=== CyberJacob is now known as zz_CyberJacob === valeech_ is now known as valeech [02:01] lazyPower, mbruzek: just browsing around and found this interesting pertaining to kub + lets encrypt https://github.com/kelseyhightower/kubernetes-letsencrypt-tutorial [02:02] that guy is a bad ass, check out some of his other work too if you have a minute .... super cool [02:06] https://github.com/kelseyhightower/vault-controller [02:56] xenial+reactive+python3 jenkins refactor -> https://github.com/jamesbeedy/juju-charm-jenkins [02:56] jamespage: ^ [03:23] boom https://jujucharms.com/u/creativedrive/jenkins/4 [05:38] bdx, oh nice [05:38] beisner, ^^ [05:38] bdx has been busy [08:08] Good morning Juju world === frankban|afk is now known as frankban === ezobn is now known as Guest98417 [10:12] bdx: I feel like I'm constantly the bearer of bad news [10:12] bdx: https://github.com/jenkinsci/jenkins-charm [10:12] bdx: in the future, I think we should start pushing a policy of people announcing what they're working on, we're (me) are terrible at this === ezobn is now known as Guest89709 === ezobn is now known as Guest58526 [11:38] magicaltrout: you should setup a ci it is supper easy (took me one evening). You just chekin something and your work is pushed to the store automagically. I have only good words to say! [11:40] I have CI kjackal just not for charms, I'm lazy :P [11:41] that said, its not like I code charms for a living, so I suspect our priorities differ ;) [11:42] magicaltrout: true! [11:42] got knocked back by the apache mesos folk when asking about LXC support, silly people, they're missing a trick [11:42] looks like I'll have to learn C [11:43] magicaltrout: I hope someone from juju-qa will step up and say "kjackal you are doing this all wrong! We have this jenkins plugin that will make your life super easy!" [11:43] Ahhh C! Best language ever! [11:43] magicaltrout: not joking ^ [11:44] I don't believe you [11:45] magicaltrout: Everything you do with C is like DIY! Always wear protective gear and never try this at home! [11:45] lol [11:46] I tried extending mesos for LXC a few weeks ago, it was very much trial and more error [11:47] it was funny they asked me what use case they were missing..... [11:47] why people think application containers are the best way forward I have no clue [11:50] developers build these things to protect themselves from thierselves [11:51] like Go?! ;) === icey-travel is now known as icey [12:01] I think the story goes like this: In begining there was C. Then developers got annoyed by junior developers that were clumsy with C and they created Java for the Junior devs to play with. But then Java became too popular and quantity beat quality: It was better to have more Java junior devs that few quality C devs. One C dev said I will reboot C and add some of the features I am missing so we have Go. While most C devs are [12:01] boosting C [12:01] magicaltrout: ^ [12:01] lol [12:02] yeah, i suspect you're not too wide of the mark kjackal [12:22] Hi all. we pushed our juju charm to public charm store using launchpad juju account. when we pushed it is giving like cs:~hareeshk/kaminario-cinder-10 . It was pushing with account username "hareeshk". Can we push our charm by hiding "username". Please provide me information. [12:33] rock: to do that you have to get it signed off and "recommended" by some of the charmers [12:34] so you need to ensure it has tests, tests pass, a decent README etc [12:34] then it will need to enter the review queue [12:41] magicaltrout: Hi. Oh. For that it requires the certification right. [12:42] well its not certification, they just check to make sure your charm is sane [12:43] rock: https://jujucharms.com/docs/2.0/charm-review-process [12:44] doesn't seem to explain how to get it in the review queue though kjackal [12:49] Kjackal/magicaltrout : Thanks. [12:50] rock kjackal http://review.jujucharms.com [12:54] marcoceppi: Thanks. How much time it will take to review? [12:54] rock: we're a little behind, but we'll try to take a look quickly [12:56] marcoceppi: OK. Thanks. [14:25] marcoceppi: yeah ... this is really defeating .... the jenkins you posted doesn't even show up in the charm store ... is in not published for some reason? [14:25] bdx: it just landed [14:25] I'm really sorry man [14:25] jesus [14:25] we're not going to cock this up again [14:26] you can expect us announcing the charms and layers we're working on, from the entire company, much earlier [14:26] entirely [14:31] possibly we could make an addition to charm build where on every build it would post the charm name and author name/email to and endpoint somewhere [14:31] or something of that fashion === daniel is now known as Guest19937 === Guest19937 is now known as Odd_Bloke [14:35] marcoceppi: why not target xenial on that? [14:37] bdx: we are [14:37] we will target xenial only for that charm, I'm building and uploading now [14:47] bdx: I feel soo bad, but I should check now - what other charms/layers you got in the works? I want to make sure we don't curb stomp your dreams [14:48] marcoceppi: no worries, practice makes perfect [14:49] marcoceppi: zuul, gitlab, documize, etherpad-lite, mailman3 [14:49] marcoceppi: I've yet to start on zuul [14:49] bdx: yeah, I'm interested as well in gitlab, and I know Tom Barber has a start of that somewhere [14:50] marcoceppi: documize, etherpad-lite, mailman3 are done minus tests [14:50] otherwise, i'm not awayre of any of those in flight [14:50] ok [14:50] I only have a start on gitlab too [14:50] beisner: do you'll have anything around zuul? [14:51] I should the trout about that for sure [14:51] magicaltrout:^ [14:52] marcoceppi: I have a reactive rewrite of kibana I've been deploying .... but I'm guessing the whole elastic stack should be consuming a base layer of some sort eh? [14:52] jhobbs, only via integration with upstream gerrit + zuul. we don't have our own zuul deployed though. [14:52] bdx: dude, that's awesome! we were jsut about to start moving elasticsearch to 5.0.0 and layered! [14:52] bdx: where's your kibana? [14:53] https://github.com/jamesbeedy/layer-kibana [14:53] \m/ [14:53] lazyPower: ^^ [14:54] thats great news \o/ [14:59] here are another few layers I've hacked up recently [14:59] https://github.com/jamesbeedy/layer-nginx-passenger [14:59] https://github.com/jamesbeedy/layer-postfix [15:00] https://github.com/jamesbeedy/layer-git-deploy [15:00] https://github.com/jamesbeedy/layer-apache2 [15:00] https://github.com/jamesbeedy/layer-phalcon [15:00] https://github.com/jamesbeedy/layer-consul-base [15:02] bdx - we should seriously, sync and divvy up the work on the elastic stack [15:02] bdx - the goalpost got moved on me, and its not coming as soon as I was hoping. I'm wiling to put in some time if you are to split that work up and deliver the 5.0 elastic stack [15:03] lazyPower: fully down man .... I've been training up a two man charming team at work ... few hours a day ... its been great all around [15:03] we would be interested in teaming up on that [15:03] for sure [15:03] word, calendar sync on thursday then to meet/greet and figure out what we need to do? [15:03] i have some early planning docs that i need to fish up and dust of [15:04] totally .... I'm in palm springs till sunday and wont be around much, lets shoot for next week sometime [15:04] monday? [15:04] that sounds even better [15:05] i may need to resched due to having team members at kubecon, i'll be running support for them [15:05] but tentatively that sounds great to me [15:05] sweet! [15:05] excellent :) [15:06] bdx preferred time? [15:06] 9am my time [15:06] so like an hour from now [15:06] does that work for you? [15:07] want me to send you a calendar invite? [15:08] bdx you've got mail [15:32] Marcoceppi: do you also have a Jenkins-slave going on? [15:33] marcoceppi: having Jenkins and Jenkins slave on xenial will be huge .... Lxd local provider + juju 2.0 can't really happen [15:33] On any other series [15:34] For Jenkins <-> juju ci pipelines I'm thinking it will be crucial to have this [15:43] bdx - there's an open jenkins layer review we could use your eyeballs on [15:44] ooo and to my surprise its already landed [16:11] lazyPower, marcoceppi: not sure I can accept the invite -> https://s12.postimg.org/8lcvygobx/Screen_Shot_2016_11_02_at_9_10_36_AM.png [16:11] its on my calendar though [16:32] nm, git it, accepted [16:32] solid [16:32] bdx - should we make this a hangout on air and invite the community at large? [16:33] i can also reach out to my contacts at elastic to see if they are interested in participating from a curious onlooker position. They might be able to lend a body for planning and insights. [16:33] lazyPower: totally [16:33] lazyPower: my guys over here want to be part of this too [16:34] bdx ok, let me see what i can do when jcastro surfaces. He's got all this down to a science [16:34] thanks for agreeing to help chair this effort :) [16:34] magicaltrout - if you're interested, you're invited toooooo [16:35] who needs help? [16:36] i'm always interested in taking on yet more unpaid work! ;) [16:37] lazyPower: hangouts on air went away, the tldr is you'd do it on youtube like how with gaming.youtube [16:37] there's a little panel for livestreaming, etc. [16:38] ahhh [16:38] magicaltrout - well i didn't mean it like that, but i appreciate the sentiment :D <3 [16:38] lol [16:38] whatever :P [16:38] magicaltrout - i was just thinking you likely have a stake in this as its solving little data problems too [16:39] this is true [16:39] i've been abusing people at work recently over elastic stuff [16:39] man this new kibana looks so sweet though [16:39] i'm [16:39] sayin [16:39] they did a bang up job on the 5.0 release [16:40] it's like, I don't even want to use the old one at all anymore [16:41] jcastroyeah, we all want it [16:41] sorry they kept teasing me with getting to work on that stack :\ [16:41] we had other bigger ticket items [16:42] dont forget fast data [16:42] soooo free time it is :D [16:42] admcleod1 - que? [16:42] big, little and fast data [16:42] fast big data? [16:42] is there a slow data too? [16:42] absoutely [16:42] what about obtuse data vs acute data? now we just sound classist. [16:42] have you ever been introduced to Map Reduce? [16:42] irc is pretty slow data [16:42] once [16:42] it was horrible [16:43] MapR is about as admcleod1 on a good day [16:43] ooooo [16:43] databased on filesystems backed by remote block devices over gigabit is slow data. [16:43] he's such a snuggly bunny [16:43] currently i'm dealing with TB's of data... on NFS [16:44] now that is slow data [16:44] magicaltrout - did you use glusterfs-nfs or just plain ole nfs? [16:44] oh, pretty much what jrwren said :) [16:44] lazyPower: not my call, plain NFS [16:44] * lazyPower nods [16:44] either way, if you want fast data, don't store it on NFS [16:44] * lazyPower pins that message to the channel [16:45] well the other day I got someone complaining that copying 100GB of data on an NFS cluster on EC2 took 1 1/2 hours [16:45] 4.9 kernel adds supoprt for NFS4.2 COPY... so you can look forward to that speed up ;] [16:45] this is correct! then doing it on a machine locally took a few seconds === lazyPower changed the topic of #juju to: Welcome to Juju! || Docs: http://jujucharms.com/docs || FAQ: http://goo.gl/MsNu4I || Review Queue: http://review.juju.solutions || Unanswered Questions: http://goo.gl/dNj8CP || Youtube: https://www.youtube.com/c/jujucharms || Juju 2.0 release notes: https://jujucharms.com/docs/stable/reference-release-notes [16:46] evilnickveitch - did we get 2.0.1 release notes? i thought i saw a merge email fly by on friday but i cant seem to find it [16:46] magicaltrout: was it EBS backed? EBS can be very slow sometimes. [16:46] yeah jrwren [16:47] magicaltrout: have you tried fastnfs? [16:47] my Sys Admin guy is putting his eggs into the Amazon EFS basket and waiting to migrate to that [16:48] you making up words again admcleod1 ? [16:48] he is [16:48] thanks [16:48] yes, yes i am [16:48] made me look [16:48] jerk [16:48] hehe [16:49] lazyPower, not any that made their way to docs - I think it was just an email to the list. i will check into that [16:49] ok thanks, i must have conflated the mail to the list with a doc pr. :) my b === admcleod1 is now known as admcleod [16:52] lazyPower, there was a doc PR to update the sources for the new release, but no notes, so your brain isn't that wonky [16:52] evilnickveitch - how much longer before i've earned the 'mad scientist' badge? [16:52] s/before/until/ [16:53] I think you are halfway there [16:53] \m/, yesssss [17:01] halfway to zeno maybe [17:52] stokachu: battlemidget! [17:52] have you tried doing conjure-up, juju, and lxd all from snaps? === frankban is now known as frankban|afk === zz_CyberJacob is now known as CyberJacob [18:47] jcastro: yea it doens't work [18:47] not easily anyway [18:49] jcastro: https://github.com/conjure-up/conjure-up/tree/master/snapcraft thats our snap build stuff [18:49] havent ran it in a bit though [18:54] yeah, the one I'm using is also trying to launch observable kubernetes [20:11] Is docker charms supported with localhost LXD setup? I've tried both the 'observable-swarm' and 'swarm [20:11] deanman - those bundles are old, but in short, that configuration is not supported today [20:12] deanman - there is work being done to make running docker containers in lxd a better story, right now you can even pilot some of that manually with lxd profiles (the lxd docker profile is available in newer lxd packages) but you'll still find a lot of limitations there [20:13] lazyPower: Can you suggest a working charm to test docker locally using Juju or only safe bet is using some cloud infrastructure? [20:13] deanman - the only way you can achieve a layer-docker based charm working locally today is to use a VM backed substrate like KVM or VirtualBox (both on the manual provider, RIP KVM provider) so a majority of the work on those charm are performed in the cloud on aws/gce/azure [20:15] lazyPower: Thanks for the hints! === CyberJacob is now known as zz_CyberJacob