| *** LinPeiWen has joined #openstack-kolla | 00:55 | |
| *** ivan_lin has quit IRC | 01:33 | |
| *** ivan_lin has joined #openstack-kolla | 01:34 | |
| *** ricolin has joined #openstack-kolla | 03:08 | |
| *** skramaja has joined #openstack-kolla | 04:30 | |
| *** vishalmanchanda has joined #openstack-kolla | 05:11 | |
| *** ricolin_ has joined #openstack-kolla | 05:52 | |
| *** Luzi has joined #openstack-kolla | 07:02 | |
| *** amoralej|off is now known as amoralej | 07:35 | |
| *** amoralej is now known as amoralej|off | 07:35 | |
| *** bengates has joined #openstack-kolla | 08:00 | |
| *** bengates has quit IRC | 08:04 | |
| *** bengates has joined #openstack-kolla | 08:05 | |
| *** bengates has quit IRC | 08:07 | |
| *** bengates has joined #openstack-kolla | 08:08 | |
| *** bengates has quit IRC | 08:11 | |
| *** bengates has joined #openstack-kolla | 08:12 | |
| *** andrewbonney has joined #openstack-kolla | 08:13 | |
| *** bengates has quit IRC | 08:17 | |
| *** bengates has joined #openstack-kolla | 08:17 | |
| *** rpittau|afk is now known as rpittau | 08:19 | |
| *** ug_egon has joined #openstack-kolla | 08:20 | |
| *** jbadiapa has joined #openstack-kolla | 08:59 | |
| mgoddard | morning | 09:00 |
|---|---|---|
| *** amoralej|off is now known as amoralej | 09:04 | |
| *** e0ne has joined #openstack-kolla | 09:10 | |
| hrw | morng | 09:24 |
| *** gfidente has joined #openstack-kolla | 09:24 | |
| hrw | yoctozepto: argh. | 09:27 |
| hrw | let me deploy something on local machine | 09:35 |
| hrw | hm. one month old images. time to build new ones | 09:36 |
| mnasiadka | morning | 09:42 |
| *** kevko has joined #openstack-kolla | 09:54 | |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible stable/victoria: nova-cell: Stop printing ceph keys in output https://review.opendev.org/c/openstack/kolla-ansible/+/782118 | 10:03 |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible stable/ussuri: nova-cell: Stop printing ceph keys in output https://review.opendev.org/c/openstack/kolla-ansible/+/782119 | 10:03 |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible stable/train: nova-cell: Stop printing ceph keys in output https://review.opendev.org/c/openstack/kolla-ansible/+/782120 | 10:04 |
| openstackgerrit | Michal Nasiadka proposed openstack/kayobe master: Bump max Ansible version to 2.10 https://review.opendev.org/c/openstack/kayobe/+/780898 | 10:05 |
| mgoddard | hrw: what's the status of https://review.opendev.org/c/openstack/kolla/+/781494 ? | 10:06 |
| hrw | mgoddard: need to check why it failed on k-a | 10:08 |
| hrw | mgoddard: other than that: mergeable | 10:08 |
| mgoddard | ok | 10:08 |
| hrw | INFO:kolla.common.utils.nova-libvirt:Trying to push the image | 10:10 |
| hrw | ERROR:kolla.common.utils.nova-libvirt:Get http://127.0.0.1:4000/v2/: dial tcp 127.0.0.1:4000: connect: connection refused | 10:10 |
| hrw | feels CI issue | 10:10 |
| kevko | hrw: i know why you failed :) | 10:10 |
| hrw | kevko: where? | 10:10 |
| kevko | hrw: in https://review.opendev.org/c/openstack/kolla/+/781494 | 10:10 |
| hrw | kevko: then share that knowledge | 10:11 |
| kevko | hrw: you have to rebase on top of my CI fix patch -> https://review.opendev.org/c/openstack/kolla-ansible/+/781151 | 10:11 |
| hrw | kevko: please rebase it then. | 10:12 |
| hrw | did recheck - should take care of it | 10:12 |
| hrw | kevko: thanks | 10:12 |
| kevko | hrw: oh, it's kolla project ..so It should work against current kolla-ansible master because it is already merged | 10:13 |
| kevko | hrw: so, probably recheck will work | 10:13 |
| hrw | kevko: that's why recheck. | 10:13 |
| kevko | +1 | 10:13 |
| hrw | redeploying 0322-1 images on aarch64 to check do they work at all | 10:14 |
| *** ricolin_ has quit IRC | 10:16 | |
| hrw | ok, destroyed. time for deploy | 10:28 |
| hrw | rabbitmq-- | 10:35 |
| hrw | 6efd5a69bb9 puchatek.lan:5000/master-0322-1/debian-source-rabbitmq:11.1.0 "dumb-init --single-…" 2 minutes ago Restarting (0) 46 seconds ago rabbitmq | 10:36 |
| *** bengates has quit IRC | 10:37 | |
| hrw | Error: This RabbitMQ version cannot run on Erlang 22.1.7 (erts 10.5.5): minimum required version is 22.3 (erts 10.7) | 10:37 |
| hrw | yay | 10:37 |
| *** bengates has joined #openstack-kolla | 10:38 | |
| hrw | oh rmq, how I dislike you | 10:38 |
| *** cah_link has joined #openstack-kolla | 10:38 | |
| kevko | :D | 10:48 |
| hrw | it also shows that X cycle will start with merging https://review.opendev.org/c/openstack/kolla/+/772479 patch | 10:49 |
| hrw | to get rid of Buster | 10:49 |
| openstackgerrit | Merged openstack/kolla-ansible master: Trivial fixes in grafana-guide doc. https://review.opendev.org/c/openstack/kolla-ansible/+/781488 | 10:51 |
| *** brtknr has quit IRC | 10:56 | |
| hrw | mgoddard: turns out that you and mnasiadka ignored Zuul failing on centos stream 8 binary. | 10:56 |
| mgoddard | hrw: doh | 10:56 |
| mgoddard | hrw: and you, to be fair :D | 10:57 |
| hrw | mgoddard: I did not noticed that too | 10:57 |
| mgoddard | hrw: now we have the root cause, thanks | 10:57 |
| hrw | I need to take some PTO or something | 10:57 |
| hrw | just this @#%@^@%@ pandemic blocks me from going somewhere to rest a bit on such PTO | 10:58 |
| mgoddard | hrw: rest at home? | 10:58 |
| hrw | unable to | 10:59 |
| hrw | bbiab | 11:00 |
| mnasiadka | centos binary? do we care? :) | 11:07 |
| openstackgerrit | Piotr Parczewski proposed openstack/kolla-ansible master: Add Alertmanger metric target(s) https://review.opendev.org/c/openstack/kolla-ansible/+/782210 | 11:55 |
| openstackgerrit | Piotr Parczewski proposed openstack/kolla-ansible master: Add Alertmanger metric target(s) https://review.opendev.org/c/openstack/kolla-ansible/+/782210 | 11:59 |
| *** cah_link has quit IRC | 12:11 | |
| *** brtknr has joined #openstack-kolla | 12:13 | |
| *** cah_link has joined #openstack-kolla | 12:28 | |
| hrw | ok. I broke Debian builds for a while | 12:32 |
| hrw | erlang shit | 12:32 |
| *** cah_link has quit IRC | 12:35 | |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: rabbitmq: bump Erlang requirements to v23 https://review.opendev.org/c/openstack/kolla/+/782247 | 12:38 |
| hrw | (venv3) 13:32 (0s) hrw@jagular:kolla$ ~/devel/docker-utils/kolla-build-images.sh debian source '--push --registry puchatek.lan:5000 ' master-0322-2 | 12:40 |
| hrw | and another rebuild... | 12:40 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: make DVR work on VLAN tenant networks https://review.opendev.org/c/openstack/kolla-ansible/+/780919 | 13:02 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: make DVR work on VLAN tenant networks https://review.opendev.org/c/openstack/kolla-ansible/+/780919 | 13:04 |
| mnasiadka | mgoddard, yoctozepto: Finally worked out octavia and OVN with CI - https://review.opendev.org/c/openstack/kolla-ansible/+/761872 - reviews please | 13:05 |
| hrw | kolla-ansible destroy++ | 13:05 |
| *** brinzhang has joined #openstack-kolla | 13:05 | |
| openstackgerrit | Merged openstack/kolla master: Build Masakari for Debian binary https://review.opendev.org/c/openstack/kolla/+/780409 | 13:13 |
| openstackgerrit | Merged openstack/kolla-ansible stable/ussuri: docs: fix registry mirror example https://review.opendev.org/c/openstack/kolla-ansible/+/781207 | 13:13 |
| openstackgerrit | Merged openstack/kolla-ansible stable/train: docs: fix registry mirror example https://review.opendev.org/c/openstack/kolla-ansible/+/781208 | 13:13 |
| openstackgerrit | Merged openstack/kolla-ansible stable/victoria: docs: fix registry mirror example https://review.opendev.org/c/openstack/kolla-ansible/+/781206 | 13:14 |
| openstackgerrit | Merged openstack/kolla-ansible stable/ussuri: nova-cell: Stop printing ceph keys in output https://review.opendev.org/c/openstack/kolla-ansible/+/782119 | 13:14 |
| *** amoralej is now known as amoralej|lunch | 13:21 | |
| hrw | so, who +2 https://review.opendev.org/c/openstack/kolla/+/781494 | 13:30 |
| hrw | so we get centos stream 8 binary working? | 13:30 |
| mgoddard | hrw: centos binary job failed... | 13:35 |
| *** sean-k-mooney has joined #openstack-kolla | 13:37 | |
| hrw | ha! I got you to look ;D | 13:38 |
| hrw | checking locally why it failed | 13:40 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: centos: enable PowerTools repo in images which needs it https://review.opendev.org/c/openstack/kolla/+/781494 | 13:47 |
| hrw | mgoddard: fixed watcher and changed commit message | 13:48 |
| hrw | ok, time to dig in nova | 13:56 |
| yoctozepto | mgoddard, mnasiadka, hrw: what to do about quay? | 13:59 |
| hrw | go for it | 14:00 |
| * hrw -> firefighting mode | 14:00 | |
| openstackgerrit | Martin Chlumsky proposed openstack/kolla-ansible master: Switch octavia-api to wsgi running under apache. https://review.opendev.org/c/openstack/kolla-ansible/+/779893 | 14:06 |
| *** skramaja has quit IRC | 14:06 | |
| hrw | yoctozepto: Debian/aarch64 will be fixed | 14:13 |
| *** luksky has joined #openstack-kolla | 14:16 | |
| hrw | forgot to bump libvirt-python ;( | 14:25 |
| *** amoralej|lunch is now known as amoralej | 14:25 | |
| *** e0ne has quit IRC | 14:27 | |
| hrw | meh, linaro infra issues | 14:40 |
| hrw | so Debian stays broken for a while | 14:40 |
| kevko | mgoddard: around ? did u check mariadb-refactor ? if no, could you check if this is acceptable ? because if no.. i have reworked branch which i can push.. but i will be happy if you can check firstly current version | 14:41 |
| mgoddard | yoctozepto: discuss on wednesday | 14:43 |
| *** Luzi has quit IRC | 14:55 | |
| yoctozepto | mgoddard: ok | 15:01 |
| yoctozepto | fine with my time | 15:01 |
| mgoddard | yoctozepto: I've added it to the agenda | 15:05 |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: Ubuntu: AppArmor config for seed VM provisioning https://review.opendev.org/c/openstack/kayobe/+/782272 | 15:08 |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: Enable bare metal testing for Ubuntu https://review.opendev.org/c/openstack/kayobe/+/782273 | 15:10 |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: refactor zuul config to use templates https://review.opendev.org/c/openstack/kayobe/+/782276 | 15:17 |
| *** schwicht has quit IRC | 15:32 | |
| *** schwicht has joined #openstack-kolla | 15:35 | |
| *** e0ne has joined #openstack-kolla | 15:39 | |
| openstackgerrit | James Kirsch proposed openstack/kolla master: Add LetsEncrypt images for cert request/renewal https://review.opendev.org/c/openstack/kolla/+/741339 | 16:08 |
| *** vishalmanchanda has quit IRC | 16:11 | |
| yoctozepto | mgoddard: thanks | 16:20 |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: refactor zuul config to use templates https://review.opendev.org/c/openstack/kayobe/+/782276 | 16:30 |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: Enable bare metal testing for Ubuntu https://review.opendev.org/c/openstack/kayobe/+/782273 | 16:35 |
| *** amoralej is now known as amoralej|off | 16:46 | |
| *** ysirndjuro has joined #openstack-kolla | 17:04 | |
| *** rpittau is now known as rpittau|afk | 17:11 | |
| *** e0ne has quit IRC | 17:13 | |
| kevko | mgoddard: did you check my new patchset ? | 17:14 |
| mgoddard | kevko: hi, not yet | 17:14 |
| mgoddard | kevko: will look later | 17:14 |
| kevko | mgoddard: i've changed few things .. locally I have reworked whole chain (mainly moved set_fact to defaults or groups vars) | 17:15 |
| kevko | but not pushed yet .. | 17:15 |
| kevko | mgoddard: i have it in case you will not like my current +1 jenkins verified chain :) | 17:16 |
| mgoddard | kevko: looks like this is the only change: https://review.opendev.org/c/openstack/kolla-ansible/+/766952/43..47/ansible/roles/mariadb/tasks/recover_cluster.yml | 17:17 |
| mgoddard | kevko: this comment isn't really addressed: https://review.opendev.org/c/openstack/kolla-ansible/+/766952/42..47/ansible/roles/mariadb/tasks/loadbalancer.yml | 17:20 |
| kevko | mgoddard: this is one of changes i've locally changed .. but it's hard to rebase whole chain .. so i not pushed yet .. | 17:22 |
| mgoddard | ok | 17:22 |
| kevko | mgoddard: wait a minute | 17:22 |
| kevko | mgoddard: this is diff what i currently have -> http://paste.openstack.org/show/803782/ | 17:28 |
| kevko | mgoddard: but still testing whole chain | 17:28 |
| kevko | mgoddard: is it acceptable ? | 17:29 |
| kevko | mgoddard: i had to move it in group_vars/all as user can run with --tags haproxy, or --tags mariadb and list of default shard/cluster members has to be accessible from both places .. | 17:31 |
| kevko | because tasks/loadbalancer.yml is running always ..if you are running --tags haproxy or --tags mariadb ..but of course ...you know it | 17:31 |
| openstackgerrit | James Kirsch proposed openstack/kolla master: Add LetsEncrypt images for cert request/renewal https://review.opendev.org/c/openstack/kolla/+/741339 | 17:32 |
| kevko | *and removed set_facts tasks | 17:32 |
| mgoddard | kevko: will need to review the whole patch, but it does look like a better approach. mariadb_default_database_shard_id is good, although I don't know if 'default' captures the intent | 17:33 |
| mgoddard | kevko: what does it mean to be in the default shard? | 17:34 |
| mgoddard | haproxy load balancing? anything else? | 17:34 |
| mgoddard | maybe mariadb_haproxy_shard_id is clearer? | 17:34 |
| kevko | mgoddard: i have to have id int because in proxysql configuration i need it to generate writer,writer_backup,reader_group etc | 17:35 |
| kevko | it is identificator of mariadb_shard | 17:35 |
| kevko | or let's say mariadb_cluster | 17:35 |
| *** bengates has quit IRC | 17:36 | |
| kevko | default shard means ..user didn't fill value in inventory ..so default it is 0, to work group_by grouping | 17:36 |
| hrw | yoctozepto, mnasiadka, osmanlicilegi: https://review.opendev.org/c/openstack/kolla/+/781494 | 17:37 |
| *** gfidente is now known as gfidente|afk | 17:40 | |
| hrw | mgoddard: I am going to drop ubuntu/aarch64 from CI | 17:42 |
| hrw | mgoddard: it needs newer erlang and I do not have any interest in keeping it alive | 17:43 |
| mgoddard | kevko: yes, although by making it a variable it's configurable. We also need to document what happens for nodes in shard 0 | 17:43 |
| mgoddard | hrw: :( ok | 17:44 |
| hrw | looks like no one uses ubuntu in kolla anyway | 17:44 |
| yoctozepto | lol | 17:44 |
| hrw | because no one noticed that we use Victoria UCA | 17:44 |
| kevko | mgoddard: as haproxy can LB only one mariadb shard/cluster (one tcp port) ..it is that one which will have VIP:3306 configured ..other shards/cluster will be just deployed | 17:45 |
| yoctozepto | but but... should not we use Victoria UCA? :P | 17:45 |
| kevko | mgoddard: if proxysql will be turned on, there will be VIP:3306 for all shards/clusters and accessible via root_shard_ID | 17:45 |
| yoctozepto | wallaby still in devel | 17:46 |
| kevko | hrw: we are using ubuntu as baremetal ..but debian images ... | 17:46 |
| kevko | *debian docker images | 17:46 |
| hrw | yoctozepto: we ARE in wallaby. we USE wallaby packages IF they are available | 17:46 |
| hrw | yoctozepto: we use RDO wallaby. | 17:47 |
| hrw | we used UCA develbranch in past | 17:47 |
| hrw | I was bumping it since train but in wallaby I was less involved in kolla than usual so no one notices | 17:48 |
| hrw | I was bumping it since train but in wallaby I was less involved in kolla than usual so no one noticed | 17:48 |
| yoctozepto | oj tam oj tam | 17:48 |
| yoctozepto | we check on release | 17:48 |
| hrw | that's why I suspect that ubuntu images were not used by anyone ;D | 17:48 |
| kevko | oj tam oj tam ? :D | 17:48 |
| hrw | kevko: Polish | 17:48 |
| kevko | no ved toto :) sa mi to zdalo :D | 17:49 |
| kevko | yeah, understood :D | 17:49 |
| hrw | kevko: we joked few times in past that Kolla one day will switch to Polish | 17:49 |
| kevko | hrw: no problem, understand polish :D | 17:49 |
| yoctozepto | we will call the project | 17:49 |
| yoctozepto | klej | 17:49 |
| yoctozepto | and klej-buduj | 17:50 |
| kevko | joo :D | 17:50 |
| yoctozepto | jo na pewno :D | 17:50 |
| kevko | funny language :) | 17:51 |
| kevko | no offense .. | 17:51 |
| hrw | yoctozepto: you forgot about KajObe | 17:51 |
| yoctozepto | hrw: kanabe | 17:52 |
| kevko | https://www.mojevideo.sk/video/287ec/pas_jest_kurwa_zapienty_(nastvany_polsky_kamionista).html | 17:52 |
| kevko | :D | 17:52 |
| hrw | kevko: we think similar about czechoslovak | 17:52 |
| kevko | yeah, i know :) | 17:52 |
| mgoddard | kevko: I think there are 3 modes of usage | 17:52 |
| yoctozepto | yeah, some of the dissimilarities make it funny to the other party | 17:53 |
| kevko | mgoddard: listening | 17:53 |
| mgoddard | 1. old school, single default shard, haproxy | 17:53 |
| * yoctozepto feels oldschool | 17:53 | |
| mgoddard | 2. new school, one or more shards, all proxysql | 17:53 |
| * yoctozepto wants to join new school | 17:53 | |
| mgoddard | 3. cells, one old school haproxy balanced shard + multiple (single node?) non-load balanced shards | 17:54 |
| yoctozepto | why can't it use proxysql too? | 17:54 |
| mgoddard | yoctozepto: it could | 17:55 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: base: use UCA Wallaby https://review.opendev.org/c/openstack/kolla/+/782308 | 17:55 |
| yoctozepto | then let's make it so | 17:55 |
| mgoddard | yoctozepto: but that would push all cell DB traffic through proxysql, which you might not want | 17:55 |
| kevko | mgoddard: oh, it's too complex, don't you think ? | 17:56 |
| yoctozepto | mgoddard: why not more proxysql instances? | 17:56 |
| hrw | https://review.opendev.org/c/openstack/kolla/+/782247 waits for brave cores wanting to help fix Debian/aarch64 | 17:56 |
| yoctozepto | later this week I will see to that stack | 17:56 |
| mgoddard | yoctozepto: out of scope :) | 17:56 |
| mgoddard | 3 is the CERN cells model | 17:56 |
| yoctozepto | mgoddard: :-(((((((((((( | 17:56 |
| kevko | mgoddard: well, honestly it is more easy to implement as it is now .. | 17:57 |
| kevko | mgoddard: i'm deleting mariadb haproxy cfg and stop/start haproxy to free 3306 to start proxysql if enabled | 17:58 |
| mgoddard | anyway, those modes were a suggestion | 17:58 |
| mgoddard | if it's just as simple as 1 or 2, then I think some logic could be simplified | 17:59 |
| kevko | mgoddard: if you are saying that you don't want to have all traffic via proxysql ... you probably want to have proxysql/haproxy on different hosts | 18:01 |
| mgoddard | well, maybe | 18:01 |
| kevko | but, in this case you have to also another VIP and another keepalived | 18:01 |
| kevko | i have it written like that ...but you said on begginning it's too complex .. | 18:01 |
| kevko | so, i've written it on one VIP/keepalived when haproxy OR proxysql can be turned on .. | 18:02 |
| kevko | but, not saying it is not possible to implement in follow up patches .. | 18:03 |
| mgoddard | it's possible, but out of scope for now | 18:03 |
| mgoddard | anyway, I'm just trying to reverse engineer use cases from code | 18:03 |
| kevko | but honestly my chain is growing, further, the harder it is to rebase | 18:03 |
| mgoddard | indeed | 18:04 |
| kevko | mgoddard: we are using it now as router to several database shards/clusters | 18:04 |
| kevko | mgoddard: nova, neutron, and others | 18:04 |
| mgoddard | ok, you tell me how it's supposed to work | 18:05 |
| mgoddard | would I ever mixe HAProxy & ProxySQL in front of mariadb? | 18:05 |
| mgoddard | I guess not due to VIP port clash | 18:05 |
| kevko | mgoddard: because connections to DB ( nova(specially) and neutron was overkilled) ..and we had to split traffic .. | 18:06 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: ubuntu: drop not existing erlang repo on obs.linaro.org https://review.opendev.org/c/openstack/kolla/+/782311 | 18:06 |
| mgoddard | kevko: I get the pure proxysql use case | 18:06 |
| mgoddard | kevko: what I'm trying to work out, is what other combinations it's supposed to support | 18:07 |
| kevko | mgoddard: but, of course i wrote it as replacement for mariadb LB .. | 18:07 |
| kevko | mgoddard: better monitoring of galera (native), and what is cool ..you can see direct sql statement to DB, gather statistics about connections, sql statements, and ton of usefull informations | 18:08 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: CI: drop ubuntu-aarch64 job https://review.opendev.org/c/openstack/kolla/+/782312 | 18:08 |
| kevko | mgoddard: and you can also configuring proxysql without restart ! | 18:09 |
| kevko | mgoddard: just configure in internal DB and load config .. | 18:09 |
| *** andrewbonney has quit IRC | 18:09 | |
| mgoddard | kevko: I don't need the proxysql sales pitch :D | 18:09 |
| kevko | :D | 18:09 |
| mgoddard | kevko: just want to know, if using haproxy, are multiple shards supported, and if so, what happens with them? | 18:10 |
| mgoddard | (I think that's the question I've been trying to ask | 18:10 |
| mgoddard | need to go, child calls | 18:10 |
| kevko | mgoddard: well, we are currently running it in production and it's fine .. but my fear grows every day when it is not occupied in the master, because to carry such patches is not easy | 18:10 |
| kevko | merged in master | 18:11 |
| kevko | mgoddard: if you are using haproxy, but define several shards in inventory ... you will have default shard behind VIP ..other shards only deployed ..without LB | 18:12 |
| kevko | mgoddard: of course, you can rework internal_haproxy_members to listen on port 3306+shard_id for example ..so you will have several ports maintained by haproxy ..doing LB for serveral shards | 18:13 |
| kevko | mgoddard: but from my perspective of view it is useless .. because you can just use proxysql .. and have only one 3306 port | 18:14 |
| kevko | mgoddard: as it is implemented now (with small changes)- it's possible to have haproxy for default shard ..and proxysql for others | 18:15 |
| *** jonaspaulo has joined #openstack-kolla | 18:15 | |
| kevko | mgoddard: small changes should be change proxysql to listen on different VIP:port for example 3307 ..and testing when: id != default_shard_id | 18:16 |
| kevko | mgoddard: on the end you will have mariadb role which is able to install several shards/clusters ...and you can also configure ids in inventory ..and based on that ids you can split that traffic this way | 18:17 |
| kevko | mgoddard: what is not very nice is ..that in this case every service which will have DB on another shard will have to regenerate config to point to VIP:3307 | 18:18 |
| kevko | as it is now .. you just switch haproxy with proxysql ..and everything works ..nothing changed from services view | 18:19 |
| kevko | hrw: looks like no one uses ubuntu in kolla anyway <<< how did you noticed this ? is there some statistics ? | 18:41 |
| yoctozepto | kevko: he just assumed this because noone bumped the uca release for wallaby... | 18:55 |
| kevko | uca ? | 18:55 |
| yoctozepto | kevko: Ubuntu Cloud Archive | 18:55 |
| yoctozepto | it's analogous to RDO in the CentOS world | 18:55 |
| yoctozepto | or debian's per-openstack-release repos | 18:56 |
| kevko | btw, guys, if you want to check current deb packages for OS -> you can check here -> https://debian.kevko.ultimum.cloud/ | 18:56 |
| kevko | ah, ok | 18:56 |
| *** hrw has quit IRC | 19:13 | |
| *** rockey has joined #openstack-kolla | 19:17 | |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: Fix baremetal SSH https://review.opendev.org/c/openstack/kayobe/+/782324 | 19:50 |
| *** kevko has quit IRC | 20:24 | |
| *** suff has joined #openstack-kolla | 20:39 | |
| *** jonaspaulo has quit IRC | 20:43 | |
| *** kevko has joined #openstack-kolla | 20:47 | |
| suff | Hi guys! I want to add heartbeat variables for octavia to upstream. Before my MR, I need to create bug at launchpad? | 20:51 |
| suff | Or features? | 20:51 |
| openstackgerrit | Piotr Parczewski proposed openstack/kolla-ansible master: Disable usage collection in Kibana https://review.opendev.org/c/openstack/kolla-ansible/+/782330 | 21:21 |
| *** e0ne has joined #openstack-kolla | 21:25 | |
| openstackgerrit | Piotr Parczewski proposed openstack/kolla-ansible master: Disable usage collection in Kibana https://review.opendev.org/c/openstack/kolla-ansible/+/782330 | 21:28 |
| openstackgerrit | Dmitry Grachev proposed openstack/kolla-ansible master: Add octavia default heartbeats vars https://review.opendev.org/c/openstack/kolla-ansible/+/782332 | 21:42 |
| *** hrw has joined #openstack-kolla | 21:44 | |
| yoctozepto | suff: it is very likely enough to propose the change with proper commit message, documentation etc. | 22:01 |
| yoctozepto | we will let you know what's missing | 22:01 |
| yoctozepto | also, we are mostlly in European timezones here so don't feel bad if we don't reply THAT LATE :D | 22:02 |
| *** kevko has quit IRC | 22:07 | |
| suff | yoctozepto: Ok, thanks! I am in European timezone too :) | 22:15 |
| *** e0ne has quit IRC | 22:28 | |
| *** e0ne has joined #openstack-kolla | 22:29 | |
| openstackgerrit | James Kirsch proposed openstack/kolla-ansible master: Add support for LetsEncrypt-managed certs https://review.opendev.org/c/openstack/kolla-ansible/+/741340 | 22:33 |
| *** e0ne has quit IRC | 22:41 | |
| *** e0ne has joined #openstack-kolla | 22:46 | |
| *** e0ne has quit IRC | 22:51 | |
| *** e0ne has joined #openstack-kolla | 23:08 | |
| *** e0ne has quit IRC | 23:09 | |
| openstackgerrit | James Kirsch proposed openstack/kolla-ansible master: Add support for LetsEncrypt-managed certs https://review.opendev.org/c/openstack/kolla-ansible/+/741340 | 23:21 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!