[03:00] hi, is Juju on cloud or can it be installed on my own server? [03:07] kyanh: it can be installed on your own server [03:13] hloeung, thanks a lot. I read on the site but it's hardly to figure the juju architecture :) [10:16] I have problem with setting some unit data and then accessing it on a interface-layer [10:23] http://paste.ubuntu.com/23450338/ [10:23] The ipv6_addr seems to be empty always [11:23] anrah: I wouldn't go into IPv6 territory at the moment. Is it a requirement for you? [11:27] kjackal: partly yes.. As I use chef below the Juju and I've written my cookbooks in a way that it prefers ipv6 over ipv4 [11:28] and also I have some requirements that IPv6 should be used if it's available on the platform [11:34] anrah: I know that there is work going on on IPv6 as we speak, rick_h should have more info on this [11:34] in general also I'm little confused am I able to use certain units unit data on my interfaces? [11:36] anrah: normally your reactive layer will call the functions of the interface passing any information required [11:45] kjackal: hi, I'm at some Juju roadmap planning meetings right now [11:46] kjackal: rick_h is the one who is working on making sure that ALL the IPs/subnets on an interface are made available to charm authors. [11:46] nice [11:47] anrah: I will try to get some more detail, though he is currently in a meeting [11:49] mramm: anrah is into the telco field. This might help in connecting him with the right people. [11:50] kjackal, yea definitely. [11:50] anrah: if you could share some details on what your use case is that would be helpful [11:52] anrah: could you send an email with your question to me at mark.ramm-christensen at canonical ? We would very much like to work with you to demo your stuff to various operators we are talking to at the moment, and I'd like to connect folks on that level. [11:53] I will also track down and see when the address stuff is going to land, and see if there's a good way to get what you want right now. [11:54] mramm: Sure! We have already discussed with Max and Milan about our approach [11:54] anrah: That's good news. [11:56] I was just out doing a workshop at an operator here in Europe last week, and it is always great to have more VNFs to demo at such things. [11:57] and the code you pasted above is great -- uses the latest reactive stuff, looks clean and easy to explain to high level telco audiences [11:57] very nice! [12:00] mramm: Apologies for butting in, but is there any chance that you guys will make the Juju roadmap somewhat? It would be really great to know where the project is headed and what will be focused on. [12:00] *public [12:01] zeestrat: Sure, I'll try to get the details from the team as soon as they are finalized (probably next week) pushed to the mailing list. [12:01] zeestrat: also, what we have as a roadmap is what Canonical will push -- Juju is open source so it's possible that other things will get done [12:02] mramm: That sounds great. Thanks [12:27] tvansteenburgh, morning [12:27] tvansteenburgh, https://github.com/juju-solutions/bundletester/pull/71 would help with getting the openstack-charmers charms which are currently failing cwr passing again I think [12:30] beisner, ^^ [12:40] tvansteenburgh, if you want to test it against cs:~james-page/rabbitmq-server-requirement [12:40] that has a small delta to install from test-requirements.txt [12:42] marcoceppi, do bundletests for cwr require use of virtualenv flag? [12:43] jamespage: I'm not 100% sure, ses is the one who would know best, followed by cory_fu [12:44] marcoceppi, okies - just trying to get our openstack gate story and cwr lined up so we can keep both happy :-) [13:59] jamespage, marcoceppi: reviewing now [14:32] jamespage, marcoceppi: Save for a brief window last week, CWR tests are run inside charmbox to ensure a consistent, clean environment, since the venv doesn't help with apt packages that the tests might install. [14:32] So, no, they don't require the virtualenv flag. [14:33] cory_fu: ok so we're good with just the requirements change I've proposed then [14:33] that's helpful I tink [14:34] cory_fu, https://review.openstack.org/#/c/395574/ running through UOSCI for gate test [14:35] jamespage, tvansteenburgh: Not that I think that having an explicit option is an issue, but couldn't you achieve the same effect by including '-rrequirements.txt' or '-rtest-requirements.txt' in the list of python-packages? [14:36] cory_fu, that might work [14:36] cory_fu, can we get cwr pointed at cs:~james-page/rabbitmq-server-requirement whilst we test this out? [14:49] jamespage: We will need that BT change merged and released, build-cloud updated, charmbox rebuilt, https://hub.docker.com/r/seman/cwrbox/builds/ rebuilt, and the Jenkins slave(s) updated. I don't have access to do the last two. (Obviously, there are too many moving parts in the current CWR system, so we're looking to streamline that.) [14:50] tvansteenburgh: Do you have access to the Jenkins slave(s)? Could you do the last two on there? [14:50] cory_fu: which slave [14:51] tvansteenburgh: xenial-slave [14:58] cory_fu: looks like no. i can ask sinzui if it's urgent? [14:59] tvansteenburgh: I guess it depends if Seman is still out sick today [15:01] cory_fu - jsut for future reference, when we rebuild charmbox, cwrbox triggers a rebuild [15:02] so we *can* update the cwrbox [15:02] lazyPower: The timing on the builds made me think that it wasn't causative, but that's good to know [15:02] yeah, i cant help with the slaves, but i can help w/ the dockerhub stuff :P [15:03] And the docker image is pulled for each build, so that's good. [15:14] if anyone wants me to import goods into the USA next year once the trade deals are scrapped.... let me know! ;) [15:15] lazyPower: did you reschedule that elastic stack chat? [15:18] magicaltrout - not yet, looking for a good date next week. what works for you? [15:18] how sober do i need to be? [15:19] its just planning/chat so 70/30? [15:20] okay not Wednesday or Thursday lazyPower [15:20] other than that i'm good [15:21] if you make it a bit later than you did this time, its less likely to clash with any PST meetings I have [15:21] they're always around 8-9am [15:22] of course the trade off their is i'll have drunk more ;) [15:36] no worries i'm armed to the gills against drinking engineers ;) [15:36] i'll shoot for a 10am PST meeting on Tuesday [15:47] cool [15:47] admcleod_: you coming next week or not then? [16:07] magicaltrout - i just sent the updated calendar invite, thanks for keeping on me about that :) [16:07] i also attached a meeting agenda doc, and you should have edit rights, feel free to add to the agenda topic list if you have anything thats burning from your perspective [16:39] heya balloons [16:39] I'm updating the juju instructions in the kubernetes documentation, do we have a section in our docs on how to install via snaps? I would like to include other distros if possible. [16:39] +1000 to that notion [17:53] tvansteenburgh, kwmonroe: Replied to your comments on https://github.com/juju/amulet/pull/166 [17:53] tvansteenburgh: Missed your reply about the comment. Adding that now [17:56] tvansteenburgh, kwmonroe: Actually, I just noticed that Amulet already includes path.py in its requirements, so I can just say "for filename in Path(dir).files():" [18:10] hows it going everyone? [18:11] Whats up bdx [18:11] Merlijn and I just hashed out some initial work on juju <-> jenkins integration ideas, its on this etherpad -> http://184.72.79.132/p/jenkins-collab [18:11] we would love any insight/feedback anyone has [18:12] we had to cut it short today, but it will be an ongoing project so ... stay tuned! [18:16] bdx: interesting... what's that jenkins charm got compared to the upstream jenkins? [19:10] lutostag: xenial [19:10] experimenting with local lxd in the build process [19:11] needed xenial [19:12] bdx: https://github.com/jenkinsci/jenkins-charm supports xenial [19:12] i think important info here is this charm layer landed last week ^ [19:12] true [19:13] did we ping the list with that info? [19:13] lazyPower: good point... the communication is lacking for sure [19:14] lazyPower: marcoceppi had plans for making that the promulgated jenkins charm, anything I can do to help that process? [19:14] lutostag - i think we just need a RevQ submission for that [19:14] lazyPower: ok, thanks, will do [19:15] bdx: sorry, our fault for not making it wider known, I hate when that happens [19:16] lutostag: its cool man ... it would be sweet to get that promulgated though [19:17] lutostag: theres no series in the metadata .... does this mean it supports xenial by default lol? [19:18] bdx: should work on trusty + xenial, but I think we'll just try to get it promulgated for xenial, so we don't need to worry about a trusty upgrade story === thomi_ is now known as thomi [19:46] icey, https://github.com/juju/theblues/issues/49 [19:47] beisner: good one :) [21:17] marcoceppi, ubuntu charm file links leads to this -> https://api.jujucharms.com/charmstore/v5/~marcoceppi/xenial,trusty,precise/ubuntu/archive/Makefile [21:18] when seeing the charm details from my local juju gui === zz_CyberJacob is now known as CyberJacob === CyberJacob is now known as zz_CyberJacob [23:55] marcoceppi, beisner: where does the jenkins-slave interface come from?