f0o | noonedeadpunk: considering it's just tuesday (~2 days since the distributed fip) I see about 1.5Gbps more traffic than before while the Routers have significant less context switches and overall load. They now only deal with routers (and unfortunately ipv6) | 06:52 |
---|---|---|
noonedeadpunk | how come with ipv6? | 07:11 |
f0o | ipv6 is routed, so it goes though the neutron routers | 07:11 |
noonedeadpunk | ok, I guess it's us having ipv6 acting pretty much same as ipv4 - just a route instead of nat. | 07:13 |
f0o | so you just have the tenants on the actual public vlan? | 07:13 |
f0o | and have a /48 (or larger) onlink? | 07:13 |
noonedeadpunk | no, not really, we have a subnet pool and a /64 added to the public vlan. and then tenants get their own /64 from subnet pool which is indeed routed | 07:15 |
noonedeadpunk | but we also don't let tenants to create ports directly in ipv4 either | 07:15 |
noonedeadpunk | so public access only through fips | 07:16 |
f0o | but that sounds similar to what we have | 07:16 |
noonedeadpunk | yeah | 07:16 |
f0o | v6 goes through the tenant router which will always make it reside/terminate on the gateway noeds | 07:16 |
f0o | nodes* | 07:16 |
noonedeadpunk | I just got a bit confused as same applies for ipv4 for us, except it's nat and not route... | 07:17 |
f0o | luckily no NAT involved but for some reason openvswitch still uses conntrack for v6 eventhough its entirely routed | 07:17 |
noonedeadpunk | (or well, sometimes it;s route as well) | 07:17 |
noonedeadpunk | yeah, that what's kinda surpising | 07:17 |
f0o | I wonder if thats just an oversight from neutron/ovn in the flow designs since v6 can really just be stateless entirely | 07:18 |
f0o | wait | 07:18 |
f0o | I lied | 07:18 |
f0o | just dumped my conntrack -L and only see v4 now | 07:19 |
f0o | maybe it was some leftovers | 07:19 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move argument parsing for dynamic_inventory to generate https://review.opendev.org/c/openstack/openstack-ansible/+/945025 | 07:21 |
noonedeadpunk | ok, then it feels more in line with what I'd expect to see :) | 07:24 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move argument parsing for dynamic_inventory to generate https://review.opendev.org/c/openstack/openstack-ansible/+/945025 | 07:39 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move argument parsing for dynamic_inventory to generate https://review.opendev.org/c/openstack/openstack-ansible/+/945025 | 07:43 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move argument parsing for dynamic_inventory to generate https://review.opendev.org/c/openstack/openstack-ansible/+/945025 | 07:44 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move argument parsing for dynamic_inventory to generate https://review.opendev.org/c/openstack/openstack-ansible/+/945025 | 08:24 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Install inventory-manage as a console_script https://review.opendev.org/c/openstack/openstack-ansible/+/945032 | 11:45 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Symlink console_scripts to /usr/local/bin/ https://review.opendev.org/c/openstack/openstack-ansible/+/945033 | 11:45 |
noonedeadpunk | jrosser: don't you accidentally have any better idea for naming scripts? https://review.opendev.org/c/openstack/openstack-ansible/+/945032/4/setup.cfg#32 | 11:52 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: [doc] Update docs on how to use inventory scripts https://review.opendev.org/c/openstack/openstack-ansible/+/946030 | 11:59 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Stop skipping yaml-related rules https://review.opendev.org/c/openstack/openstack-ansible/+/942686 | 12:59 |
noonedeadpunk | can we also take a look at https://review.opendev.org/q/topic:%22simplify_bootstrap%22 ? | 13:00 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use FQCN for modules and fix YAML issues https://review.opendev.org/c/openstack/openstack-ansible/+/946042 | 14:04 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use FQCN for modules and fix YAML issues https://review.opendev.org/c/openstack/openstack-ansible/+/946042 | 14:06 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Stop skipping FQCN ansible-lint rule https://review.opendev.org/c/openstack/openstack-ansible/+/946043 | 14:08 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Stop skipping FQCN ansible-lint rule https://review.opendev.org/c/openstack/openstack-ansible/+/946043 | 14:45 |
noonedeadpunk | #startmeeting openstack_ansible_meeting | 15:01 |
opendevmeet | Meeting started Tue Apr 1 15:01:10 2025 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:01 |
opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:01 |
opendevmeet | The meeting name has been set to 'openstack_ansible_meeting' | 15:01 |
noonedeadpunk | #topic rollcall | 15:01 |
noonedeadpunk | o/ | 15:03 |
NeilHanlon | o/ | 15:03 |
noonedeadpunk | #topic office hours | 15:05 |
noonedeadpunk | so I have secured a room on PTG for us on Apr 9 from 15 to 17 UTC | 15:06 |
noonedeadpunk | I have not prepared etherpad with agenda though.... | 15:06 |
noonedeadpunk | #link https://etherpad.opendev.org/p/osa-flamingo-ptg | 15:07 |
NeilHanlon | woo! | 15:09 |
NeilHanlon | thank you for doing that | 15:09 |
noonedeadpunk | I will fill it with the content later today and will send a ML | 15:09 |
noonedeadpunk | I failed to look into EL10 so far, unfortunatelly | 15:12 |
noonedeadpunk | though some small progress in polsihing/finalizing things | 15:12 |
NeilHanlon | same, honestly | 15:15 |
noonedeadpunk | #endmeeting | 16:08 |
opendevmeet | Meeting ended Tue Apr 1 16:08:01 2025 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 16:08 |
opendevmeet | Minutes: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-01-15.01.html | 16:08 |
opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-01-15.01.txt | 16:08 |
opendevmeet | Log: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-01-15.01.log.html | 16:08 |
noonedeadpunk | I got completely distracted during the meeting, but basically - given that coordinate release is tomorrow - we need to do some beta at very least soonish | 17:06 |
noonedeadpunk | and basically - do feature freeze | 17:06 |
noonedeadpunk | mgariepy: NeilHanlon jrosser can you kindly look at https://review.opendev.org/c/openstack/openstack-ansible/+/945115 and all their dependencies so that I could prepare role freeze as well? | 17:08 |
noonedeadpunk | damiandabrowski: ^ :) | 17:16 |
noonedeadpunk | there're quite some patches if count for dependencies | 17:20 |
noonedeadpunk | https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/945089 is not so obvious one :) | 17:21 |
noonedeadpunk | thanks!!! | 17:33 |
mgariepy | hmm this one i guess will need 945089 merged first ? https://review.opendev.org/c/openstack/openstack-ansible/+/945090 | 17:34 |
noonedeadpunk | yeah | 17:38 |
noonedeadpunk | and plugins as well | 17:39 |
noonedeadpunk | but I can take care of rechecks :) | 17:39 |
mgariepy | will rabbitmq repos stop being a mess at some points? | 17:40 |
noonedeadpunk | eh... | 17:40 |
noonedeadpunk | I wonder if it's accidental or on purpose | 17:40 |
noonedeadpunk | or just lack of funding | 17:41 |
mgariepy | backlog upgrading, and that's the one part that is not fun haha | 17:41 |
mgariepy | how recent version are you guys running with lxb ? | 17:41 |
noonedeadpunk | I don't run lxb newer then 2023.1 tbh | 17:42 |
noonedeadpunk | ovs/ovn is rest | 17:42 |
mgariepy | ok | 17:42 |
mgariepy | i do have all 3 also. | 17:42 |
mgariepy | but i wonder how far i can go with lxb | 17:42 |
mgariepy | 2023 on 22.04 with lxb i guess? | 17:43 |
noonedeadpunk | it should be "fine" untile 2024.2 including | 17:43 |
noonedeadpunk | lxb is dropped in 2025.1 | 17:43 |
mgariepy | ok | 17:43 |
noonedeadpunk | we do have some ci tests for neutron with lxb | 17:43 |
mgariepy | do you have a playbook to migration lxb to ovn in a 4 minutes downtime ? :) | 17:44 |
mgariepy | haha | 17:44 |
noonedeadpunk | lol, no, not yet | 17:45 |
noonedeadpunk | frankly lxb footprint is super limited for me atm | 17:45 |
noonedeadpunk | so unlikley folks will invest smth into such migration... | 17:46 |
noonedeadpunk | unless there will be ready-to-go solution | 17:46 |
mgariepy | yeah | 17:47 |
mgariepy | i do only have 1 cloud with lxb and it's running on top of vlan mostly, so i guess migration beside that the stuff aroud FIP and so on should be quite easy. | 17:47 |
noonedeadpunk | yeah, vlan is likely relatively easy | 17:48 |
mgariepy | vxlan > geneve would be something else. | 17:48 |
jrosser | mgariepy: we still have lxb on Dalmatian | 18:22 |
* jrosser nightmare migration….. | 18:22 | |
mgariepy | i can imagine | 18:22 |
mgariepy | noonedeadpunk, how were you mitigating slurp upgrade before zed ? did you update the nova serviec version in db direclty ? | 18:57 |
noonedeadpunk | I did yes | 19:11 |
noonedeadpunk | But Yoga -> 2023.1 is seamless | 19:11 |
mgariepy | was only nova the issue i guess ? | 19:12 |
mgariepy | i was planning to do : xena > zed > 2023.1 (for reasons) but i'll do yoga instead of Zed then :) haha | 19:14 |
noonedeadpunk | I did Xena -> 2023.1 fwiw | 19:16 |
noonedeadpunk | and yes, bumping nova rpc version in db was needed | 19:16 |
mgariepy | well. i'll jump through yoga. | 19:17 |
mgariepy | since it's not that long to do. | 19:17 |
opendevreview | Merged openstack/openstack-ansible-plugins master: Move remaining playbooks to collection https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/945089 | 19:26 |
opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Freeze roles for 31.0.0.0b1 release https://review.opendev.org/c/openstack/openstack-ansible/+/946083 | 19:49 |
opendevreview | Merged openstack/openstack-ansible master: Use first utility host as default host for openstack_resources https://review.opendev.org/c/openstack/openstack-ansible/+/945085 | 19:57 |
opendevreview | Merged openstack/openstack-ansible master: Move remaining playbooks to collection https://review.opendev.org/c/openstack/openstack-ansible/+/945090 | 23:57 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!