=== mup_ is now known as mup
[06:37] <kjackal> Good morinig Juju world!
[07:37] <Naz> Good morning, I have 2 questions please :) , 1- How could I move a service (Application/unit/machine) from a hosting model A to a model B?          2- Could I deploy the same application twice? (I mean, using the SAME charm, I want to run two machines serving the same application but using different configuration]).
[07:47] <Naz> for question 2, obviously, I am talking about deployment of same app twice to SAME model.
[08:00] <zeestrat> Naz: Not a juju dev but, 1: I think Juju is currently limited to migrating model migrations, see: https://jujucharms.com/docs/2.1/models-migrate, rick_h might know if migrating individual services is in the pipeline
[08:04] <zeestrat> Naz: 2: To deploy the same application but with different names and configurations, use "juju deploy <charm> <name-of-application>", so "juju deploy mysql different-mysql" for example
[08:12] <Naz> @Zeestrat, Big than you, much appreciated. However, Probably, I didnt explain well, for 1, I am not talking about a model as a whole migrating from a controller A to controller B, I need to migrate a machine and its service from a Model A to a Model B. For 2, Good tip, Thank you :)
[08:23] <zeestrat> Naz: No problemo. I'd suggest reaching out to the juju mailing list (https://lists.ubuntu.com/mailman/listinfo/juju) to ask about 1.
[08:37] <boolman> hi, what does this mean, I cant find any information about this error: machine-3: 08:36:52 ERROR juju.provisioner cannot start instance for machine "3/lxd/0": unable to setup network: host machine "3" has no available device in space(s) "internal"
[08:38] <boolman> what is it trying to do?
[08:47] <zeestrat> boolman: If think juju is trying to setup network for the LXD container on machine 3 connected to a network space called internal, but cannot find it on the machine. Are you using maas with juju? If so, what versions? What are you trying to deploy?
[08:50] <boolman> zeestrat: the machine 3 has a network interface on space 'internal' and a configure IP from that network. the space/subnet shows in 'juju spaces'
[08:50] <boolman> I'm trying to deploy openstack
[08:54] <zeestrat> boolman: Gotcha. Are you using the openstack bundle from https://jujucharms.com/openstack-base/ by any chance?
[08:58] <boolman> zeestrat: yes I believe so
[09:02] <zeestrat> boolman: Alright. I'm not sure how that bundle works per today with spaces. You might want to ask jamespage and/or folks in #openstack-charms
[09:02] <boolman> ok thx
[09:09] <boolman> think I figured it out, lxd requires the space on a bridge port, not a physical port.
=== rumble is now known as grumble
[10:16] <erik_lonroth_> kwmonroe: https://pastebin.com/aC7pmiqh  --- it seems from me debugging the install-hook, the  "bigtop" installation is failing. What happens here is yet covered in smoke for me.
[10:25] <kwmonroe> yeah erik_lonroth_, the bigtop apt repository is not reachable.  did you set http/apt proxies on the model?
[10:25] <erik_lonroth_> Yes I did. Let me run the "apt-get update" manually to see what happens. Just a sec.
[10:28] <erik_lonroth_> That fails utterly
[10:28] <kwmonroe> heh
[10:28] <erik_lonroth_> https://pastebin.com/RjHZF2SQ
[10:29] <kwmonroe> ok erik_lonroth_, you can get here, right?  http://bigtop-repos.s3.amazonaws.com/releases/1.2.0/ubuntu/16.04/x86_64/dists/bigtop/contrib/binary-amd64/Packages
[10:35] <kwmonroe> i'm gonna assume you can get there, but you can't wget that url from a juju deployed unit.  that reeks of a proxy issue.  does your network require you to whitelist unknown domains?
[10:39] <kwmonroe> also, run "juju model-config | grep proxy" just to make double sure your proxy settings look right
[10:48] <erik_lonroth_> It is proxy issues and I've added them to our whitelist now and trying again with a clean model
[10:57] <cnf> so in juju, with maas 2.2, how can you filter on subnets now that spaces are vlan based, instead of subnet based?
[11:10] <jam> cnf: you would still have subnets in your space, wouldn't you? and "how can you filter" in what UI? MaaS UI? 'juju status' ?
[11:10] <cnf> jam: constraints
[11:11] <cnf> and i have vlans with multiple subnets, where i want to specifiy juju needs to use a specific subnet
[11:12] <cnf> in 2.1 i can use maas spaces, but with 2.2 they moved spaces to vlans
[11:12] <cnf> so that doesn't work anymore
[11:57] <boolman> does juju/maas create a local user with password? I can't deploy on my nodes and I want to login from console
[11:58] <erik_lonroth_> :kwmoroe It "almost" worked this time. The state I'm now in for some good 20 minutes is for spark/0    (juju status -> "spark/0*                   maintenance  idle   4        10.54.83.216                        configuring spark in yarn client mode"
[11:58] <erik_lonroth_> spark/0 never leave this state for me....
[11:58] <rick_h> boolman: it should setup the ubuntu user with ssh access from the ssh key in MAAS or your local machine it was bootstrapped from
[12:02] <boolman> rick_h: no I dont think it's installed my ssh key
[12:02] <boolman> certainly cannot login
[12:05] <jam> boolman: ssh ubuntu@X.Y.Z ?
[12:05] <cnf> boolman: maas adds your ssh keys
[12:05] <cnf> boolman: did you add your ssh keys to maas?
[13:05] <boolman> cnf: yes I've added my key to maas, I can ssh into other nodes. but not the 3 I'm trying to figure out why the deployment failed
[13:06] <boolman> by the looks of it, the ubuntu install is successfull, but on firstboot when running cloud-init it fails to find the maas datasource or something
[13:06] <boolman> thats why I need to login
[13:07] <admcleod> boolman: is it possible you have an MTU mismatch somewhere?
[13:07] <boolman> not likely, but I can double check
[13:08] <admcleod> boolman: might be worth it, can cause problems with cloud-init (not so much "cant find" though)
[13:13] <boolman> admcleod: hmm ok I do see some strange stuff in the config on this crappy switch
[13:14] <boolman> hm no that was 'system mtu'
=== scuttle|afk is now known as scuttlemonkey
[13:45] <vlad_> Hey guys is there a 5th follow up post for this series of blog posts? (Here's the 4th and last one I could find): https://insights.ubuntu.com/2016/01/31/nodes-networking-deploying-openstack-on-maas-1-9-with-juju/
[13:47] <rick_h> vlad_: check out http://blog.naydenov.net/
[13:49] <vlad_> rick_h: And to everyone else on here. This is the best IRC ever!
[13:55] <erik_lonroth_> =)
[13:59] <boolman> admcleod: No the mtu settings should be fine. however untill I can login to the node, I cant be sure
[14:36] <kwmonroe> erik_lonroth_: any luck on spark/0 coming out of "maintenance"?  if not, can you paste a "juju status && juju show-status-log spark/0 -n 100"?
=== frankban|afk is now known as frankban
[15:10] <kwmonroe> hello friends!  jujubox:devel and charmbox:devel docker images have been updated with juju rc3.  happy modelling!
[16:35] <Budgie^Smore> o/ juju world :)
[16:39] <wpk> 🙌
[17:14] <lazyPower> 👋 budgie
[18:08] <thumper> wpk: some people are more capable with emoji than others, my limits lie with :-) and o/
[18:09] <cnf> 👃🏻🙈🍌
[18:16] <thumper> cnf: I can't tell if that is telling a story or just random stuff...
[18:18] <cnf> monkey smelt but didn't see the banana!
[18:18] <cnf> that, or money didn't want to see the penis, and wanted a banana
[18:18] <cnf> you pick