[07:17] <ak_dev> thedahac, tvansteenburgh, rick_h, kjackal : if you havet not already deployed, please do try to deploy this new bundle
[07:18] <ak_dev> https://usercontent.irccloud-cdn.com/file/NXjEsqoq/bundle.yaml
[07:18] <ak_dev> you won't have to run any extra commands, just deploy :)
[07:18] <ak_dev> thanks again
[07:36] <kjackal> hi ak_dev testing it on aws now. thank you
[07:37] <ak_dev> kjackal: hey, thanks for testing it :-)
[07:37] <ak_dev> some pods dont run on these though, maybe you can debug better
[08:06] <kjackal> ak_dev: you are right containers do not start, although the bundle deployes...
[08:07] <kjackal> let me open an issue and put there some logs with what I see
[08:07] <ak_dev> kjackal: some of them start right?
[08:07] <ak_dev> yeah sure that would be great!
[08:08] <kjackal> ak_dev: yes some pods seem to start, others crashloop
[08:09] <ak_dev> kjackal: yeah cool, so thats exactly what happened with me too
[08:09] <ak_dev> great then, so it works for you too (kind of)
[08:12] <kjackal> ak_dev: any preference where I should open this issue?
[08:12] <kjackal> ovn charm layer?
[08:12] <ak_dev> kjackal: no idea, I dont exactly know where the issue is, or what exactly it is
[08:13] <ak_dev> do you think it could be something with those pods?
[08:13] <ak_dev> or with OVN?
[08:14] <kjackal> ak_dev: I do not know right now. I'll open the issue under ovn layer and we can expand to whatever component has the real problem
[08:15] <ak_dev> kubectl describe pod <pod_name> gave me something like "connection refused to the pod ip : port" for health check
[08:15] <ak_dev> kjackal: yeah, sure
[08:19] <ak_dev> kjackal: I also forgot to tell you, the open port thing works, obviously you might have noticed that now, thanks for the help you provided before :-)
[08:26] <kjackal> ak_dev: I saw that, I think the open ports you are doing is a bit aggressive. :)
[08:26] <ak_dev> kjackal: haha yeah
[08:27] <ak_dev> but i noticed once, when I deployed without all those ports open, services werent starting
[08:27] <ak_dev> maybe poor judegement from my part, will sure fix it after the more important problems get solved, i think
[08:55] <armaan> jamespage: Hello, I am running Mitaka (deployed with juju) on Trusty and now I want to upgrade to Newton. IIRC, for that I will have to first upgrade Trusty to Xenial. Could you please let me know whether apt-get dist-upgrade is enough for upgrading trusty to xenial and I don't need to upgrade charms?
[08:56] <jamespage> armaan: the trusty->xenial hop is not currently possible - there are a few nuances
[08:56] <jamespage> specifically lxc vs lxd
[08:56] <jamespage> and the fact that juju still thinks the application series is trusty
[08:57] <jamespage> that said, the charms do support both series
[08:59] <armaan> jamespage: Ok, so how could I upgrade a production Mitaka environment to Newton?
[08:59] <armaan> which is running on trusty
[08:59] <jamespage> you can't at the moment
[09:10] <armaan> jamespage: Ok, so in your opinion what should be upgrade strategy for me here?
[09:21] <jamespage> armaan: limited choice atm
[09:21] <jamespage> armaan: build a second xenial based cloud, migrate workloads
[09:21] <jamespage> or I think the juju team are doing some work on a documented process (i.e. not fully automatic) for doing in-place - but thats WIP
[09:23] <armaan> jamespage: ok and this is also what canonical does for their customers who want to upgrade from Mitaka?
[09:24] <jamespage> I think there have been a few who have taken that first route, and some who are waiting for the second option
[09:25] <jamespage> armaan: was this a juju 2.x based deployment?
[09:25] <jamespage> or is it the older 1.25.x series?
[09:25] <armaan> jamespage: older 1.25.x series?
[09:25] <jamespage> armaan: which version of juju are you using?
[09:26] <armaan> jamespage: juju 1.25
[09:30] <jamespage> armaan: again there is not currently an upgrade process from 1.25.x to 2.x.x
[09:30] <jamespage> that's part of the work the juju team have on
[09:35] <armaan> jamespage: ok, thanks for letting me know.
[15:18] <kjackal> ak_dev, the kubernetes team appointed some time in this two week sprint to support the work you are doing on ovn. Idealy we would like to have your bundle fully functional at the end of this two week sprint. After that so we can distil the tasks needed to have your bundle and cdk converge.
[18:41] <ak_dev> kjackal: hey! sorry was out for a while, that great! who do I contact for help from the team?
[18:42] <ak_dev> Yeah, we will try and complete this at the end of the further two weeks
[18:42] <ak_dev> thanks for helping out :-)
[19:24] <kjackal> ak_dev: EU time its me and US time you can ping Cynerva or ryebot, but we show a preference towards opening issues
[19:25] <ak_dev> kjackal: oh okay, cool, I will open an issue whenever I feel the need to
[19:33] <kjackal> thanks ak_dev
[19:33] <ak_dev> kjackal: no, thank you :-)