| *** dviroel is now known as dviroel|pto | 00:40 | |
| *** ysandeep|out is now known as ysandeep | 01:25 | |
| *** ysandeep is now known as ysandeep|afk | 04:04 | |
| *** ysandeep|afk is now known as ysandeep | 04:58 | |
| *** ysandeep is now known as ysandeep|lunch | 07:45 | |
| *** ysandeep|lunch is now known as ysandeep | 09:10 | |
| opendevreview | Merged openstack/openstack-ansible master: Fix dynamic-address-fact gathering with tags https://review.opendev.org/c/openstack/openstack-ansible/+/855977 | 09:22 |
|---|---|---|
| *** ysandeep is now known as ysandeep|afk | 10:17 | |
| opendevreview | Merged openstack/openstack-ansible master: Use proxy protocol v2 to send client info to galera backend servers https://review.opendev.org/c/openstack/openstack-ansible/+/854810 | 11:30 |
| *** ysandeep|afk is now known as ysandeep | 11:30 | |
| spatel | jamesdenton around | 14:03 |
| spatel | Is it ok to just create dummy br-vxlan interface and not attach to anything on LAN? | 14:06 |
| *** ysandeep is now known as ysandeep|out | 14:16 | |
| dokeeffe85 | Hi all, have I come to the right place to get help/support with OSA | 14:25 |
| jamesdenton | sure | 14:41 |
| jamesdenton | spatel don't see why not | 14:41 |
| jamesdenton | hi dokeeffe85 - ask away | 14:41 |
| spatel | jamesdenton thanks!! ! | 14:44 |
| spatel | We are not using vxlan so thinking to have one less VLAN in my fabric | 14:46 |
| jamesdenton | oh, gotcha. I wonder if you can just leave that out altogether, or if there is some expectation it be there | 14:46 |
| dokeeffe85 | Thanks, new here and trying to get my head around OSA. I actually ran the playbooks to completion before and I had a working cluster but made some errors and decideded to do it freshly on new machines with new Ubuntu 20.04 installs. I have a deploy host, controller and two computes. The setup-hosts was causing me an issue but that was my own fault, once that was solved I moved onto setup-infrastructure and it's failing on me in two places. | 14:47 |
| dokeeffe85 | The first fails but continues running and the second fail kills the playbook. I will send both in the order they happen | 14:47 |
| jamesdenton | yes, please use paste.openstack.org and show the errors | 14:48 |
| dokeeffe85 | Ah ok, something else I didn't know existed :) Just drop them in there with no comments? | 14:49 |
| jamesdenton | comments help, too. you'll get a unique URL you can paste here that we can look at | 14:49 |
| dokeeffe85 | Ah brilliant, thanks | 14:49 |
| dokeeffe85 | https://paste.openstack.org/show/bvQMIcLKsuuswTBfYjwv/ | 14:52 |
| jamesdenton | ahh yes, i have seen this | 14:52 |
| jamesdenton | probably easiest to "umount /var/www/repo" before re-running the playbook | 14:53 |
| jamesdenton | oh one sec | 14:53 |
| jamesdenton | try the umount, then re-run the playbook and see if it still fails | 14:54 |
| dokeeffe85 | will do, thanks | 14:56 |
| jamesdenton | which version are you deploying? | 14:56 |
| dokeeffe85 | I just pulled this git clone -b master https://github.com/openstack/openstack-ansible.git /opt/openstack-ansible and followed the instructions here: https://docs.openstack.org/project-deploy-guide/openstack-ansible/latest/deploymenthost.html | 15:00 |
| jamesdenton | cool - my only recommendation is you might want to consider stable/yoga vs master, since it would be the latest "supported" version vs one that's constantly being iterated on | 15:01 |
| dokeeffe85 | Cool, thanks. Should have got on here sooner rather than pull my hair out :) Yeah that returned the same error on the "Get list of repo packages" task. I have to leave for the day now but I can check in again tomorrow, I will do as you say with the stable/yoga repo and see if that helps. Thanks again | 15:34 |
| jamesdenton | no problem, i'm sure we can get it worked out | 15:35 |
| spatel | jamesdenton can i see vxlan tunnel status in brctl command? | 15:50 |
| jamesdenton | no | 15:55 |
| jamesdenton | there's really no "tunnel", per say, that's established (like ipsec) | 15:55 |
| andrewbonney | noonedeadpunk: I missed your mention earlier. I'll discuss internally and see if what we have might make sense to refactor into a role we could contribute | 15:56 |
| jamesdenton | it's just endpoints that communicate using vxlan headers and encap | 15:56 |
| NeilHanlon | ooh (: openinfra summit in vancouver! | 16:03 |
| jamesdenton | time to get our poutine on | 16:04 |
| NeilHanlon | i can sell it to my wife on that alone :) | 16:17 |
| mgariepy | poutine in vacouver.. pfff | 16:19 |
| mgariepy | not the right side of the country for that my friend;p | 16:19 |
| NeilHanlon | spoken like a true quebecois ;) | 16:29 |
| spatel | jamesdenton Thanks!! Reason i am asking question because this is working openstack and because of VLAN conflict i want to remove physical VLAN for VxLAN | 16:33 |
| spatel | but planning to keep vxlan bridge in place | 16:33 |
| jamesdenton | gotcha. should be fine, there's no persistent traffic or anything that would "fail" if you break that, other than vxlan networks (which you don't have) | 16:33 |
| spatel | ++1 | 16:34 |
| noonedeadpunk | andrewbonney: well, I was thinking in general, to mainly to get rid of resources management in roles. And we should also have some role for images lifecycle management | 17:09 |
| noonedeadpunk | jamesdenton: fwiw, jsut removing ports did not solve the issue | 17:12 |
| noonedeadpunk | as flows likely were stil trying to route through them | 17:12 |
| jamesdenton | did removing ports + restarting the agent to rebuild flows help? | 17:13 |
| noonedeadpunk | So I just live migrated all instances across computes to "re-wire" them which has helped | 17:13 |
| jamesdenton | ahh ok | 17:13 |
| noonedeadpunk | maybe I should have done emer-reset but was quite afraid | 17:14 |
| jamesdenton | were you able to get everything restored? | 17:14 |
| *** dviroel is now known as dviroel|out | 17:58 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!