/srv/irclogs.ubuntu.com/2018/06/21/#juju.txt

wallyworldvino: the bzr command is "bzr revision-info"00:06
kelvinwallyworld, do u think we should define the device constrains at `juju/juju/device` or somewhere else?00:08
wallyworldhmmm. i think "devices" is ok for now00:10
kelvinin the root dir `/devices`?00:11
vinook.00:16
vinowallyworld : and for mercurial hg equivalent is hd id --id00:19
wallyworldgreat00:19
wallyworldkelvin: yeah, i think so for now00:19
kelvinwallyworld, great, thx00:20
thumperwallyworld, kelvin: how about core/devices?01:35
thumperi had wanted to move the /network into core/network01:35
thumperas long has it has no state or api dependencies01:35
kelvincore/devices looks a good one,01:36
kelvinyes, then we can shorter the root dir ls01:37
wallyworldthumper: we should move storage as well then02:13
thumperwallyworld: yes02:13
wallyworldanastasiamac: if you have a moment? https://github.com/juju/names/pull/8903:05
wallyworldveebers: want to talk about resources? thumper has stodd me up for our 1:1 :-)03:11
veeberswallyworld: hah, sure thing03:11
thumperwallyworld: we did have a long catch up this morning :)03:11
wallyworldwe did. but you could have told me :-)03:12
=== mup_ is now known as mup
thumperwallyworld: sorry, was planning to but got stuck talking to xavpaice03:14
kelvinwallyworld, could I have ur a few minutes?03:25
kelvinthinking if it's reasonable to consolidate the struct (params struct) at api and apiserver in single place.03:36
kelvinto have the api schema shared for client, server side, or other dependencies.03:39
wallyworldkelvin: sure, just finished talking to chris, free now03:44
kelvinc u in hangout03:45
kelvinnow?03:45
babbageclunkthumper: will the hub drop messages if the handler takes too long?04:08
babbageclunk(I mean, if a subscribed handler takes too long)04:09
anastasiamacwallyworld: reviewed..04:46
wallyworldty04:46
wallyworldanastasiamac: i pushed a change to use a common method04:52
kelvinwallyworld, could u take a look the PR again when u have time? thx05:06
vinowallyworld: have a min ?06:35
wallyworldsure06:35
vinoHO06:35
wallyworldam in standup06:36
=== frankban|afk is now known as frankban
thumperbabbageclunk: no07:54
babbageclunkgood07:54
thumperdid you want to chat about it?07:54
* thumper is back in the office for a bit07:55
babbageclunkno, it's fine07:55
babbageclunkthanks!07:55
thumperjam: are you back around?07:55
jamthumper: hey, just got back, quite the adventure.. are you still here?09:04
thumperum... was just about to leave, but can chat09:10
thumperjam: did you want to catch up?09:11
jamthumper: sure. Just chatting with manadart, will be available in 2min.09:15
thumperkk09:16
manadartNeed a review for: https://github.com/juju/juju/pull/883809:29
stickupkidmanadart: i'll take a look09:33
stickupkidmanadart: just out of curiosity, but jam will probably want to vet the whole thing (?)09:34
manadartstickupkid: Assume so.09:38
stickupkidmanadart: done - just a few questions10:06
manadartstickupkid: Shoot.10:06
stickupkidmanadart: questions in the PR10:23
manadartstickupkid: Ta.10:28
naturalbluehi. I am trying to deploy a juju maas controller and it starts to deploy but i notice later on it looks for an IP (that was MAAS's old IP address) and then fails to deploy. Is there a way to get the details of what it is looking for. I have changed all ips and restarted all services on MAAS. I am not sure why it is still looking for the old address.11:05
BlackDexnaturalblue: Did you re-deploy juju?11:07
BlackDexthe controller?11:07
BlackDexIf not, i think you need to reconfigure your cloud providers in juju11:08
naturalblueah that maybe it. i think although i removed the cloud provider after the controller, i possible changed its ip after i added it back in.11:14
naturalblueI will give it atry and let you know.11:14
naturalblueBlackDex: Thansk11:14
naturalblueno, i just checked with juju show-cloud openstack-maas and the endpoing shows the correct ip (i.e. the new IP)11:15
naturalbluebut cloud-init is definitely looking for the old address and then aborting11:17
BlackDexduring boot?11:17
BlackDexpxe boot?11:17
BlackDexPlease check the if the IP's are correct during the following command: `sudo dpkg-reconfigure maas-region-controller maas-rack-controller`11:19
naturalblueduring pxe boot, after i ask to bootstrap a new controller, maas rolls the OS, then juju starts the deploy of the controller. Watching the console its installing but suddenly i see a cloud-init script that references the old ip address of maas and the system shutsdown11:21
naturalbluethe ip in `sudo dpkg-reconfigure maas-region-controller maas-rack-controller` was incorrect, it was the old one11:22
naturalbluei didnt realise i had to run that after chaning the ip from the maas web ui and rebooting11:23
naturalblueits asking for the api address now on the dpkg-reconfigure. Do i put in thr port of 5240 or is that just for the web gui11:25
BlackDex524011:56
BlackDexthat is the API port11:56
BlackDexport 80 wil be deprecated in 2.4 or 2.5 if i'm correct11:56
BlackDexyea, you can change those in a file somewhere, but this is better, and it restart the services needed after :)11:57
rick_h_naturalblue: it might be in the cache for the cloud in .local/share/juju/clouds.yaml12:02
naturalblueBlackDex: rick_h_ : Thanks12:14
manadartstickupkid: Are you in any position to push your changes that further strangle out rawProvider?12:15
manadartstickupkid: I am having rewrite provider tests and it makes no sense to be stubbing/mocking out changes to something we are ditching.12:16
magicaltroutanyone know why if i try and deploy to a machine into an lxd container12:17
magicaltroutit sets my resolv.conf seemingly incorrectly and the charm fails to apt update?12:17
rick_h_magicaltrout: bionic?12:17
rick_h_magicaltrout: there was a bug around that for bionic I know that was worked on and a fix landed12:17
magicaltroutthe controller is on bionic the machines are xenial12:18
rick_h_magicaltrout: https://bugs.launchpad.net/juju/+bug/176431712:18
mupBug #1764317: bionic LXD containers on bionic hosts get incorrect /etc/resolve.conf files <bionic> <cdo-qa> <cdo-qa-blocker> <foundations-engine> <kvm> <lxd> <network> <uosci> <juju:Fix Committed by ecjones> <juju 2.3:Fix Released by ecjones> <https://launchpad.net/bugs/1764317>12:18
enrico_Hello, I am trying to deploy a service with juju. I want to deploy it inside an lxd container. Till now ok. But I want to add 2 NIC, how can I do?12:29
stickupkidmanadart: i can do, but it's broken a few things... let me just fix what i've got so you can at least see what i've done12:43
stickupkidmanadart: this is a very much work in progress - https://github.com/juju/juju/compare/develop...SimonRichardson:lxd-schema?expand=112:56
stickupkidmanadart: i'm in the middle of refactoring the tests, before I get to the construction of the provider - although I may change my mind whilst I'm writing the tests12:56
manadartstickupkid: Ta.13:06
=== zeus is now known as Guest21990
=== Guest21990 is now known as zeus
naturalbluewhen i try to install juju gui with juju upgrade-gui i get this "ERROR cannot upgrade to most recent release: cannot retrieve Juju GUI archive info: error fetching simplestreams metadata: invalid URL "https://streams.canonical.com/juju/gui/streams/v1/index.sjson" not found"14:24
naturalblue^^but i have tried a wget and can retrieve the file from both the client and the controller14:25
naturalblue^^ please ignore this. turns out i had a proxy variable in /etc/environment which get reapplied on a reboot. removed now and variables unset.14:30
Guest53Hi I just booted three ceph osd that were shutdown.  One came up fine but the other two say they are stuck in upgrading.  The log message I am getting is DEBUG config-changed Error ENOENT: error obtaining osd_hostname_luminous_start.15:34
Guest53Any idea where this luminous start file is located?15:34
=== Guest53 is now known as acss
acssAnyone know why I would see the following in a juju ceph-osd unit log? 2018-06-21 16:03:06 INFO juju-log Monitor config-key get failed with message: b''16:07
rick_h_acss: hmm, some sort of unexpected config value the charm couldn't parse?16:07
acssThis would be based on my ceph.conf file?16:08
acssI am confused because I have another OSD functioning fine with the same charm configuration.16:30
=== frankban is now known as frankban|afk
rick_h_acss: sorry, ran for lunch stuff. So this is the juju config16:43
rick_h_acss: so you can see what you've got with juju config ceph-osd16:43
acssokay16:43
acsswhere is this located?16:44
rick_h_acss: so it's the config juju is told to send to the charms and changes to that config triggers the charm's config-changed event which then causes it to evaluate things and update any local files/apis/settings that need updating for the software in question16:47
acssokay thank you16:47
=== Guest16456 is now known as jose
sfeolehello all... we have bootstrapped a controller out on AWS cloud and deployed rabbitmq, juju status displays the internal cloud IP address.   However, i want to query rabbitmq via the assigned floating ip address, will juju correctly translate/route  floatingip -> internal cloud ip?21:23
rick_h_sfeole: so you need to expose the rabbit unit for the floating IP to be the one to work since by default the firewall won't allow any traffic to it21:32
sfeolepatriciadomin, ^^21:32
sfeolerick_h_, so just,  juju expose <unit>   ?21:33
rick_h_sfeole: juju expose rabbitmq or whatever the application name is21:33
rick_h_sfeole: that'll set the firewall rules as long as the charm is setup to expose the rabbit port21:33
sfeolerick_h_, ok thanks,21:33
sfeolepatriciadomin, ^^21:33
patriciadominrick_h_ sfeole thx21:36

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!