[13:56] <magicaltrout> is it time for beer yet?
[14:34] <rick_h> magicaltrout: wfm
[15:36] <D4RKS1D3> Hi, I am looking for some documentation of the scheduler in Juju, thanks
[15:48] <rick_h> D4RKS1D3: what do you mean by the scheduler?
[15:49] <admcleod_> thumper: what room are you guys in?
[15:50] <thumper> admcleod_: 1409
[17:08] <rick_h> I need a juju show theme song. Something to get everyone pumped up ahead of things
[17:08] <rick_h> dum dum dum da da dum dum dum
[17:09] <rick_h> Juju show in 51minutes reminder wheeee
[17:49] <zeestrat> rick_h: Can't go wrong with some 80's glam metal: https://www.youtube.com/watch?v=CmXWkMlKFkI
[17:50]  * rick_h chants "go big hair go big hair!" and clicks link
[17:51] <rick_h> zeestrat: lol, there you go. I want to hear screaming fans before I click that "go live" button
[17:52] <zeestrat> rick_h: Question for y'all here or on the show. What's the status of the juju snap channels. I'm a bit confused as I thought it went from edge>beta>candidate>stable, however right now stable and beta has 2.2.4 while candidate has 2.2.5+2.2-d75e780
[17:53] <rick_h> zeestrat: oh hmmm...well that sounds like an oops skipping a channel there.
[17:53] <rick_h> balloons: ^
[17:55] <rick_h> thumper: juju show in 5 if any sprinters want to step aside and join in the fun
[17:55] <balloons> zeestrat, yes, that sounds about right. We're prepping to release a beta of 2.3 that will soon hit that channel
[17:56] <balloons> so for the moment, yea, it's a little odd that beta is technically behind, but we don't want to ship any further 2.2 series into that channel. So it's just been matching the stable channel
[17:56] <rick_h> https://hangouts.google.com/hangouts/_/yckznpagk5e4ladwik7ah4yvk4e to join the party live and https://www.youtube.com/watch?v=_yMx129uhYc to watch
[17:56] <zeestrat> balloons: Gotcha. Thanks for the info!
[17:56] <rick_h> balloons: but there's 2.2.4 and 2.2.5?
[17:57] <balloons> rick_h, there's no intention for a 2.2.5, but if one where to come, that's what it would be called :-)
[17:57] <rick_h> balloons: yea, I was expecting that to be the same as stable tbh
[17:59] <balloons> rick_h, I thought it made sense to push 2.2.5 into it for folks who might want to try it. zeestrat, rick_h, essentially we've gone with edge == tip of development, beta == "stable" development with a published agent, candidate == "tip of latest stable", stable == latest stable
[17:59] <rick_h> balloons: oic, I wasn't aware of that ok.
[18:00] <rick_h> bdx: you joining today?
[18:02] <kwmonroe> ah crud rick_h, me chromiums burned down
[18:02] <kwmonroe> restarting, biab
[18:03] <kwmonroe> now to reopen 74 tabs :/
[18:03] <rick_h> doh! wrecking crew through kwmonroe's house
[18:03] <Keenlovel> https://www.youtube.com/watch?v=_yMx129uhYc Is this a repeat ?
[18:03] <zeestrat> balloons: Are there any docs highlighting that on jujucharms.com? I just was just a bit jarring as I've been reading charm store docs where I'm interpreting those channels to be a bit differently (mind you I understand that charm store and juju core are two entirely separate things :) ) https://jujucharms.com/docs/2.1/charms-deploying#channels
[18:03] <kwmonroe> oh noes rick_h!  i'm no longer allowed to join the vid.. did you guys go live?
[18:03] <rick_h> Keenlovel: not yet
[18:04] <rick_h> kwmonroe: so I kicked you thinking you'd rejoin
[18:04] <rick_h> Keenlovel: we're about to go live once kwmonroe gets his puter to behave
[18:04] <kwmonroe> i'm back!
[18:05] <magicaltrout> jesus
[18:05] <balloons> zeestrat, they should be mirrrored, and they page you linked has it correct
[18:06] <magicaltrout> i should dial in and walk into the charm engineering room
[18:06] <balloons> zeestrat, I was just attempting to stretch the idea of candidate containing a release candidate -- technically it's not yet an rc since we don't do rc's for minor point releases
[18:07] <balloons> zeestrat, and there are currently no betas for juju, that's why it's the same as stable
[18:07] <zeestrat> kwmonroe: Yeah, I was wondering what was up the PPA's too
[18:07] <balloons> would it make more sense to you to have beta also contain 2.2.5?
[18:10] <kwmonroe> +1 balloons.  if the devel ppa is to live on, i think it should at least be at the same level of the stable ppa.  and devel would get 2.2.5 before the stable ppa.
[18:11] <Keenlovel> rick_h: check the question for marco c      - in the youtube chat windows - ty
[18:11] <balloons> kwmonroe, devel won't get 2.2.5 before stable
[18:12] <balloons> kwmonroe, that's kind of the distinction I was making
[18:14] <kwmonroe> gotcha balloons -- what's the point of the devel ppa if it doesn't get the next release before stable?
[18:14] <balloons> kwmonroe, it will get the next release before stable -- we just don't have any stabilized version of it yet
[18:15] <kwmonroe> balloons!  you just said "kwmonroe, devel won't get 2.2.5 before stable", followed by "kwmonroe, it will get the next release before stable".  color me conflicted.
[18:16] <balloons> kwmonroe, basically any X.Y.Z to X.Y.Z+1 version goes into the stable ppa directly. We don't have rc ppa's
[18:16] <balloons> kwmonroe, that said X.Y to X.Y+1 will go into devel first
[18:17] <thelaugher> rick_h: the new launchpad group (for jujucharm & Marco Ce ) is here https://is.gd/2YReLj - Please contact him on the mailing list to join this group as I am on mobile (please just email him for myself, thankyou)
[18:17] <zeestrat> Just from the side regarding candidate channel, I expect it to be what I see in stable next as stated in [1], "candidate: used to vet uploads that should require no further code changes before moving to stable." so I was a bit confused when you say there will most likely not be a 2.2.5 release. [1]: https://snapcraft.io/docs/reference/channels#risk-level
[18:18] <zeestrat> But if it's documented anywhere, that helps
[18:18] <kwmonroe> ok balloons, i think i get it :)
[18:18] <balloons> kwmonroe, basically 2.2 is stable, 2.3 is unstable. There are no released versions of 2.3 (yet). New versions of 2.2 are minor point releases and are just released
[18:19] <kwmonroe> yeah balloons - i got ya.  it took seeing you write X.Y+1 for it to click for me
[18:19] <balloons> kwmonroe, feedback welcome. I don't think you are the first to be confused. It's clear as mud for me
[18:21] <balloons> zeestrat, I JUST tried the idea out with 2.2.5; so it's interesting you've noticed it already
[18:22] <zeestrat> balloons: In the process of setting up some automated testing of my charms with travisci targeting different juju snap channels so that's why :)
[18:27] <balloons> zeestrat, perhaps it's interesting that we would put something in rc we wouldn't end up releasing. It's certainly possible we do a 2.2.5, but it was planned as the last in the series before 2.3. However, in case we release it, having folks on the channel running it and talking about it is useful. We put the rc's for 2.2.3, 2.2.4 in there as well fo
[18:27] <balloons> r the same reason. The interesting thing is now with 2.2.5, it may not see a release if 2.3 arrives as expected
[18:28] <balloons> but for example, 2.2.4 wasn't as planned persay as 2.2.3
[18:32] <SolutionJUJU> rick_h QUESTION: Where is the bug-list for the mysql jujucharm located at ?
[18:34] <zeestrat> rick_h: Question: I know you're working on those magical docs with all tips and tricks for running juju in prod. Is there any work being done on the charm developer getting started guide? It could really use a bit of polish and a fully fledged example project that is kept up to date with an opinionated selection of the current best practice tooling (charmhelpers, automated functional testing with bundletester, amulet,
[18:34] <zeestrat> matrix or what not). It still mentions the charm review queue for example.
[18:35] <kwmonroe> https://github.com/marcoceppi/charm-mysql/issues for mysql bugs
[18:36] <zeestrat> rick_h: Your screen is not in focus.
[18:37] <zeestrat> kwmonroe: Can you shout at Rick ^
[18:38] <zeestrat> Haha
[18:46] <rick_h> zeestrat: sorry, for your question there on the docs. They're always getting some love. I don't know where the updates on the new review queue process and such are at atm. Will check it out.
[18:46] <rick_h> and sorry for the screen driver fail. One of these days I'll figure out how this hangout/YouTube thing works.
[18:48] <zeestrat> rick_h: Cool. You mentioned some blogpost by the kubernetes guys in the beginning? Care to share the link?
[18:48] <rick_h> zeestrat: sure thing, it's https://insights.ubuntu.com/2017/09/27/patch-cdk-1-build-release/
[18:48] <rick_h> and I'm updating the show notes as we speak
[18:48] <kwmonroe> zeestrat: rick_h:  followup on the doc request.. i've opened https://github.com/CanonicalLtd/jujucharms.com/issues/493
[18:51] <rick_h> ty kwmonroe
[18:51] <kwmonroe> rick_h: https://www.youtube.com/watch?v=_yMx129uhYc&feature=youtu.be&t=1988
[18:51] <rick_h> oh no.../me cringes
[18:53] <zeestrat> kwmonroe: I created an issue over at juju/docs earlier (https://github.com/juju/docs/issues/2145). Not sure if that was the right place now. What's the diff between the two repos again?
[18:56] <kwmonroe> zeestrat: heh, i really don't know.  i always use the jujucharms.com repo for issues with content on the web site.
[19:16] <chamar> Hi.  Is it possible to use conjure-up on already installed VM?  Just use conjure-up to deploy and bootstrap kubernetes, for example?
[19:38] <marcoceppi> chamar: it is, if you install conjure-up on that VM and install LXD
[19:52] <chamar> marcoceppi:  Ho.  You're the one of told me to have a look at conjure-up / mass on Reddit :)  But if I conjure up on that server with LXD, it would be a local install (ie: single node cluster), right?
[19:59] <marcoceppi> chamar: hey o/ so you want to use a few VMs of already installed Ubuntu?
[20:03] <chamar> marcoceppi: Correct.  I did work perfectly fine with MAAS and a KVM backend.  It did create all the VM and bootstrap k8s.  The issue is that MAAS looks for a virtual network called either "maas" or "default".  I want my VM to use bridge adatpor instead.
[20:03] <chamar> So I was thinking of installing my VM, and then use conjure-up to bootstrap my cluster by maybe modifying the template somehow?!
[20:13] <marcoceppi> chamar: so, you /can/ do everything on a single VM
[20:13] <marcoceppi> but it'd be like, for a test cluster
[20:14] <marcoceppi> chamar: problem is, conjure-up doesn't really support this mode, but you can do it directly yourself, it's pretty straight forward I'll paste the instructions here
[20:15] <marcoceppi> cory_fu_: stokachu conjure-up doesn't support manual provider, right?
[20:16] <chamar> marcoceppi: yeah.  I do expect some tweaking and manual step.  but I really like the way conjure-up bootstrap everything ;)
[20:16] <chamar> And the web presenation were nice, too.
[20:16] <marcoceppi> chamar: yeah, it's sweeeeet, the manual steps aren't that much more  and I'll make sure to give you a good run down in like 20 mins
[20:17] <chamar> no rush.  I'll stay connected for a few hours.
[20:26] <cory_fu_> marcoceppi: stokachu is on vacation, FYI.  conjure-up does not support manual provider, but you can use the Application Configuration screen to tag units to specific MAAS nodes, which might be an alternate approach (especially if combined with kvm or lxd placement)
[20:27] <marcoceppi> cory_fu_: cheers
[20:54] <marcoceppi> chamar: I'm just about done with the blog post
[20:54] <chamar> marcoceppi: sweeeet!  Thanks a ton.  I'll give that a shot this evening in my lab and report back.
[20:55] <chamar> I saw that an issue was already open regarding the fact that you can't use bridge networking with MAAS / KVM backend (as a Pod)
[20:57] <marcoceppi> yeah, after this I'm going to do a "MAAS + KVM" blog post where MAAS is deployed using docker on your host so it's easier to set up in small labs
[21:00] <chamar> Yeah. It work fine on my lab.  But the fact that it forces to use NAT network make it's a bit useful for a Prod setup.
[21:00] <chamar> (the NATed network is "behind" your host)
[21:19] <marcoceppi> I foolishly made a 1 core 1 GB vm
[21:19] <marcoceppi> so the process is slow
[21:49] <chamar> haah! I did the same .. slowiisssshh
[22:20] <marcoceppi> chamar: would you be willing to try two vms? It's going to be really complicated with just one VM because of the whole CNI stuf
[22:21] <chamar> marcoceppi: yup..  I might try to put it further to have at least 3 etcd
[22:22] <marcoceppi> chamar: okay, cool, I'll update the blog a little bit to reflect tow machines and publish shortly
[22:22] <marcoceppi> it's easy to see how to scale up once you read through for two machines
[22:23] <chamar> Exactly what I thought... Once I understand the pattern, I should be able to scale it ;)
[22:23] <marcoceppi> perfect, and I still think once you get comfortable with this, MAAS will really help accelerate the process, but I can respect a learn to walk then run ramp up
[22:26] <chamar> Yeah.  Thanks for that. btw, just got a call and they'll shutdown the power at home (a tree falls on some electrics cable outside)
[22:26] <chamar> can you drop me a email when it's ready? (no rush, again)
[22:28] <marcoceppi> chamar: yeah, send me in a PM
[22:28] <marcoceppi> I plan on publishing tonight, if I don't publish the same day it'll be a draft forevermore