=== natefinch-afk is now known as natefinch === thumper is now known as thumper-afk === natefinch is now known as natefinch-afk === mup_ is now known as mup === frankban|afk is now known as frankban [09:00] Hello Juju World! === jaro is now known as Guest90451 [12:38] kjackal: howdy [12:38] hi rick_h_! [12:51] o/ juju add-unit hello-world [13:24] magicaltrout i am excite! We have a meeting in ~ 30 minutes, and there's a new com truise album! I dont know what i did to deserve today, but it must have been something right [13:25] lazyPower: lol, my brain turned that into "a new tom cruise album" [13:26] rick_h_ - i do the same thing sometimes ;) California based electronic artist != eccentric billionare movie star === natefinch-afk is now known as natefinch [13:41] is there anyone running xenial who can verify a bug for me? [13:44] balloons - i've got several xenial containers/vm's. Whats up? [13:46] lazyPower, I'm trying to figure out bug 1614959. You need a xenial with juju-1-default installed and to have not upgraded to 1.25.6 yet [13:46] Bug #1614959: /usr/bin/juju is missing after upgrade to 1.25.6

[13:46] weird, let me try to repro -- i'm headed into a meeting shoirtly but will ping back with results if i can repro [13:56] that awkward moment when you realize a system restore nuked your VM's because you omitted them from the backups... [13:58] rick_h_: my double vision also turned that into Tom Cruise [13:59] https://www.youtube.com/watch?v=jTgttHHjGRI [13:59] maybe that will help [14:00] lazyPower, I'd like the output of all those commands in the bug before and after the upgrade. ls -al /usr/bin/juju* and dpkg-divert --list | grep juju [14:00] balloons - ack, will do. just wrapped up creating a new vm [14:01] magicaltrout - jwaiting on you sir https://hangouts.google.com/hangouts/_/canonical.com/troutpocalypse?authuser=0 [14:02] my hangout says its requesting a join [14:02] * magicaltrout tries again [14:03] oh i was joining as meteorite.bi [14:03] * magicaltrout swaps [14:03] that awkward moment when you realize you have the hangout locked up from the casuals [14:54] lazyPower: Amazon Music just said [14:54] "We ran the numbers and we think you'll like this station" [14:54] * lazyPower smirks [14:54] It was Little Mix.... [14:54] :'( [14:54] little mix? [14:54] i'm not familiar [14:54] girl band [14:54] their learning algo sucks [14:55] heh, do you have family members using your prime acct? [14:55] that's a sure fire way to mux up their learning algs [14:55] aye. No one else uses it, the only stuff in it is Weather Report, Blue Brothers, and a few punk albums [14:55] not sure what planet they're on [14:57] "we ran the numbers and the record label paid us to promote this factory band" [14:58] ^ [14:58] i think thats more than likely the culprit [14:58] hehe [14:58] maybe they mistyped s/the/our [14:59] oooh its so nice to nearly be able to read a screen today [14:59] that said I have a 4k screen with gnome-terminal on full zoom [15:01] magicaltrout - they say trouts are wall-eyed [15:01] do you feel like you're looking through a fish-eye lense? [15:01] hehe it does at times [15:01] its weird today [15:02] sometimes i can see with perfect clairty for a split second [15:02] then it switches to fuzzy [15:02] but keeps repeating for a few minutes [15:02] like my eyes remember how to focus and then forget [15:02] magicaltrout: are you ok? detached cornea? [15:02] na jrwren lasek [15:03] i'm living vicariously through magicaltrout's surgery. i've considered it myself bu ti'm hyper scared to have someone lazering on my eyeballs [15:03] hehe [15:03] i can read a license plate from 36 meters today when i tested it [15:03] which is pretty cool [15:03] its just the close up stuff that i'm having issues with [15:03] but thats normal [15:04] for 7 - 14 days [15:04] so if i do it, it'll be over xmas shutdown [15:04] depends what you have done as well lazyPower [15:04] i cant afford the time away from reading a monitor :( too much to do, too little time [15:04] lasek or lasik [15:04] lasik is more standard and much quicker recovery [15:04] but my cornea was too thin [15:05] i have astigmatism so i'm likely to have the extra expensive and tricky lazering [15:05] yeah i had an astigatism as well [15:05] it wasn't cheap, but i don't like my lenses and glasses get on my nerves [15:05] agreed [15:05] they fog up in winter, and slide off your nugget in summer [15:06] yup [15:06] the only perk is transitions [15:06] i always have sunglasses on hand [15:06] i've only had glasses for a month. :p [15:06] i've been wearing glasses since i was 12 :/ [15:06] except i now have hipster sunglasses without the problems of a prescription ;) [15:06] i'm over it [15:06] yeah i had them around 12 [15:06] i like lenses because of the all round vision [15:07] but they just wear your eyes out [15:10] yeah [15:10] eyestrain is a real thing [15:10] I tend to work until my eyes hurt and thats when i know its time to EOD [15:10] not the best policy by any means [15:11] hmm [15:11] i don't get eyestrain [15:11] maybe thats why i just keep going [15:11] I never felt eyestrain, but maybe its because I've trained my eyes since I was 6yo :p [15:11] And now with fancy pants lenses doing that whole polarization filtering thing... even less eyestrain, not that I ever noticed. [15:16] well assuming my near sighted vision clears up i'm well happy with this investment [15:17] even if i need glasses for some near sighted stuff which isn't the plan, i'd still consider it money well spent [15:19] fair enough :) [15:20] funnily when i had my checkup the other day i couldn't see anything out of my right eye [15:20] today its much clearer than the left eye [15:21] they need to hurry up and align so I can write talk content ;) [15:21] your cones and rods are working against you [15:22] lazyPower, did you try replicating the bug? [15:23] balloons - i have the vm stood up, was just about ot circle back and run through the log [15:26] balloons - give me 15 minutes and i'll context switch to this until i have your log output [15:28] lazyPower, thank you. The key is to ensure juju-1-default is diverting /usr/bin/juju to juju-1 before and after the upgrade [15:45] rick_h_: so far for the summit I only have your talk and dimiter's on networking for core-ish talks. [15:45] rick_h_space is filling up so if you want more core-ish talks please lmk asap === frankban is now known as frankban|afk [15:58] kwmonroe: petevg: I have a question regarding the Barcelona style issue that we have, what zookeeper charm should we use for promulgated bundles? [15:59] balloons - to be very specific, if i install juju-1.25=1.25.5-0ubuntu3 and then upgrade [15:59] kjackal: I think that we should use the trusty version. [15:59] thats the path we are looking to have tested? [16:00] petevg: the one we have from james, right? [16:00] kjackal: correct. The one that's already in the store. [16:00] The xenial one isn't promulgated yet. [16:00] And when we do promulgate it, we'll probably do so without the need for the openjdk relation. [16:00] cory_fu: does the above sound correct? [16:00] petevg: the xenial is promulgated [16:01] https://jujucharms.com/zookeeper/ (xenial only) [16:01] lazyPower, yes. And make sure juju-1-default is installed [16:02] kwmonroe: cool! In that case, kjackal: I think that you should still use the trusty one, if the rest of your bundle is trusty stuff. Though if you want to wait for the xenial one to get re-deployed without the need for the openjdk relation, then you could use the xenial one. [16:02] lazyPower, and record the values before and after so we can vet it looks good before (and works), and doesn't afterwards [16:02] petevg: kjackal, the trusty version (james') has to be explicitly namespaced (cs:~charmers/zookeeper/trusty): https://jujucharms.com/u/charmers/zookeeper/trusty [16:03] kwmonroe: Did we deploy a trusty version of the new zookeeper charm, then? [16:05] Hi, can we have a package name without any extension say .zip or tar.gz in resources ? [16:06] balloons http://paste.ubuntu.com/23070607/ [16:06] no petevg.. bigtop zk for trusty is not promulgated. that version is only available in bd-dev: https://jujucharms.com/u/bigdata-dev/zookeeper/trusty [16:08] kwmonroe: have we promulgated the bt spark? (You are our official release manager!) [16:11] negative kjackal, latest spark with your most recent "Spark tests need to pass" fixes is https://jujucharms.com/u/bigdata-dev/spark/trusty/9 [16:12] kjackal: i was going to promulgate after checking out your "spark houskeeping" review [16:12] Cool, thank you [16:12] balloons - all that looks sane to me [16:13] i've nuked and repeated twice now... unable to reproduce [16:14] petevg: I got multiple agents in failed state when deploying the hadoop bundle, digging into this [16:14] kjackal: cool. thx for investigating. [16:15] lazyPower, that makes me very happy to hear.. Although I'm still not sure why others are seeing it [16:15] lazyPower, thank you [16:15] * lazyPower nods [16:15] happy to help, sorry it took so long [16:15] who knew friday would be busy when you come back from time off ;) [16:17] shilpa: i just checked out the latest resources docs (https://jujucharms.com/docs/2.0/developer-resources), and it seems like there is some logic to check the extension.. so i'm not sure if you can omit one. the test would be to specify a non-extension filename in metadata.yaml and see if juju allows you to subsequently attach a non-extention file to that charm. [16:17] shilpa: here's the sentence that makes me think there's some extension logic happening: "The filename is what Juju will name the file locally when it is downloaded. Juju will check the extension on the file being uploaded and will prevent files with different extensions from being uploaded." [16:21] shilpa: based on this bug, it seems you should be able to define a filename with no extension in metadata.yaml and attach a resource with no extension: https://bugs.launchpad.net/juju-core/+bug/1578383 [16:21] Bug #1578383: incorrect extension on resource upload

[16:21] petevg: this is in the ganglia nodes but it seems unrelated... http://pastebin.ubuntu.com/23070770/ [16:23] kjackal: interesting. I think that I may have seen something similar. Will dig around in the logs in a bit (right now, my local juju is busy bootstrapping a new aws controller). [16:23] thanks kevin, sometime back i had tried without specifying extension in metadata file, when i did juju attach, it was throwing some error. [16:24] i will try once again without extension in metadata and do juju attach. [16:29] kjackal: hmmm. I don't seem to have that particular error in my current ganglia node. You're right that it's probably not related, though. [16:30] petevg: going to queue the tests on a different machine [17:11] jcastro: k, tganks fornthe heads up. have we reached out to bac about something around the charmstore/gui? [17:12] are you asking me or telling me? [17:12] jcastro: and any suggestions on core thing that would go over well for the audience? /me isn't sure who' coming to this one [17:12] jcastro: asking [17:12] jcastro: and slightly suggesting if the answer is no [18:01] how do you specify local charms in a juju2 bundle? [18:01] it seems to have changed [18:02] nvm, giving it the full path works [18:03] cholcombe: I think relative path works too, it just needs to start with ./ [18:18] does charm push take into account layer+reactive charms? I'm guessing I push the built bits correct? [18:21] cholcombe: charm push knows nothing about layer or reactive. push the built bits. [18:25] jrwren, ok cool that's what i thought [18:45] marcoceppi: what is secretstorage ? [18:46] kwmonroe: are you able to get bundletester to work at all with juju beta 15? I fail on a KeyError in jujuclient (it looks like its due to some consistencies in how "juju show-model" and "juju switch" name the models). [18:55] bdx: it's like a dependency or a dependency of a dep that got bumped and broke a bunch of stuff [18:56] bdx: beisner has the details [18:59] hi, do you guys know what's up with juju(2) storage? [18:59] I can't get a pool list [18:59] $ juju-2.0 storage pool list --filesystem [18:59] "machine-pool" is not a valid machine tag [18:59] "machine-list" is not a valid machine tag [18:59] juju-2.0 help storage also doesn't match, at all, what's described in https://jujucharms.com/docs/2.0/charms-storage [19:00] ahasenack: yes, there's a bug open on the docs to get it caught up to the 2.0 cli updates [19:01] rick_h_: according to help, there isn't any subcommand for storage anymore, is that right? [19:01] juju help commands | grep pool [19:01] ah [19:01] ahasenack: yes, like everything else in 2.0 there's a pool "noun" and verbs to list, etc [19:02] list-storage-pools lists the types, not the pools I created [19:02] I guess that's when "juju-2.0 storage" comes in [19:03] ahasenack: hmm, there's show-storage for a specific one, is storage the list-storage equiv that shows the created ones? [19:03] * rick_h_ needs to play with that more [19:04] I grepped for pool :) [19:04] heh [19:04] * ahasenack refines the grep [19:04] ok, I see [19:04] lots of aliases [19:04] since we are still in beta, these aliases could be dropped [19:04] ahasenack: yes, definitely [19:05] ahasenack: another bug that's filed to clean those up somewhere. [19:10] I think pool, count and size are in the wrong order [19:11] The acceptable format for storage constraints is a comma separated [19:11] sequence of: POOL, COUNT, and SIZE, where [19:11] ... [19:11] juju add-storage u/0 data=ebs,1024,3 [19:17] ahasenack: heh yes, agree [19:54] lazyPower: for the kubernetes bundle on openstack, looks like they need to be able to add an external keyserver... so if there is firewalled access I'll have to clear the firewalls right [19:54] wolsen - there's quite a few external dependencies there [19:54] external key server, external image registry, and github access required to clone easy_rsa [19:54] lazyPower: ack thx, let me just enable the squid proxy in general [20:56] hiya, bradm: are you still maintaining the bip charm? There are a couple of open PRs against it (start at https://code.launchpad.net/~josvaz/charms/trusty/bip/client_side_ssl-with_helper-lp1604894/+merge/301802) that need some love from the maintainer (and possibly a redirection to your branch). [21:32] Have a strange error. We are building new layer-docker charms that passed proof and build back in March . The exact same code now produces the following error when running 'charm build ./hss' : [21:32] Traceback (most recent call last): [21:32] File "/usr/bin/charm-build", line 9, in [21:32] load_entry_point('charm-tools==2.1.2', 'console_scripts', 'charm-build')() [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/__init__.py", line 673, in main [21:32] build() [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/__init__.py", line 516, in __call__ [21:32] self.generate() [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/__init__.py", line 467, in generate [21:32] self.formulate_plan(layers) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/__init__.py", line 408, in formulate_plan [21:32] self.plan = self.plan_layers(layers, output_files) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/__init__.py", line 319, in plan_layers [21:32] next_layer / BuildConfig.DEFAULT_FILE, True) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/config.py", line 81, in add_config [21:32] c.configure(config_file, allow_missing) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/config.py", line 69, in configure [21:32] tactic = load_tactic(name, basedir) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/build/tactics.py", line 661, in load_tactic [21:32] obj = utils.load_class(dpath, basedir) [21:32] File "/usr/lib/python2.7/dist-packages/charmtools/utils.py", line 321, in load_class [21:32] dpath, workingdir)) [21:32] OSError: Unable to load tactics.docker.DockerWheelhouseTactic from /home/juju/charms/deps/layer/layer-basic [21:33] Any advice is appreciated [22:54] cloudguru: have you updated the layers recently? [23:15] i've done apt update .. how do you update layer-docker layers ? [23:55] It looks like the charm pulls in the latest layer-docker .. I also tried to pull in from git but had the same result [23:55] could this be a simple permissions issue ?