[07:31] <stickupkid> FYI: i'm updating jenkins :p
[09:26] <jam> stickupkid: our recent merge commits have started mixing line endings (\r\n vs \n). so if you do 'git log --first-parent' you'll see a bunch of \r for the description parts.
[09:26] <jam> stickupkid: is that from the plugin change? or did we break something else?
=== nmn is now known as nmnde
=== nmnde_ is now known as nmnde
[15:54] <danboid> The juju api uses port 37017 (via IP4) to communicate to its controllers right?
[15:54] <danboid> It seems 17070 (as stated in the docs) is IP6 only
[15:55] <danboid> My juju controller is only listening on 37017 for IP4
[15:57] <danboid> Should my juju controller be listening on port 17070 for IP4?
[15:58] <danboid> If so, how do I make it?
[16:07] <danboid> I expect I need to tell the juju controller to only use IPv4 do I? How do I do that?
[16:10] <danboid> Ah! Seems 37017 is just for mongo
[16:21] <danboid> I've tried disabling IPv6 in LXD but thats only had the effect of preventing the juu controller starting anything on 17070. Now I only see mongodb ports on my controller
[16:33] <babbageclunk> danboid: sorry, everyone in the juju team is at a team meetup at the moment so we haven't been watching the channel. If your controller isn't listening on 17070 then it sounds like it's not getting to the point of the API server starting up. If you look in the logfile /var/log/juju/machine-0.log, do you see any errors about api-server?
[16:35] <danboid> babbageclunk, Will do if it happens again. I ended up deleting the controller container and creating a new one but running juju bootsrap has paused at `Waiting for address`
[16:36] <danboid> Its been waiting for address for a few minutes now
[16:36] <danboid> Not a good sign I'd imagine
[16:37] <danboid> I have disabled IPv6 in LXD now
[16:37] <danboid> lxc network set lxdbr0 ipv6.address none
[16:37] <babbageclunk> danboid: if you run `lxc list` can you see the container running?
[16:38] <babbageclunk> (and an IP address?)
[16:38] <danboid> yes
[16:38] <danboid> There is a new container but it has no IP
[16:39] <babbageclunk> ah, ok - it sounds like that'll be the problem
[16:39] <danboid> Stop a delete it?
[16:39] <danboid> *and
[16:40] <babbageclunk> if you launch a lxd container with `lxc launch ubuntu:` does it start ok and get an address?
[16:41] <danboid> It created fine...
[16:41] <babbageclunk> hmm
[16:42] <danboid> no address tho
[16:42] <danboid> The host machine is using IP4
[16:42] <danboid> I created lxdbr0 with lxc init
[16:43] <danboid> Creating juu controllers was working before I disabled IPv6. Maye I should re-enable it
[16:43] <danboid> re-enable it in lxd that is
[16:44] <babbageclunk> I think juju requires it to be turned off - but I don't remember specifically how it should be configured
[16:46] <danboid> This is where I got the idea from https://github.com/juju/docs/issues/2965
[16:51] <babbageclunk> danboid: I think that's right - it's the same error juju will report if it detects ipv6 turned on on lxd
[16:52] <danboid> I don't remember seeing any such error
[16:56] <babbageclunk> It looks like we configure the bridge to turn ipv6 off if we create it (because we run lxd init ourselves)
[17:00] <babbageclunk> danboid: can you run `lxc network show lxdbr0` for me?
[17:01] <danboid> babbageclunk, It was like this:
[17:01] <danboid> babbageclunk, https://gist.github.com/danboid/6ce3fd2b0f298fc76630985c8d25d6fd
[17:02] <danboid> I have since deleted that container and its rebooting
[17:02] <danboid> What I thought I'd try is
[17:02] <danboid> lxc network set lxdbr0 ipv6.nat false
[17:03] <danboid> but it wasn't letting me, hence the reboot :)
[17:03] <babbageclunk> danboid: just looked at my config - I have ipv6.nat set to true
[17:04] <danboid> but ip6 addressing disabled? Seems a bit odd
[17:04] <babbageclunk> that's the only difference I see from your config
[17:04] <babbageclunk> yeah, I wouldn't expect it has any effect
[17:06] <babbageclunk> danboid: sorry, this isn't an area I know much about - I'm tempted to suggest uninstalling and reinstalling lxd.
[17:08] <babbageclunk> danboid: I have to go now unfortunately - sorry to leave you in the lurch!
[17:12] <danboid> babbageclunk, OK, thanks for your help!
[18:01] <danboid> Fixed it!
[18:01] <danboid> I had to add a grub kernel boot parameter on the lxd host machine to disable ipv6
[18:01] <danboid> ipv6.disable=1
[18:02] <danboid> update grub, reboot... tada!
[18:13] <danboid> How long is juju destro-controller supposed to take?
[18:13] <danboid> I'm trying to destroy some controllers that were never used
[18:14] <danboid> I deleted their conrainers first tho, maybe thats the problem?
[18:15] <danboid> I presume it would delete their containers if I'd nor already done that
[18:16] <danboid> I don't think its working
[18:17] <danboid> `juju controllers --refresh` doesn't seem to work either
[19:11] <rick_h> dannf:  yea, if the containers are already gone just use "juju unregister"
[19:11] <rick_h> sorry danboid ^
[19:11] <rick_h> who's gone...
=== exsdev0 is now known as exsdev