| noonedeadpunk | mornings | 07:38 |
|---|---|---|
| noonedeadpunk | 1 more vote needed for this one https://review.opendev.org/c/openstack/openstack-ansible/+/880607 | 07:38 |
| noonedeadpunk | And I will push Y release on top of it | 07:38 |
| noonedeadpunk | Also these are required for Z bump https://review.opendev.org/q/parentproject:openstack/openstack-ansible+branch:%255Estable/zed+status:open+ | 07:39 |
| noonedeadpunk | adjutant is already being re-checked | 07:39 |
| damiandabrowski | hi | 07:54 |
| jrosser | morning | 08:05 |
| noonedeadpunk | regarding haproxy - I think this is the next thing to land https://review.opendev.org/c/openstack/openstack-ansible/+/880775 | 10:03 |
| opendevreview | Merged openstack/openstack-ansible stable/yoga: Gather generic masakari facts https://review.opendev.org/c/openstack/openstack-ansible/+/880607 | 10:24 |
| opendevreview | Merged openstack/openstack-ansible-os_octavia master: Do not limit IP prefix for DHCP rule https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/880804 | 10:41 |
| opendevreview | Merged openstack/openstack-ansible-os_octavia master: Change default CIDR for security_group https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/880544 | 10:41 |
| opendevreview | Merged openstack/openstack-ansible-ceph_client stable/zed: Improve regexp for fetching nova secret from files https://review.opendev.org/c/openstack/openstack-ansible-ceph_client/+/880493 | 10:56 |
| KSR | Hi' | 11:07 |
| noonedeadpunk | o/ | 11:07 |
| KSR | I'm about to upgrade a Openstack Ussuri to Victoria - I've understood the most recent stable openstack-ansible victoria branch is victoria-em, but is that based on 22.4.4 or 22.4.0 which is the most recent according to https://docs.openstack.org/releasenotes/openstack-ansible/victoria.html ? | 11:09 |
| KSR | I mean.. It seems like there might be a deviation between what is released, and what is announced on the docs site | 11:12 |
| noonedeadpunk | ugh... | 11:31 |
| noonedeadpunk | got distracted :( | 11:31 |
| opendevreview | Merged openstack/openstack-ansible stable/zed: bump openstack_hosts role to resolve openvswitch3.1 problem on Rocky https://review.opendev.org/c/openstack/openstack-ansible/+/880826 | 11:46 |
| opendevreview | Merged openstack/openstack-ansible stable/zed: Gather generic masakari facts https://review.opendev.org/c/openstack/openstack-ansible/+/880606 | 11:46 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-galera_server master: Define backup randomized delay in defaults https://review.opendev.org/c/openstack/openstack-ansible-galera_server/+/880950 | 12:09 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-galera_server master: Define backup randomized delay in defaults https://review.opendev.org/c/openstack/openstack-ansible-galera_server/+/880950 | 12:10 |
| opendevreview | Merged openstack/openstack-ansible master: Fix blazar haproxy service https://review.opendev.org/c/openstack/openstack-ansible/+/880775 | 12:53 |
| noonedeadpunk | Adjuntant passed now https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/879858 | 13:09 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Add way to periodically trim Nova DB https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/880971 | 13:13 |
| noonedeadpunk | ^ this is a thing I was thinking to do for quite a while now, but never had a time | 13:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Add way to periodically trim Nova DB https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/880971 | 13:17 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_cinder master: Add way to periodically trim Cinder DB https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/880975 | 14:02 |
| opendevreview | Merged openstack/openstack-ansible master: Add is_nest property for container_skel https://review.opendev.org/c/openstack/openstack-ansible/+/869762 | 14:51 |
| spatel | noonedeadpunk jamesdenton question with OVS it use LinuxBridge for security group so too many layers here. why don't use OVS built-in firewall? is there any issue to use native firewall of OVS and remove linuxbridge? | 15:41 |
| noonedeadpunk | I think it's jsut historical and/or due to some limitations. Maybe James know more. And there was option to use native firewalling for OVS as well | 15:42 |
| mgariepy | i use ovs flow and it works just fine . | 15:43 |
| noonedeadpunk | yeah, so likely it's jsut default that should be easy to override | 15:45 |
| mgariepy | our doc does deploy ovs on top of a lxb. so i guess it's why it's like that. | 15:45 |
| mgariepy | i have my tunnel interface > ovs direcly. no lxb between them, | 15:46 |
| spatel | Where is that setting to tell don't use lxb and use ovs firewall? is that in ml2 file? | 15:47 |
| noonedeadpunk | `neutron_firewall_driver`? | 15:48 |
| noonedeadpunk | Default is 'iptables_hybrid' but you can set it to 'openvswitch': https://opendev.org/openstack/openstack-ansible-os_neutron/src/branch/master/releasenotes/notes/override-ovs-firewall-driver-0ba4dee50aff725c.yaml | 15:49 |
| mgariepy | https://paste.openstack.org/show/bf08d5bx8YL4UVQ6GZVF/ | 15:50 |
| noonedeadpunk | spatel: btw this release not also explains the reason why it was like that | 15:50 |
| noonedeadpunk | mgariepy: yeah, there's easier way now... | 15:50 |
| mgariepy | on victoria there was a small issue but it's fixed in neutron now: | 15:51 |
| mgariepy | yeah, i'll revisit at some point :D | 15:51 |
| noonedeadpunk | mgariepy: was you migrating from hyprid or initial setup was like that? | 15:51 |
| mgariepy | initial like that | 15:51 |
| noonedeadpunk | I kinda wonder if I can just switch to native firewall... | 15:52 |
| spatel | This is awesome if we can switch with just config option | 15:52 |
| noonedeadpunk | I never tried that but eager to know :D | 15:52 |
| mgariepy | not sure if it would flush the iptables | 15:53 |
| spatel | OVN deployment by default use OVS based ACL so its not new anymore :) | 15:53 |
| noonedeadpunk | well, flush iptables is not a rocket science. I think it won't | 15:53 |
| noonedeadpunk | But I kinda wonder how long it might take to apply rules in ovs | 15:54 |
| NeilHanlon | poor Linux bridges, always getting hate | 15:54 |
| spatel | how does existing rules get consume by OVS ? | 15:54 |
| noonedeadpunk | I love LXB | 15:54 |
| NeilHanlon | hehe | 15:54 |
| noonedeadpunk | And hate ovs - it's so troublesome... | 15:54 |
| spatel | LXB is amazing... no doubt but poor thing end of life | 15:54 |
| noonedeadpunk | it's so much more clear and straightforwad with lxb... | 15:55 |
| spatel | if OVS is troublesome then just think of OVN.. hehehe!! | 15:55 |
| noonedeadpunk | Well, ovn at least does not have namespaces... | 15:56 |
| spatel | agreed!! +1 | 15:56 |
| noonedeadpunk | except metadata... | 15:56 |
| spatel | I don't understand why metadata isn't part of OVN, why its so hard to implement in flow? | 15:57 |
| noonedeadpunk | I have literally no idea and wasn't trying to understand that yet | 15:57 |
| spatel | Upgrading today my cloud from Wallaby to Xena :) hope today is good day!! | 15:58 |
| noonedeadpunk | that should be quite straighforward | 15:59 |
| noonedeadpunk | spatel: fwiw https://review.opendev.org/c/openstack/releases/+/878118 | 15:59 |
| noonedeadpunk | there're quite some changes from last release | 15:59 |
| noonedeadpunk | and that's close to last one - next will be just EM | 16:00 |
| spatel | so go with 2646.0 | 16:00 |
| spatel | 24.6.0 | 16:00 |
| noonedeadpunk | or you can `git checkout 6d863d13458bd185550f3c8ea457b9fda821c7c5` as well, which will be the same content. But might mess up venv names | 16:01 |
| spatel | oh wait.. what is that mess? | 16:02 |
| noonedeadpunk | well, names will be like 24.5.1.dev** | 16:03 |
| noonedeadpunk | or smth | 16:03 |
| jrosser | you can always apply your own custom tag if you want | 16:04 |
| noonedeadpunk | OR, after bootstrap you can edit /usr/local/bin/openstac-ansible and replace value for `OSA_VERSION` | 16:05 |
| noonedeadpunk | but yeah, still in pip freeze output would not be neat... | 16:05 |
| noonedeadpunk | OR! You can just create a local tag after checkout :) | 16:06 |
| spatel | why do i need to go through this process? what if people just following official upgrade doc? | 16:07 |
| noonedeadpunk | Like `git checkout 6d863d13458bd185550f3c8ea457b9fda821c7c5; git tag -a 24.6.0 -m "Local osa tag"` | 16:07 |
| noonedeadpunk | spatel: it's if you want to do this today | 16:07 |
| noonedeadpunk | As it's unlikely that patch I've provided for releasing will be merged | 16:08 |
| noonedeadpunk | And no new version is created until it's merged | 16:08 |
| NeilHanlon | noonedeadpunk: btw re https://review.opendev.org/c/openstack/openstack-ansible/+/880826 - was that the 'bump versions' script you were talking about yesterday? would love to follow along to see how it's done | 16:10 |
| spatel | oh ok.. so if i just do -> git checkout 6d863d13458bd185550f3c8ea457b9fda821c7c5; git tag -a 24.6.0 -m "Local osa tag" in that case I don't need to do anything else and just regular upgrade process correct? | 16:10 |
| noonedeadpunk | NeilHanlon: https://docs.openstack.org/openstack-ansible/latest/contributor/periodic-work.html#osa-cli-tooling | 16:10 |
| NeilHanlon | ty! reading :) | 16:10 |
| noonedeadpunk | Just replace `evrardjp` with `noonedeadpunk` | 16:11 |
| mgariepy | i wonder if the iptables rules would be flushed or not if you change the config and restart the agent. | 16:13 |
| mgariepy | sorry i was busy in another meeting. | 16:14 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: [doc] Update releasing tool location https://review.opendev.org/c/openstack/openstack-ansible/+/881030 | 16:14 |
| mgariepy | when i deployed my ovs one my plan was to switch to ovn at some point so i did use the flow directly. | 16:14 |
| noonedeadpunk | Well, if you change to firewall to ovs I bet iptables won't be flushed on it's own | 16:15 |
| mgariepy | but i got hit by this : https://bugs.launchpad.net/neutron/+bug/1912651 | 16:15 |
| NeilHanlon | OVS makes me flashback to working with XenServer on a daily basis | 16:15 |
| noonedeadpunk | looks like it's baclported a well | 16:16 |
| mgariepy | also if you switch the firewall you might want to get rid of the lxb that is not needed. | 16:16 |
| mgariepy | yep. it was fixed correctly :D | 16:16 |
| mgariepy | when you invest time poking ppl it does happens. | 16:16 |
| NeilHanlon | who needs firewalls, anyways. i thought it was all zero trust now | 16:16 |
| noonedeadpunk | NeilHanlon: I was lucky and had old-school Xen with LXB :) | 16:16 |
| noonedeadpunk | lol | 16:16 |
| NeilHanlon | noonedeadpunk: i'm quite friendly with the XCP-NG folks still... happy they're doing good things with the garbage state Citrix left it in | 16:17 |
| NeilHanlon | but gosh OVS was always such a PITA. mostly due to the version. newer OVS is a lot better | 16:18 |
| NeilHanlon | we'd have flow tables just.. drop out ? | 16:18 |
| noonedeadpunk | well we had terrible struggle with OVS due to it's gcc compatability issues | 16:20 |
| noonedeadpunk | So it was getting stuck from time to time and each restart was painful due to flow being lost iirc. | 16:20 |
| noonedeadpunk | Such a nightmare | 16:21 |
| noonedeadpunk | The solution they've proposed was gcc update :) | 16:21 |
| NeilHanlon | oof | 16:22 |
| admin1 | spatel, i did a 22 -> 26 recently, following offical upgrade docs .. the only point where i had to override was for new erlang versions for rabbitmq that was not the tags | 18:22 |
| admin1 | i just went to the last tag of every release | 18:22 |
| spatel | That is a very big jump 22 to 26 | 18:33 |
| admin1 | 22 -> 23 -> 24 -> 25 -> 26 :D | 18:35 |
| opendevreview | Merged openstack/openstack-ansible-os_keystone master: Rename keystone_ssl to keystone_backend_ssl https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/879379 | 18:46 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_glance master: Move _glance_available_stores to defaults https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/880872 | 19:10 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible master: Add support for TLS backends https://review.opendev.org/c/openstack/openstack-ansible/+/879085 | 19:10 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_zun master: Add TLS support to zun backends https://review.opendev.org/c/openstack/openstack-ansible-os_zun/+/881140 | 19:40 |
| psymin | Which distribution and version am I likely to have the most success with while using openstack-ansible? | 20:15 |
| damiandabrowski | psymin: we currently test ubuntu, debian and rocky in our CI so I'd pick one of them | 20:32 |
| psymin | is the output for CI public somewhere? | 20:33 |
| damiandabrowski | when it comes to openstack-ansible version, I'd recommend latest stable(so zed) | 20:34 |
| damiandabrowski | yes, you can check the results for each patch for ex. here: https://review.opendev.org/c/openstack/openstack-ansible/+/880775/1?tab=change-view-tab-header-zuul-results-summary | 20:34 |
| damiandabrowski | but I'm not aware of any periodic CI jobs | 20:35 |
| damiandabrowski | you can also find some graphs from our CI here: https://grafana.opendev.org/d/8cbc9d7032/openstack-ansible?orgId=1&from=now-7d&to=now | 20:35 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_adjutant master: Ensure service is restarted on unit file changes https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/879829 | 20:46 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_adjutant master: Add TLS support to adjutant backends https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/881147 | 20:55 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_magnum master: Ensure service is restarted on unit file changes https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/879970 | 21:03 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_magnum master: Add TLS support to magnum backends https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/879918 | 21:04 |
| psymin | Should Rocky Linux 8 work with openstack-ansible to deploy Yoga as hinted on this page? https://docs.openstack.org/project-deploy-guide/openstack-ansible/yoga/deploymenthost.html | 21:35 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!