14:00:56 #startmeeting TripleO Edge Squad Meeting 14:00:56 Meeting started Thu Feb 7 14:00:56 2019 UTC and is due to finish in 60 minutes. The chair is slagle. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:57 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:59 ping slagle, csatari, jaosorior, owalsh, fultonj, gfidente, hjensas, jtomasek, bogdando, dtantsur, rbrady, d0ugal, toure, abishop 14:01:00 The meeting name has been set to 'tripleo_edge_squad_meeting' 14:01:00 #info remove or update your nick from the Meeting Template on the etherpad if you want (or don't want) to be ping'd for the start of the meeting 14:01:03 #link https://etherpad.openstack.org/p/tripleo-edge-squad-status 14:01:05 Anyone can use the #link, #action, #help, #idea, #agreed, and #info commands, not just the moderatorǃ 14:01:13 hello. who's around? 14:01:16 o/ 14:01:17 o/ 14:01:22 o/ 14:02:23 o/ 14:02:33 #topic Agenda 14:02:35 * Review past action items 14:02:38 * Goals/action items for the week 14:02:42 #topic Review past action items 14:03:03 slagle/fultonj try multi-ceph deployments with DCN 14:03:16 sorry i didn't get to that this week 14:03:21 i didn't make any progress last week. but I'm bringing up an environment now 14:03:27 cool 14:03:51 o/ 14:03:52 fultonj: i'll let you know how I make out. right now i'm hitting some issues with network isolation just in the multistack setup 14:04:03 not related to ceph 14:04:18 slagle: so you're not using standalone; two VMs? 14:04:40 fultonj: right. i'm trying to go with the single undercloud / multi-stack approach 14:04:52 Sorin Sbarnea proposed openstack-infra/tripleo-ci master: WIP: Enable tripleo-buildcontainers for fedora 28 https://review.openstack.org/635458 14:05:13 we'll have to use standalone for the CI, but I think undercloud will be needed for the officially supported arch 14:05:29 +1 14:06:23 * slagle fultonj (et al) get split control plane ci job to not crash on podman container restart 14:06:47 still crashing on edge deployment 14:06:55 #link http://logs.openstack.org/88/615988/14/check/tripleo-ci-centos-7-split-controlplane-standalone/cd1a87f/logs/subnode-2/home/zuul/standalone_deploy.log.txt.gz 14:07:03 Sagi Shnaidman proposed openstack-infra/tripleo-ci master: Prevent second logs collection for OVB jobs https://review.openstack.org/635510 14:07:27 * fultonj got pulled into non-edge work last week 14:07:50 plan is to reproduce locally to find cause, but ran into my own env issues 14:08:11 #action fultonj (et al) get split control plane ci job to not crash on podman container restart 14:08:25 * fultonj will try again next week 14:08:35 * slagle review https://review.openstack.org/#/c/632089/ 14:08:52 quique|rover|eat, but you still need the rt repo to be there when enabling it 14:09:25 i'll revisit the docs patches this week 14:09:50 i should have some updates after getting an environment stood up 14:10:14 #action review https://review.openstack.org/#/c/631489 and https://review.openstack.org/#/c/632089 14:10:45 #topic Goals/action items for the week 14:11:12 slagle: this is done by dib 14:11:14 the other thing I wanted to mention that we should be thinking about is if what recommended architectures we want to support 14:11:16 sshnaidm: I mean 14:11:50 we know we want a ceph cluster per DCN site for local storage 14:11:58 but there may be others? 14:12:28 other storage backends? 14:12:49 yes 14:12:53 or different roles even 14:13:16 e.g., we have roles for both HCI nodes at the edge as well as separate roles for compute/ceph 14:13:19 there's ephemeral only and there's ceph ; the ceph case assumes active/active cinder which not all cinder backends support 14:13:45 Merged openstack/tripleo-heat-templates stable/rocky: Adding dependency for NetworkDeployment in 'server_resource_name'Deployment https://review.openstack.org/635450 14:14:02 right, i wasn't sure if there are others beyond that 14:15:07 just trying to cover all the bases 14:15:10 ok, anything else? 14:15:18 maybe a simple NFS backend? 14:15:48 yea something like that 14:15:53 e.g. if you have a NFS server indepdent at the edge site, we could have a way for that to work for less demanding storage workloads 14:16:21 fultonj slagle I am not sure about cinder/nfs driver active active 14:16:37 there could be bugs in the nfs driver 14:16:52 but we should be able to get things configured correctly at least 14:17:00 gfidente: sure. it's just a thought. i just want to make sure we document and provide environments/roles for the architectures that are supported 14:17:04 Kamil Sambor proposed openstack/tripleo-heat-templates master: Switch default neutron ML2 mechanism driver to OVN https://review.openstack.org/593056 14:17:05 ack +1 14:17:16 gfidente: i was thinking about cinder but about just putting instances on a provider network which an NFS filter could also live on and share out to 14:17:26 fultonj: NFS gets deprecated afaik 14:17:31 so -1 from my side 14:17:40 I mean its osp support 14:18:16 or is it glusternfs?.. whatever 14:18:18 it would be external to openstack but we can drop it if we want to stay 100% in openstack 14:18:36 somthing very simple for someone who wants to store some stuff but don't need full blown ceph 14:19:12 if not a good idea, then we can drop (but i wasn't thinking about cinder) 14:19:15 I'd tend to propose manila and cephfs instead 14:19:21 bogdando: me too 14:19:30 but for someone who doesn't want to do ceph 14:20:32 ok, thanks everyone! 14:20:35 #endmeeting