| opendevreview | Merged openstack/openstack-ansible-galera_server stable/xena: Bump mariadb version. https://review.opendev.org/c/openstack/openstack-ansible-galera_server/+/864052 | 05:30 |
|---|---|---|
| *** ysandeep|out is now known as ysandeep | 06:44 | |
| gokhani | hi folks, I am now trying ceph integration, but on glance side I am getting this bug https://bugs.launchpad.net/glance/+bug/1916482/. it seems uwsgi issue. how have you solved this issue ? maybe wee need some configs on haproxy or uwsgi side ? | 09:02 |
| noonedeadpunk | gokhani: `glance_use_uwsgi: false` in user_variables and re-run os-glance-install.yml | 09:10 |
| gokhani | noonedeadpunk: ok I am trying now. cinder integration is also workes. And another question, previously I used nfs as a shared storage for vms but my all vms are on cinder and at nfs side I only see intances console.log. ı want to change my vm storage to rbd. can it break if we change it directly from nfs to rbd ? | 09:17 |
| noonedeadpunk | oh yes, it will | 09:18 |
| noonedeadpunk | or well | 09:18 |
| noonedeadpunk | you meant you used nfs for nova "ephemeral" stprage by all VMs are booted from volume and cinder was on ceph from beginning? | 09:19 |
| noonedeadpunk | then answer is no:) | 09:20 |
| gokhani | partially yes, all vms are booted from volume and cinder was on fc backend not ceph. | 09:21 |
| gokhani | I will also try convert from fc to ceph | 09:22 |
| noonedeadpunk | well, that is actually fun part... | 09:28 |
| gokhani | noonedeadpunk: glance has worked :) | 09:28 |
| * noonedeadpunk are right now in progress of migration from NFS to Ceph for cinder volumes | 09:28 | |
| gokhani | noonedeadpunk: ok good luck noonedeadpunk and me too :) | 09:33 |
| *** ysandeep is now known as ysandeep|lunch | 09:35 | |
| gokhani | noonedeadpunk: is it possible to use rbd for ephemeral vms for newly added compute nodes ? ı think it is possible and I want to be sure :) | 09:36 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Fix inspector auth config for ironic-api https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864395 | 09:47 |
| noonedeadpunk | um, yes, and I think you can indeed jsut switch existing ones given all are bfv now | 10:01 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/wallaby: Set Wallaby to stable branch after EM-ing https://review.opendev.org/c/openstack/openstack-ansible/+/864268 | 10:06 |
| jrosser | looks like our ironic+swift CI doesnt work due to SSL troubles | 10:07 |
| jrosser | i wonder if we should re-do that so swift is not involved | 10:12 |
| noonedeadpunk | haven't look into that... Shouldn't we sort out ssl stuf regardless? | 10:18 |
| jrosser | we should | 10:19 |
| noonedeadpunk | tbh every time I looked into ironic/swift and was thinking to decouple that, I ended with tought that this should be fixed anyway | 10:20 |
| noonedeadpunk | and that integration was looking as some real test for swift. But yep, eventually we should just cover swift with better testing rather then test swift with ironic... | 10:22 |
| jrosser | i still am wtf at os_ironic | 10:23 |
| jrosser | just trying to make enough patches so that it at least works | 10:23 |
| jrosser | but really, needs more fundamental rewrite | 10:24 |
| noonedeadpunk | btw, I was looking at /healtcheck uri and wondering why we don't use it... | 10:26 |
| noonedeadpunk | disregard that actually :) I will try to check it in more details and propose smth I guess | 10:27 |
| noonedeadpunk | regarding os_ironic, well... I can't put much cycles into it now, as even don't have hardware to play with at the moment :( | 10:29 |
| jrosser | trouble with os_ironic is that is was developed with is_metal=true, and then consideration of containers added as an afterthought | 10:41 |
| *** ysandeep|lunch is now known as ysandeep | 10:52 | |
| *** frenzy_friday is now known as frenzy_friday|doc_appt | 11:05 | |
| *** dviroel_ is now known as dviroel | 11:15 | |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Remove duplicate creation of nginx config directory https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864403 | 11:26 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Ensure nginx config is present for ironic inspector https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864406 | 12:41 |
| nixbuilder | Need a little clarification... in https://docs.openstack.org/openstack-ansible-os_neutron/latest/app-openvswitch-dvr.html#openstack-ansible-user-variables it says that "Bridges specified here will be created automatically"... so I am assuming the bridge being referenced is the br-provider. And then later on before the br-vxlan configuration it says "The bridge name should correspond to a pre-created Linux bridge or OVS bridge | 13:24 |
| nixbuilder | " . So just to be sure... I do not need to create br-provider but I do need to create br-vxlan prior to installation. | 13:24 |
| gokhani | noonedeadpunk: fistly, at cinder integration it worked successfully, but after changing active active config (because my other backend doesn't support active-active ) false, cinder stucked in volume creating | 13:35 |
| gokhani | do we need to do extra things, after changing active-active config ? | 13:36 |
| nixbuilder | My confusion comes from another document (https://docs.openstack.org/openstack-ansible-os_neutron/latest/app-openvswitch.html), it shows only br-mgmt being configured ahead of time) in the /etc/network/interfaces but has the same statement that br-vxlan "The bridge name should correspond to a pre-created Linux bridge or OVS bridge" | 13:37 |
| jrosser | jamesdenton: can you help with the OVS question? ^^ | 13:55 |
| jamesdenton | Hi - yes. So, the name of the interface in the Linux kernel needs to correspond to the label given in openstack_user_config.yml - br-vxlan is common, but br-overlay is better. The IP applied to that interface will be used to fill in local_ip for the respective network agent | 14:02 |
| jamesdenton | those docs are being worked on to clear up some confusion, thanks for pointing it out | 14:02 |
| jamesdenton | that really applies to the use of a dedicated vlan/interface for overlay traffic. If you omit that network and ignore those directions, the mgmt ip of the host is used instead. | 14:04 |
| jamesdenton | for better or for worse | 14:05 |
| nixbuilder | jamesdenton: OK... thanks for the clarification. And yes, the docs are a little confusing in this regard. | 14:22 |
| *** frenzy_friday|doc_appt is now known as frenzy_friday | 14:33 | |
| *** ysandeep is now known as ysandeep|out | 15:10 | |
| noonedeadpunk | jrosser: if you have ironic somewhere handy - can you check if /healthcheck is returing 200/204 there? | 15:21 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_octavia master: Enable /healthcheck for Octavia API https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/864419 | 15:23 |
| *** dviroel is now known as dviroel|lunch | 15:28 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_barbican master: Update barbican api paste https://review.opendev.org/c/openstack/openstack-ansible-os_barbican/+/864420 | 15:42 |
| jrosser | 8-O https://review.opendev.org/c/openstack/nova/+/864418/1/doc/source/admin/virtual-gpu.rst | 15:54 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_sahara master: Update api-paste to enable healthcheck https://review.opendev.org/c/openstack/openstack-ansible-os_sahara/+/864422 | 15:55 |
| noonedeadpunk | O_O | 16:00 |
| noonedeadpunk | I kind of wonder.... How/why nova would pick that device... | 16:01 |
| noonedeadpunk | ` Mediated device creation is now deprecated in 2023.1 Antelope` | 16:01 |
| jrosser | noonedeadpunk: what release ironic for healthcheck? i have Yoga here | 16:06 |
| jrosser | which fwiw is 401 | 16:07 |
| noonedeadpunk | jrosser: and can you add override into ironic.conf with [healthcheck]/enabled=True ? | 16:09 |
| jrosser | noonedeadpunk: seems to work https://paste.opendev.org/show/bgA3MgAKWdTH9A7EV1EB/ | 16:12 |
| noonedeadpunk | nice, thanks! | 16:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_ironic master: Enable /healtcheck for Ironic API https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864423 | 16:14 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_ironic master: Enable /healtcheck for Ironic APIs https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864423 | 16:16 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use /healthcheck URI for backends https://review.opendev.org/c/openstack/openstack-ansible/+/864424 | 16:19 |
| noonedeadpunk | I have also discovered recently passive checks for backends - based on the responce codes from API. But I kind of afraid to use that for some reason... | 16:20 |
| noonedeadpunk | https://www.haproxy.com/blog/how-to-enable-health-checks-in-haproxy/#passive-health-checks | 16:21 |
| *** dviroel|lunch is now known as dviroel | 16:26 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/yoga: Bump OpenStack-Ansible Yoga https://review.opendev.org/c/openstack/openstack-ansible/+/864428 | 16:51 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/xena: Bump OpenStack-Ansible Xena https://review.opendev.org/c/openstack/openstack-ansible/+/864431 | 17:02 |
| jrosser | jamesdenton: i got inspector doing lldp detection of switchports - baremetal port creation can be completely automatic | 17:36 |
| jamesdenton | niiiiice | 17:45 |
| jrosser | just one thing to fix - LLDP returns the mac of the port, not the switch | 17:50 |
| jrosser | so i need to get the switch hostname into local_link_connection.switch_info | 17:51 |
| jamesdenton | seem to recall you mentioning something unique to Nexus code? | 17:52 |
| jamesdenton | some bug, but prob not related to this | 17:52 |
| jrosser | when you do this by hand it seems that a switch_id is mandatory on the port, but thats a mac addr | 17:55 |
| jrosser | but you can also give a vendor specific switch_info as the hostname that neutron ssh to to adjust the port vlan membership | 17:55 |
| jrosser | i just need to join the dots in inspector to auto-populate that | 17:55 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Update variables for switchport introspection during inspection https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/864437 | 18:36 |
| *** dviroel is now known as dviroel|afk | 20:01 | |
| opendevreview | Merged openstack/openstack-ansible master: Drop usage of lxc containers proxy https://review.opendev.org/c/openstack/openstack-ansible/+/861825 | 21:32 |
| opendevreview | Merged openstack/openstack-ansible-ops master: Add support for enabling ELK stack security https://review.opendev.org/c/openstack/openstack-ansible-ops/+/862873 | 22:10 |
| opendevreview | Merged openstack/openstack-ansible-ops master: Enable cluster ID monitoring setting for beats using logstash https://review.opendev.org/c/openstack/openstack-ansible-ops/+/862874 | 22:10 |
| opendevreview | Merged openstack/openstack-ansible-ops master: Set permissions on elastic data directories https://review.opendev.org/c/openstack/openstack-ansible-ops/+/862875 | 22:10 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!