/openstack/kayobe/+/786023 | 09:06 |
| hrw | https://review.opendev.org/c/openstack/kolla/+/785871 any brave core? | 09:07 |
| kevko | hrw: checking PTG general discussion topics, and from you there is a comment ..deprecate binary images .. does it mean that you want to deprecate and remove in future ? | 09:19 |
| kevko | could I ask why ? :/ | 09:19 |
| kevko | btw, do someone have some playbook for irc bouncer installation and setup ? :D i'm lazy to install and setup manually | 09:21 |
| hrw | kevko: can not help with irc bouncer - my machine works 24/7 | 09:21 |
| kevko | hrw: and images question ? :P | 09:21 |
| hrw | kevko: using binary packages is always adding extra near-to-release work | 09:22 |
| hrw | kevko: for example now none of supported distros has ready to use binary stuff | 09:22 |
| hrw | CentOS/RDO teams work on preparing all repos, for Debian we need to bump to Bullseye, Ubuntu UCA update needs someone checking why V->W upgrade failed on CI | 09:23 |
| kevko | hrw: understand your point, but it's really usefull for users (as I am also) who build images locally on theirs side ..and are patching packages theirselves .. just adding repo for build | 09:23 |
| hrw | kevko: similar with source images - you can tell kolla 'use those git repos instead of upstream' | 09:23 |
| kevko | hrw: yeah, but binary has dependencies, recommends, suggest (debian) and it's cleaner, nicer .. i don't trust pip :( | 09:25 |
| kevko | hrw: have to go for 25 minutes ..but stay here .. i really want to discuss :P | 09:26 |
| * kevko is going afk for 30 minutes | 09:26 | |
| mnasiadka | kevko: you will get the opportunity to discuss on the PTG, that's why it's a PTG topic ;) | 09:26 |
| hrw | kevko: in perfect world we would have images ready to release in any moment. Which is possible with source ones. | 09:27 |
| hrw | kevko: with binary we need to wait for distros to keep up. Here Debian was usually worst (as repos were available after OpenStack release), Ubuntu was better as UCA repos were available early but here getting issues sorted out was usually a problem | 09:28 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Disable amphora in OVN jobs https://review.opendev.org/c/openstack/kolla-ansible/+/783669 | 09:31 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla master: move to Debian 'bullseye' https://review.opendev.org/c/openstack/kolla/+/772479 | 09:42 |
| hrw | mnasiadka: thanks! | 09:44 |
| mnasiadka | ok, should pass now, but this expect crap is something that made my day | 09:44 |
| mnasiadka | if it won't pass, we'll need to fix wsrep notify script in k-a, but let's see | 09:44 |
| *** rpittau is now known as rpittau|bbl | 09:54 | |
| openstackgerrit | Pierre Riteau proposed openstack/kayobe stable/stein: Fix installation errors with Python 2 https://review.opendev.org/c/openstack/kayobe/+/785869 | 10:07 |
| openstackgerrit | Merged openstack/kayobe stable/ussuri: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785923 | 10:08 |
| openstackgerrit | Merged openstack/kolla-ansible master: octavia: Add support for disabling amphora provider https://review.opendev.org/c/openstack/kolla-ansible/+/783699 | 10:12 |
| *** wuchunyang has quit IRC | 10:17 | |
| *** fuhrmannb has joined #openstack-kolla | 10:19 | |
| kevko | hrw: let's assume that I have found that designate image is corrupted because of python-dnspython library in version X.Y installed as dependency of designate package from APT, there is already fix in included in version A.B ...but upper-constraints is set to X.Y ...so what I need to do is fix python-dnpython library ..but only with patch (because full replacement of dnspython by A.B version breaks designate) ...so I build new package ..patch it | 10:28 |
| kevko | with fix ..and build image .. do I have this option in source setup ? I mean ..can I replace installation of dependency - small library python-dnspython with my own git-repo in specific branch with fix included ? | 10:28 |
| hrw | kevko: you clone openstack/requirements, bump upper-constraints.txt and send patch | 10:32 |
| kevko | hrw: but I don't know if dnspython is working in 1.8 or 1.9 or 1.9.999999 | 10:33 |
| kevko | hrw: i just need patch 1.7 with one line in a code | 10:33 |
| hrw | where from you get that one line? | 10:33 |
| kevko | hrw: from upstream | 10:35 |
| kevko | hrw: and have to say this is real example I've solved few days ago in debian | 10:35 |
| hrw | kevko: so from 1.7.1 or 1.8 version, right? | 10:35 |
| hrw | kevko: if bug is debian binary specific then it can probably just stay in Debian. if it is not then requirements should get bump | 10:36 |
| kevko | hrw: bugreport debian -> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=986645 , fixed version https://release.debian.org/proposed-updates/buster_diffs/dnspython_1.16.0-1+deb10u1.debdiff which will be included in next point release of debian ... and already backported in backports debian openstack repo | 10:36 |
| openstack | Debian bug 986645 in python3-dnspython "python3-dnspython: breaks designate secondary zone create" [Normal,Fixed] | 10:36 |
| kevko | hrw: it's not specific for debian ..it is real bug in dnspython library | 10:37 |
| kevko | hrw: (1.7 was example ..real version is 1.16.0) | 10:37 |
| hrw | kevko: so, it is fixed in 1.16.1? | 10:38 |
| kevko | hrw: it's fixed in dnspython (1.16.0-1+deb10u1) buster; urgency=medium | 10:39 |
| hrw | kevko: this is not #debian-kolla | 10:39 |
| hrw | kevko: so, it is fixed in 1.16.1 upstream? | 10:39 |
| kevko | hrw: no, it's not ..that's a reasn | 10:39 |
| hrw | kevko: ok. have anyone suggested 1.16.1 release? | 10:40 |
| kevko | hrw: it's fixed firstly in 2.0 | 10:40 |
| hrw | which openstack lists as 'do not use' | 10:41 |
| hrw | kevko: so, it is fixed in 1.16.1 upstream?global-requirements.txt:dnspython!=2.0.0 # http://www.dnspython.org/LICENSE | 10:41 |
| hrw | ops | 10:41 |
| hrw | global-requirements.txt:dnspython!=2.0.0 # http://www.dnspython.org/LICENSE | 10:41 |
| kevko | hrw: but 2.0 is breaking designate code (and maybe others) ..but as you can see ..fix is trivial ..and backport/patch is very easy >> https://github.com/nrhall/dnspython/commit/9403c1bdbdf562cb01ee492b207b1f26191976ca ...so my question is still same ..am I able to tell kolla ..use my git branch in my git repo ..where i have this simple fix ? | 10:42 |
| hrw | https://review.opendev.org/c/openstack/requirements/+/761432 | 10:42 |
| kevko | for this trivial fix ? | 10:42 |
| hrw | kevko: dnspython gets installed via pip, right? | 10:43 |
| kevko | hrw: well, in source yes, in binary ...it's installed as dependency from binary APT repo .. | 10:43 |
| kevko | hrw: in APT repo I have option to fix this one line and release as version with new revision VERSION-REVISION | 10:43 |
| kevko | in changelog i add note that I fixed this with oneline patch | 10:44 |
| kevko | so, if there will be mechanism to override everything what is kolla installing via pip ..i have no problem | 10:44 |
| hrw | I know | 10:45 |
| hrw | https://github.com/eventlet/eventlet/issues/619 is fun read | 10:46 |
| hrw | mnasiadka: passed that mariadb issue, failed step later ;( | 10:48 |
| mnasiadka | hrw: ok, I'll do a full deployment locally then - just did mariadb :) | 10:49 |
| hrw | "msg": "Timeout when waiting for search string MariaDB in 192.0.2.1:3306", | 10:49 |
| kevko | what problem do you have about mariadb ? | 10:49 |
| hrw | https://6ebeb1abc98ef928169c-71736a9a588da4a5a421519de66e6230.ssl.cf2.rackcdn.com/772479/18/check/kolla-ansible-debian-source/c79fec5/primary/logs/ansible/deploy | 10:50 |
| kevko | hmm, maybe it is breaking my merged patch :D | 10:51 |
| kevko | let me help you | 10:51 |
| hrw | kevko: thanks | 10:51 |
| hrw | kevko: https://review.opendev.org/c/openstack/kolla/+/772479 change | 10:51 |
| kevko | hrw: could you send me url with logs dir ? :P | 10:52 |
| hrw | I will add one more entry to deprecate list | 10:52 |
| mnasiadka | hrw: ah, we need a patch in k-a | 10:52 |
| kevko | not only one above | 10:52 |
| mnasiadka | because wsrep-notify.sh does not have a timeout at all | 10:52 |
| hrw | kevko: https://6ebeb1abc98ef928169c-71736a9a588da4a5a421519de66e6230.ssl.cf2.rackcdn.com/772479/18/check/kolla-ansible-debian-source/c79fec5/ is enough? | 10:53 |
| mnasiadka | and it's stuck on some reason connecting to the db and doesn't allow for any other connections | 10:53 |
| hrw | mnasiadka: ;) | 10:53 |
| mnasiadka | I'm wondering if we need wsrep-notify.sh at all | 10:53 |
| kevko | thanks | 10:53 |
| mnasiadka | since mariadb-clustercheck is checking cluster state and adding nodes to haproxy when those are in good state | 10:54 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: move to Debian 'bullseye' https://review.opendev.org/c/openstack/kolla/+/772479 | 10:56 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: Debian/Ubuntu: handle APT keys in proper way https://review.opendev.org/c/openstack/kolla/+/784923 | 10:56 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: WIP: Add timeout to mysql cmd in wsrep notify https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 10:57 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla master: move to Debian 'bullseye' https://review.opendev.org/c/openstack/kolla/+/772479 | 10:58 |
| mnasiadka | ok, adding depends-on - this helped on my local env | 10:58 |
| kevko | give me 10 15 minutes | 10:59 |
| mnasiadka | kevko: take all the time you need ;) | 11:01 |
| hrw | do we have openstack release name as a template variable? | 11:06 |
| openstackgerrit | Merged openstack/kayobe stable/victoria: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785922 | 11:07 |
| kevko | hrw: yes | 11:07 |
| hrw | ok, openstack_release | 11:07 |
| kevko | :D | 11:07 |
| * kevko is going to ciggarette | 11:08 | |
| hrw | INFO:kolla.common.utils.base:'master': not a valid cloud-archive name. | 11:09 |
| mnasiadka | well, about that - kolla by default builds images with a version tag (like 11.1.0) and kolla-ansible uses openstack_release (which is for example master or wallaby) - wouldn't we want to unify those? ;) | 11:09 |
| kevko | if yes, i will vote for simple release name, not digit version | 11:10 |
| kevko | hrw: hmm, I have some dev env in k-a master (outdated 1 month ) ..so I downloaded fresh master ..and replace ..and run deploy ... do you know why i am getting this ? :D | 11:12 |
| kevko | http://paste.openstack.org/show/804431/ | 11:12 |
| hrw | http: server gave HTTP response to HTTPS client | 11:13 |
| mnasiadka | insecure-registry not configured in daemon.json? | 11:14 |
| kevko | oh, my fault ..i commented custom repo where I have master images built | 11:15 |
| kevko | in globals ..so uncommented :P | 11:15 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: centos: enable repos where needed https://review.opendev.org/c/openstack/kolla/+/785871 | 11:21 |
| hrw | yoctozepto: thanks for comments | 11:21 |
| hrw | git-review lacks '--refresh-from-gerrit' argument | 11:24 |
| hrw | 2aad1eb19 (HEAD -> review/marcin_juszkiewicz/prepare-for-kolla-wallaby-release) base: do not overwrite APT sources.list | 11:30 |
| hrw | 66c25b346 Debian/Ubuntu: handle APT keys in proper way | 11:30 |
| hrw | 0db26c864 move to Debian 'bullseye' | 11:30 |
| *** k_mouza has quit IRC | 11:37 | |
| hrw | will wait with it | 11:39 |
| kevko | hrw: if i will find something ..can I just upload to that review ? | 11:43 |
| *** kevko_ has joined #openstack-kolla | 11:47 | |
| *** kevko has quit IRC | 11:49 | |
| hrw | kevko_: once patch landed in gerrit it is project's patch so anyone can update it | 11:51 |
| kevko_ | ok | 11:52 |
| *** k_mouza has joined #openstack-kolla | 11:52 | |
| hrw | kevko_: and gerrit tracks changes between revisions so it is easy to track | 11:54 |
| kevko_ | i know :) | 11:54 |
| kevko_ | btw, when I buy your fridge magnet ..i will ask for address where to send | 11:54 |
| openstackgerrit | Will Szumski proposed openstack/kayobe-config-dev master: Adds ntp group https://review.opendev.org/c/openstack/kayobe-config-dev/+/786040 | 11:55 |
| hrw | kevko_: ok | 11:56 |
| hrw | kevko_: and thanks ;) | 11:56 |
| oklhost | hi | 11:57 |
| *** dougsz has joined #openstack-kolla | 12:00 | |
| *** dougsz has quit IRC | 12:00 | |
| oklhost | mgoddard: FYI: my colleague submitted https://bugs.launchpad.net/designate/+bug/1923589 maybe thats a case in https://bugs.launchpad.net/designate/+bug/1897936? | 12:00 |
| openstack | Launchpad bug 1923589 in Designate "pool update fails - want to set zone master for primary zones" [Undecided,New] | 12:00 |
| openstack | Launchpad bug 1897936 in Designate "Pool update fails when zones exist" [Undecided,Fix released] - Assigned to Mark Goddard (mgoddard) | 12:00 |
| *** dougsz has joined #openstack-kolla | 12:00 | |
| mnasiadka | hrw: seems we got to horizon now, but it's returning 404 :) | 12:01 |
| hrw | mnasiadka: improvement ;) | 12:01 |
| mnasiadka | hrw: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_3cd/772479/20/check/kolla-ansible-debian-source/3cdbd6c/primary/logs/kolla/horizon/horizon.txt | 12:01 |
| hrw | oklhost: mgoddard is on PTO this week | 12:01 |
| hrw | mnasiadka: hm. 3.7... | 12:02 |
| openstackgerrit | Will Szumski proposed openstack/kayobe master: Adds support for configuring chrony https://review.opendev.org/c/openstack/kayobe/+/784460 | 12:02 |
| mnasiadka | hrw: we had the same problem last cycle with Ubuntu ;) | 12:02 |
| oklhost | hrw: thanks! :-) | 12:02 |
| mnasiadka | hrw: distro_python_version points to 3.7 | 12:04 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: storm: no need to check is it Python 3 https://review.opendev.org/c/openstack/kolla/+/786042 | 12:05 |
| hrw | elif self.base in ['debian']: | 12:06 |
| hrw | self.conf.distro_python_version = "3.9" | 12:06 |
| hrw | hm. | 12:06 |
| hrw | ah. image built by k-a | 12:07 |
| hrw | mnasiadka: https://paste.centos.org/view/487aaafa | 12:08 |
| mnasiadka | hrw: yeah, but do we have python 3.9 on Debian Buster? | 12:09 |
| hrw | nope | 12:09 |
| hrw | mnasiadka: you suggest that we need bullseye node too? | 12:10 |
| openstackgerrit | Will Szumski proposed openstack/kayobe master: Adds support for configuring chrony https://review.opendev.org/c/openstack/kayobe/+/784460 | 12:11 |
| mnasiadka | hrw: well, we have the ugliest dependency you can imagine ;) | 12:12 |
| hrw | mnasiadka: you mean? | 12:12 |
| mnasiadka | hrw: Bullseye change needs k-a to use python 3.9 path for wsgi, but bumping up python to 3.9 in k-a will fail on CI jobs :) | 12:14 |
| mnasiadka | well, debian-source is non voting, so we might survive | 12:15 |
| mnasiadka | but it's a bit of a breaking change | 12:15 |
| mnasiadka | hrw: well actually we can do this: https://review.opendev.org/c/openstack/kolla/+/738994/19/.zuul.d/ubuntu.yaml | 12:17 |
| mnasiadka | and then fix it in k-a, and then remove those additional variables in kolla | 12:18 |
| hrw | mnasiadka: and in meantime infra will finish reviewing bullseye nodes/images patches | 12:19 |
| *** e0ne has quit IRC | 12:19 | |
| mnasiadka | hrw: will you update the change or should I? | 12:20 |
| hrw | mnasiadka: go for it | 12:20 |
| hrw | I am going shopping - my fridge has only penguins inside | 12:20 |
| mnasiadka | Ok, will do | 12:22 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 12:26 |
| kevko_ | hrw: are u sure that dependent review is not breaking tests ? | 12:27 |
| hrw | kevko_: which case? | 12:28 |
| hrw | kevko_: sorry, I have too many patches in queue | 12:29 |
| kevko_ | https://6ebeb1abc98ef928169c-71736a9a588da4a5a421519de66e6230.ssl.cf2.rackcdn.com/772479/18/check/kolla-ansible-debian-source/c79fec5/primary/logs/ansible/deploy | 12:29 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla master: move to Debian 'bullseye' https://review.opendev.org/c/openstack/kolla/+/772479 | 12:30 |
| hrw | kevko_: mnasiadka works on it. sync with him please | 12:31 |
| * hrw -> shopiing | 12:31 | |
| *** rpittau|bbl is now known as rpittau | 12:31 | |
| mnasiadka | kevko_: depends-on was added after this run | 12:31 |
| mnasiadka | to make it work, not the other way around ;) | 12:32 |
| kevko_ | mnasiadka: so, probably you see this https://6ebeb1abc98ef928169c-71736a9a588da4a5a421519de66e6230.ssl.cf2.rackcdn.com/772479/18/check/kolla-ansible-debian-source/c79fec5/primary/logs/kolla/mariadb/mariadb.txt | 12:32 |
| mnasiadka | yes, operation not permitted is because MariaDB 10.5+ changed mysql.user to a view | 12:32 |
| mnasiadka | and we're changing haproxy user in a hacky way | 12:33 |
| mnasiadka | as I already mentioned, I'm not sure we need that script anyway | 12:33 |
| mnasiadka | but it's not a good time to get rid of it :) | 12:33 |
| kevko_ | mnasiadka: hmm, are u sure ? this script is used in clustercheck ..based on that script clustercheck is reporting which node is OK and haproxy is enabling/disabling mariadb backend nodes | 12:36 |
| mnasiadka | kevko_: no, wsrep notify is not used in clustercheck | 12:36 |
| mnasiadka | the user it alters is, that's correct ;) | 12:37 |
| mnasiadka | still clustercheck would not ,,add'' a non-synced node to haproxy pool if it wasn't in correct state | 12:37 |
| mnasiadka | so it doesn't really matter if we rename the haproxy user to haproxy_blocked in some cases | 12:38 |
| kevko_ | mnasiadka: it is -> https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/mariadb/templates/wsrep-notify.sh.j2 | 12:39 |
| kevko_ | https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/mariadb/templates/mariadb-clustercheck.json.j2 | 12:39 |
| mnasiadka | this is wsrep notify, not clustercheck | 12:39 |
| mnasiadka | https://github.com/openstack/kolla-ansible/blob/09fcfe8cd2a88a8a425d3f975fa0347b7701347c/ansible/roles/mariadb/templates/xinetd-clustercheck.j2#L12 | 12:40 |
| mnasiadka | this is what xinetd in clustercheck is running | 12:40 |
| kevko_ | mnasiadka: oh, yes | 12:40 |
| kevko_ | mnasiadka: it just confused me because I saw it there but it's not used | 12:42 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: CI: cephadm: wait after applying cluster spec https://review.opendev.org/c/openstack/kolla-ansible/+/786043 | 12:53 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 12:55 |
| mnasiadka | bloody semicolons | 12:55 |
| mnasiadka | ; | 12:55 |
| mnasiadka | ;) | 12:55 |
| *** wuchunyang has joined #openstack-kolla | 12:58 | |
| yoctozepto | hrw: I like penguins | 13:01 |
| *** kevko__ has joined #openstack-kolla | 13:05 | |
| *** wuchunyang has quit IRC | 13:06 | |
| *** kevko has joined #openstack-kolla | 13:07 | |
| *** kevko_ has quit IRC | 13:07 | |
| kevko | mnasiadka: hmm, ok, now I understand why wsrep_notify can be removed ..because haproxy is checking via clustercheck | 13:07 |
| mnasiadka | kevko: yup, once it was haproxy doing the check - it made sense, but clustercheck is doing it's job probably better | 13:08 |
| mnasiadka | I'll raise a change to remove it in X | 13:08 |
| kevko | mnasiadka: agree, moreover, I proposed patch for proxysql (maybe you've seen already) ..where proxysql is doing check of mariadb state itself | 13:09 |
| mnasiadka | kevko: yeah, I've seen, I'm still not bought in what proxysql solves, but I guess it's nice to have some options ;) | 13:10 |
| *** kevko__ has quit IRC | 13:10 | |
| kevko | mnasiadka: but, issue from log looks like no backend is alive ..so clustercheck is reporting backends failed ..but why ? if there is user and mariadb working (from log looks like it works) why haproxy is reporting down ? | 13:10 |
| mnasiadka | kevko: well, in the log you pasted (before I adapted wsrep notify) it is really down - for some reason the mysql update this script is trying to do is hanging forever and blocking any access to the database ;-) | 13:15 |
| mnasiadka | don't ask me, I touch databases as little as I can ;) | 13:15 |
| kevko | mnasiadka: hmm, ok | 13:16 |
| *** k-s-dean has quit IRC | 13:17 | |
| *** hjensas is now known as hjensas|afk | 13:18 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: CI: cephadm: wait after applying cluster spec https://review.opendev.org/c/openstack/kolla-ansible/+/786043 | 13:27 |
| *** k_mouza has quit IRC | 13:30 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla master: move to Debian 'bullseye' https://review.opendev.org/c/openstack/kolla/+/772479 | 13:39 |
| *** k_mouza has joined #openstack-kolla | 13:39 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: CI: cephadm: retry OSD tasks https://review.opendev.org/c/openstack/kolla-ansible/+/786043 | 13:42 |
| *** Luzi has quit IRC | 13:49 | |
| ozzzo | I'm following the document here to upgrade my test cluster from Train to Ussuri: https://docs.openstack.org/kolla-ansible/train/user/operating-kolla.html#upgrade-procedure | 13:49 |
| ozzzo | It says "Because of system technical limitations, upgrade of a libvirt container when using software emulation (virt_type = qemu in nova.conf), does not work at all. " | 13:49 |
| ozzzo | But I think my test cluster can only use qemu because it's built on VMs | 13:49 |
| ozzzo | how can I upgrade it to Ussuri without upgrading the libvirt container? | 13:49 |
| *** k_mouza has quit IRC | 13:53 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 13:53 |
| priteau | ozzzo: This is a comment that dates from 2016. I wonder if it is still valid? | 13:54 |
| dardelean | ozzzo I've done a test upgrade with VMs also and it worked just fine, without touching that config, don't know which is the default option there. Can you try that? | 13:54 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 13:55 |
| ozzzo | ok I'll try it | 13:55 |
| *** gfidente has quit IRC | 13:56 | |
| *** gfidente has joined #openstack-kolla | 13:56 | |
| dardelean | ozzzo you can also have kvm as virt_type, i have it on a cluster and I use VMs as well | 13:58 |
| ozzzo | I ran into some trouble when I used kvm | 13:58 |
| ozzzo | I forget the details but I asked around and someone told me I needed qemu for double-virt | 14:00 |
| ozzzo | and that fixed the problem | 14:00 |
| priteau | You probably don't have nested virtualization enabled? | 14:00 |
| ozzzo | I'm not sure I understand the question. Is there a config item for that? | 14:05 |
| *** luksky has quit IRC | 14:05 | |
| *** luksky has joined #openstack-kolla | 14:06 | |
| *** ivan_lin has quit IRC | 14:10 | |
| *** Luzi has joined #openstack-kolla | 14:10 | |
| *** ivan_lin has joined #openstack-kolla | 14:11 | |
| kevko | ozzzo: cat /sys/module/kvm_intel/parameters/nested | 14:12 |
| kevko | ozzzo: https://docs.fedoraproject.org/en-US/quick-docs/using-nested-virtualization-in-kvm/ | 14:12 |
| ozzzo | I have "N" - I guess that must have been causing the problem | 14:14 |
| priteau | ozzzo: It's something that needs to be configured on the hypervisor itself | 14:15 |
| ozzzo | oic, the hypervisors have "Y" | 14:16 |
| ozzzo | you mean the bare-metal HV that I am building the VMs on, to build the test cluster, right? | 14:16 |
| *** macz_ has joined #openstack-kolla | 14:17 | |
| ozzzo | turns out some have it and some don't. When I had the problem I must have been on one that had "N" | 14:18 |
| *** k_mouza has joined #openstack-kolla | 14:19 | |
| priteau | You need to enable it in the KVM VM as well | 14:21 |
| priteau | But anyway, do try the nova-libvirt upgrade with a non-kvm qemu | 14:21 |
| priteau | It might be an outdated comment | 14:22 |
| hrw | https://review.opendev.org/c/openstack/kolla/+/785871 mnasiadka, yoctozepto? | 14:29 |
| ozzzo | ok I'll try it; if that doesn't work I'll change the "N" to "Y" on the hypervisors | 14:34 |
| *** samcat116 has joined #openstack-kolla | 14:42 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 14:49 |
| *** Luzi has quit IRC | 14:51 | |
| *** samcat116 has quit IRC | 15:06 | |
| *** samcat116 has joined #openstack-kolla | 15:07 | |
| *** sorin-mihai has quit IRC | 15:11 | |
| *** sorin-mihai has joined #openstack-kolla | 15:11 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: mariadb: Fix wsrep notify script https://review.opendev.org/c/openstack/kolla-ansible/+/786036 | 15:26 |
| openstackgerrit | Merged openstack/kayobe stable/stein: Fix installation errors with Python 2 https://review.opendev.org/c/openstack/kayobe/+/785869 | 15:29 |
| *** luksky has quit IRC | 15:47 | |
| *** luksky has joined #openstack-kolla | 15:50 | |
| openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Extend support for custom Grafana dashboards https://review.opendev.org/c/openstack/kolla-ansible/+/786071 | 15:53 |
| *** LinPeiWen has quit IRC | 15:53 | |
| *** LinPeiWen has joined #openstack-kolla | 16:05 | |
| hrw | 1806 images to remove. | 16:05 |
| * hrw off | 16:07 | |
| *** ivan_lin has quit IRC | 16:07 | |
| *** kevko has quit IRC | 16:19 | |
| *** rpittau is now known as rpittau|afk | 16:24 | |
| *** dougsz has quit IRC | 16:35 | |
| *** e0ne has joined #openstack-kolla | 16:35 | |
| *** kevko has joined #openstack-kolla | 16:35 | |
| *** e0ne has quit IRC | 16:38 | |
| *** e0ne has joined #openstack-kolla | 16:42 | |
| *** fuhrmannb has quit IRC | 17:00 | |
| *** bengates_ has quit IRC | 17:07 | |
| *** bengates has joined #openstack-kolla | 17:08 | |
| *** bengates has quit IRC | 17:12 | |
| *** jonaspaulo has joined #openstack-kolla | 17:12 | |
| *** gfidente is now known as gfidente|afk | 17:16 | |
| *** samcat116 has quit IRC | 17:35 | |
| *** e0ne has quit IRC | 17:38 | |
| *** samcat116 has joined #openstack-kolla | 17:57 | |
| *** kevko has quit IRC | 18:49 | |
| *** samcat116 has quit IRC | 18:49 | |
| *** andrewbonney has quit IRC | 18:54 | |
| *** samcat116 has joined #openstack-kolla | 18:59 | |
| openstackgerrit | Merged openstack/kayobe stable/train: Fix installation errors with Python 2 https://review.opendev.org/c/openstack/kayobe/+/785624 | 18:59 |
| *** kevko has joined #openstack-kolla | 19:01 | |
| openstackgerrit | Merged openstack/kayobe stable/train: Fix help texts of inventory options https://review.opendev.org/c/openstack/kayobe/+/785924 | 19:04 |
| *** ivan_lin has joined #openstack-kolla | 19:17 | |
| *** LinPeiWen has quit IRC | 19:19 | |
| *** stingrayza has quit IRC | 19:20 | |
| *** stingrayza has joined #openstack-kolla | 19:20 | |
| *** kevko_ has joined #openstack-kolla | 19:21 | |
| *** kevko has quit IRC | 19:24 | |
| *** k_mouza has quit IRC | 19:49 | |
| *** samcat116 has quit IRC | 19:53 | |
| *** vishalmanchanda has quit IRC | 19:55 | |
| *** samcat116 has joined #openstack-kolla | 20:00 | |
| *** jonaspaulo has quit IRC | 20:31 | |
| *** samcat116 has quit IRC | 21:03 | |
| *** k_mouza has joined #openstack-kolla | 21:49 | |
| *** k_mouza has quit IRC | 21:53 | |
| *** luksky has quit IRC | 22:07 | |
| *** macz_ has quit IRC | 23:26 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!