| *** k_mouza has quit IRC | 00:17 | |
| *** luksky has quit IRC | 00:22 | |
| *** unicell has quit IRC | 00:39 | |
| *** kplant has quit IRC | 01:30 | |
| *** kplant has joined #openstack-kolla | 01:48 | |
| *** kplant has quit IRC | 02:03 | |
| *** kplant has joined #openstack-kolla | 02:38 | |
| *** kplant has quit IRC | 02:43 | |
| *** kplant has joined #openstack-kolla | 03:35 | |
| *** kplant has quit IRC | 03:40 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Remove nova-consoleauth https://review.opendev.org/661427 | 04:54 |
|---|---|---|
| *** Sravan has joined #openstack-kolla | 05:23 | |
| *** Sravan has quit IRC | 05:28 | |
| *** unicell has joined #openstack-kolla | 05:43 | |
| *** unicell has quit IRC | 05:43 | |
| *** luksky has joined #openstack-kolla | 06:00 | |
| *** Sravan has joined #openstack-kolla | 06:28 | |
| *** Sravan has quit IRC | 06:32 | |
| openstackgerrit | Slawek Kaplonski proposed openstack/kolla stable/stein: Install iputils package in neutron containers https://review.opendev.org/661395 | 06:33 |
| *** dciabrin_ has joined #openstack-kolla | 07:09 | |
| *** dciabrin has quit IRC | 07:10 | |
| *** chason has quit IRC | 07:14 | |
| *** Sravan has joined #openstack-kolla | 07:32 | |
| *** Sravan has quit IRC | 07:36 | |
| *** Sravan has joined #openstack-kolla | 08:35 | |
| *** tonythomas has joined #openstack-kolla | 08:39 | |
| *** Sravan has quit IRC | 08:40 | |
| *** radeks__ has joined #openstack-kolla | 08:44 | |
| *** radeks_ has quit IRC | 08:47 | |
| *** Sravan has joined #openstack-kolla | 10:21 | |
| *** Sravan has quit IRC | 10:26 | |
| *** altlogbot_3 has quit IRC | 11:33 | |
| *** altlogbot_3 has joined #openstack-kolla | 11:34 | |
| *** kplant has joined #openstack-kolla | 12:30 | |
| *** dave-mccowan has joined #openstack-kolla | 13:20 | |
| *** kplant has quit IRC | 13:34 | |
| *** kplant has joined #openstack-kolla | 13:44 | |
| ivve | is there an easy way to point all of the sources to local (i.e. file) when building an image? | 13:45 |
| ivve | or does every single image need a definition? | 13:46 |
| ivve | kolla-build, for clarification :) | 14:09 |
| *** radeks__ has quit IRC | 14:14 | |
| *** goldyfruit has joined #openstack-kolla | 14:14 | |
| *** whoami-rajat has joined #openstack-kolla | 14:49 | |
| ivve | anyone able to check this out? https://review.opendev.org/#/c/661068/ | 14:51 |
| *** goldyfruit has quit IRC | 14:55 | |
| *** Jeffrey4l has quit IRC | 15:03 | |
| *** Jeffrey4l has joined #openstack-kolla | 15:04 | |
| *** jaicaa has quit IRC | 15:54 | |
| *** jaicaa has joined #openstack-kolla | 15:54 | |
| *** mgoddard has quit IRC | 15:54 | |
| *** hjensas has quit IRC | 15:55 | |
| *** mgoddard has joined #openstack-kolla | 15:55 | |
| *** altlogbot_3 has quit IRC | 15:58 | |
| *** luksky has quit IRC | 16:06 | |
| *** luksky has joined #openstack-kolla | 16:06 | |
| *** hjensas has joined #openstack-kolla | 16:08 | |
| *** altlogbot_3 has joined #openstack-kolla | 16:08 | |
| *** mgoddard has quit IRC | 16:10 | |
| *** sm806 has quit IRC | 16:19 | |
| *** sm806 has joined #openstack-kolla | 16:20 | |
| *** jbadiapa has quit IRC | 16:22 | |
| *** jbadiapa has joined #openstack-kolla | 16:25 | |
| ChewieBeardy | hello everyone! I'm studying kolla-ansible for the deployment of our openstack. POCs work great (you guys did an amazing job), but I have a few questions left I'm hoping you could help me with | 16:28 |
| ChewieBeardy | 1/ I've noticed the pypi release hasn't catched up to the stein release yet. Do you suggest we wait for the official release or can we use the master branch for our tests? | 16:30 |
| ChewieBeardy | 2/ we're also reworking our dc from the ground up (more like a junkyard actually, we're a school working with giveaway servers and switches), and we're considering a leaf/spine architecture | 16:31 |
| ChewieBeardy | basically, that implies a L3 fabric instead of L2, each rack is in its own subnets and they're routed to each other, rather than all of the nodes on the same broadcast domain | 16:32 |
| ChewieBeardy | would such a design pose problems for a deployment? | 16:32 |
| ChewieBeardy | naively, I'd assume that tenant networks should work fine with vxlan tunnels getting routed, and we can use routed provider networks for our uplink | 16:33 |
| ChewieBeardy | but I'm not so sure about the HA for the API and the network nodes, since it uses keepalived | 16:34 |
| ChewieBeardy | maybe the controller/network nodes should be kept in the same network? most talks I've seen with an L3 fabric for openstack didn't use tenant networks and directly plugged instances on the provider, so I have a hard time figuring out the correct arhchitecture for our use case | 16:36 |
| ChewieBeardy | thanks a lot for you time! | 16:36 |
| *** tolisbar has joined #openstack-kolla | 16:50 | |
| *** whoami-rajat has quit IRC | 16:58 | |
| *** luksky has quit IRC | 17:13 | |
| ivve | ChewieBeardy: pypi or master doesn't matter. for your second requires a longer answer. but the short one is, you can do it however you like. L2 or L3 in your physical network, or a mix of both. you can also solve some architectures with vxlans in the physical realm as well | 17:22 |
| ivve | a lot depends on how you want it to look like later on and if you have any requirements (specifically for external networks) | 17:23 |
| *** goldyfruit has joined #openstack-kolla | 17:38 | |
| *** dave-mccowan has quit IRC | 17:41 | |
| *** Sravan has joined #openstack-kolla | 17:59 | |
| *** Sravan has quit IRC | 18:40 | |
| openstackgerrit | Merged openstack/kolla-ansible stable/stein: Increase retries numbers of grafana check task https://review.opendev.org/655524 | 18:42 |
| *** hrw has quit IRC | 18:51 | |
| *** radeks has joined #openstack-kolla | 18:55 | |
| *** radeks has quit IRC | 18:57 | |
| *** radeks has joined #openstack-kolla | 18:57 | |
| *** Sravan has joined #openstack-kolla | 19:11 | |
| *** Sravan has quit IRC | 19:14 | |
| *** Sravan has joined #openstack-kolla | 19:15 | |
| *** Sravan has quit IRC | 19:16 | |
| *** radeks has quit IRC | 19:23 | |
| *** goldyfruit has quit IRC | 19:27 | |
| *** goldyfruit has joined #openstack-kolla | 19:32 | |
| *** kplant has quit IRC | 19:32 | |
| *** kplant has joined #openstack-kolla | 19:33 | |
| *** kplant has quit IRC | 19:38 | |
| *** itlinux has joined #openstack-kolla | 19:51 | |
| *** hrw has joined #openstack-kolla | 20:04 | |
| *** kplant has joined #openstack-kolla | 20:09 | |
| *** kplant has quit IRC | 20:14 | |
| *** kplant has joined #openstack-kolla | 22:11 | |
| *** kplant has quit IRC | 22:16 | |
| *** goldyfruit has quit IRC | 22:32 | |
| *** kplant has joined #openstack-kolla | 22:38 | |
| *** kplant has quit IRC | 23:23 | |
| ChewieBeardy | ivve: thank you for your answers! hadn't thought of tunneling in the undercloud | 23:37 |
| ChewieBeardy | I think as a starting point I'll try to run the network nodes in their own rack (and as such, subnet) and spread the rest around | 23:38 |
| ChewieBeardy | a bit of a failure domain, but it should be a good starting point! | 23:39 |
| *** goldyfruit has joined #openstack-kolla | 23:40 | |
| ChewieBeardy | external connectivity wise, the goal is to be a simple as possible: just one uplink to exit the DC, "normal" workflow is to create instances on a tenant network and route it to the "uplink" provider network, and we hide all of the internal fabric craziness as much as possible | 23:44 |
| ChewieBeardy | there's actually a lot of magic/hacks in the school network going on (turns out the higher ups have a whole /16 that they're barely using o_O), so the goal is to deploy everything in a happy place away from all the nonsense and snowpierce through it | 23:49 |
| *** kplant has joined #openstack-kolla | 23:50 | |
| *** kplant has quit IRC | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!