=== balloons is now known as Guest90681 === chihchun_afk is now known as chihchun [04:46] * Guest42341 What a fine day for science! === bigcat_ is now known as bigcat [07:27] good morning === kickinz1|away is now known as kickinz1 [07:45] good morning :) [08:12] good morning [09:38] good morning [09:46] Good morning all; happy Monday, and happy OK Corral Day! 😃 === _morphis is now known as morphis [09:56] what's /usr/lib/snappy-systemd/systemd-snappyhook ? [09:57] ogra_: is that one yours? ^ [11:42] dpm, thanks for the emails [11:42] I'll bring it up in the snappy standup later on as well [11:43] great, thanks dholbach [11:46] Chipaca: hi, eventually I decided to propose the simple solution for Bug #1496319 (just a symlink without filters) :) [11:46] bug 1496319 in Snappy "Could not create symlink to hw device with udev rules" [Undecided,New] https://launchpad.net/bugs/1496319 [11:50] clobrano: still needs jdstrand's and/or co's input [11:51] Chipaca: co? [11:52] mvo: hey [11:52] mvo: nice to see you again, how are you? [11:53] clobrano: "and/or his team" [11:54] Chipaca: aahh, get it now ;). It's fine :) [11:55] hey zyga! thanks, I'm very well. how are you? [11:58] mvo: I just returned from holidays, I'm settling in [11:59] mvo: I wanted to ask about snaps for the kernel and a few other bits, if you remember our conversation from the sprint [12:00] zyga: I'm not sure I do, but there is progress for kernel snaps since the sprint [12:00] mvo: is there enough bits now to come up with an image for any system with just the snaps? [12:01] zyga: http://people.canonical.com/~mvo/all-snaps/ has all thats needed. there is a example image, there is a ubuntu-device-flash, a kernel and os snap and the run to show how to drive udf in the new world. the next step is store support for the kernel/os snaps and then we can get serious about testing them [12:05] mvo: that's a lot, thanks! I'll read that and see how to start desigining skeleton for ubuntu-image [12:07] zyga: this is what we did for u-d-f to support all-snaps https://code.launchpad.net/~snappy-dev/goget-ubuntu-touch/all-snaps/+merge/275273 [12:08] zyga: it needs some cleanup, its just enough to get us going for better testing [12:09] mvo: I see, thanks [12:37] sergiusens: hey! [12:37] sergiusens: how are you :) [12:37] sergiusens: I'm back from holidays today [12:38] sergiusens: would you have a moment to sync later today? [12:40] lool: ping [12:40] shuduo: pong [12:41] zyga, sure, but later, not just now ;-) [12:42] lool: i'm working to create a webapp snap example based on tomcat-maven-webapp. i wonder how the scribpt 'wrapper' be generated? [12:42] sergiusens: I'll setup a meeting and we'll see [12:42] shuduo: see tomcat-maven-webapp/Makefile [12:42] shuduo: install -D -m755 wrapper $(DESTDIR)/bin/wrapper [12:43] bottom of snapcraft.yaml runs make [12:45] lool: i mean, is it composed by you or someone else for tomcat-maven-webapp? Is it necessary for tomcat-hosted webapp? [12:46] shuduo: it's part of the example snap and helps start tomcat properly; you're free to get rid of it [12:46] shuduo: perhaps you should mention your higher level goals; it looks like we're discussing specifics but you have a larger point to make [12:50] lool: sorry i just try to understand the structure and call stack. i'm not an expert of webapp. i just tried to learn it as our customer maywant my answer regarding example code [12:50] lool: i tried to program a tomcat-hosted spring based webapp then make it as a snap [12:51] lool: i'm curious why we need wrapper and who call it after it be installed [12:56] lool: oh, i find the part of snapcraft wrapper script explaied. it should answered my question. [12:58] shuduo: basically snaps can either expose commands or services; in the case of this example snap, a tomcat service is exposed and is launched with this wrapper; note that snapcraft will generate intermediate wrappers (e.g. to set environment variables such as LD_LIBRARY_PATH or PATH) [13:06] lool: let me read application developer manual first. thanks for your explaining. :) [13:07] shuduo: anytime; let me know if you have issues or questions with this sample === Guest90681 is now known as balloons [13:38] lool: i found my sample project seems not packed well. after I installed on snappy on vm i can't see webapp run on specified url. [13:39] shuduo: try sudo snappy service logs [13:46] nope nope nope [13:47] lool: my sample project is a webapp use tomcat8 as host, spring as framework, mvn build system. I verified on my laptop with localhost:8080/SpringMVC. then i build by snapcraft and installed on snappy system on vm. then i use IP:8080/SprinMVC but it return 404 [13:48] shuduo: can you connect to :8080? [13:48] shuduo: do you get tomcat 8 running there? [13:48] lool: yes [13:48] shuduo: if yes, then it's a webapp deployment issue [13:48] lool: then how i debug it? [13:48] shuduo: so this is more tomcat world than snappy world, but either you dump webapps preinstalled in the tomcat runtime dir, or you put the webapp to deploy in the to-be-deployed tomcat location [13:49] shuduo: typically in the later case you'd install a .war in the to-be-deployed dir [13:49] and that'd be installed in the runtime webapps dir [13:49] shuduo: checkout the logs under the runtime dir === chihchun is now known as chihchun_afk [14:19] fgimenez, so I haven't forgotten about your session for UOS :-) [14:20] we'd like to get things all scheduled this week, so it would be great if you could propose the session. [14:45] nessita: I've just gotten this TLS error: http://pastebin.ubuntu.com/12970390/ [14:48] elopio, checking! [14:48] lool: i can se the war file be placed in /var/lib/apps/spring-mvc-demo.sideload/current/webapps. I pasted the logs content at https://pastebin.canonical.com/142681/. i don't know what is clue. could you pls take a look? [14:48] elopio, but is for an image not a package? [14:50] balloons, :) ok, tbh i'm not yet sure about this, anyway i'll try to find the time and contents to prepare something and let you know [14:50] fgimenez: sorry for throwing you here... :) [14:51] you can say no, but you'll break balloons' heart. [14:52] elopio, care to look at https://code.launchpad.net/~sergiusens/snapcraft/1481122/+merge/275237 again? [14:52] nessita: generic-amd64 is a package. [14:52] that u-d-f downloads to create an image. [14:52] elopio, but the error is about downloading the package icon? [14:53] elopio, but it also downloads the icon; nessita is on the right track [14:53] what is failing is the link in the store manifest to retrieve the icon [14:53] the client should definitely retry and perhaps install without icon; on the other hand I will pursue the myapps setup about icon downloads [14:53] elopio, have a bug for this? [14:54] thanks elopio and fgimenez :-) [14:54] nessita: https://bugs.launchpad.net/snappy/+bug/1488186 [14:54] Launchpad bug 1488186 in Snappy "Failed to update: gnutls_handshake() failed" [High,Triaged] [14:55] we need to retry in many places. [14:56] elopio, thanks. I was thinking on a bug report more specific to icon download error vs package download error vs system image download error, because each one is fetched from very different places [14:57] nessita: I was thinking about marking more projects as being affected by this. But if you prefer that, I'll split it in three. [14:58] elopio, well, I would like to track ocurrences of each kind, becase I'm not sure you have noticed issues downloading a package yet? [14:58] elopio, so since each source is served from very different servers, I was thinking on tracking each separately (icons, packages and system images) [14:58] nessita: yes, http://pastebin.ubuntu.com/12896071/ [14:59] elopio, interesting [14:59] elopio, so are all these errors happening in the same place? [14:59] or is reproduceable in different envs? [14:59] ok, I'll report two more. Should I mark them as affecting a server project in addition to u-d-f? [15:00] nessita: I have seen the system image one in canonistack, and locally. The other two just from my machine, haven't seen them yet in the cloud [15:00] elopio, well, we haven't had any other reports, for example we never got reports from the phone [15:01] elopio, so it may also be a low timeout in the client for TLS timeout? [15:01] nessita, icons is a must though, it is part of the package install, just delivered outside of the package [15:02] elopio, check the date on your workstation ;-) [15:03] nessita: maybe. What I find weird is that it wasn't like this two weeks ago. [15:05] nessita: https://bugs.launchpad.net/ubuntu/+source/goget-ubuntu-touch/+bug/1510136 and https://bugs.launchpad.net/ubuntu/+source/goget-ubuntu-touch/+bug/1510138 [15:05] Launchpad bug 1510136 in goget-ubuntu-touch (Ubuntu) "Failed to install the oem package: TLS handshake timeout" [Undecided,New] [15:05] Launchpad bug 1510138 in goget-ubuntu-touch (Ubuntu) "Failed to download the oem package icon: TLS handshake timeout" [Undecided,New] [15:09] ah, I know why I haven't seen this in the cloud, we don't use u-d-f in the suite :p [15:10] elopio, those icon retries are all snappy bugs though [15:10] elopio, the icon download comes from using the snappy api [15:11] sergiusens: um, that's better. So we can fix it in one single place. [15:11] elopio, you do need to rebuild u-d-f so it is fine to leave the bug open for it [15:12] sergiusens: yes, just marking it as affecting snappy now. You guys can later mark it invalid or duplicate. [15:44] ppisati: hello. I'm starting to learn about dtb and the board leds and was wondering why the heartbeat is not blinking in bbb. [15:47] elopio: you can find which trigger that led is using, looking at the [15:47] elopio: cat /sys/class/leds/$LED/trigger [15:47] elopio: or something along that line [15:47] elopio: kernel version? [15:47] pitti: it's set to none, but the dtb tries to set it to heartbeat. [15:47] I tried echoing heartbeat in there too, which works on debian but doesn't in snappy. [15:48] elopio: do you have the heartbeat module loaded? [15:48] ppisati: I don't know. [15:48] * elopio starts the bbb again. [15:55] ppisati: 4.2.0-16-generic [15:55] not sure how to load the hearbeat module. lsmod doesn't show anything with heartbeat in the name. [15:55] lool: ping [15:55] elopio: cat /sys/class/leds/$LED/trigger [15:55] elopio: and show me the output [15:57] ppisati: http://pastebin.ubuntu.com/12970889/ [15:58] elopio, Chipaca care to take a look at https://code.launchpad.net/~sergiusens/snapcraft/1510160/+merge/275736 ? [15:58] elopio: sudo modprobe ledtrig-heartbeat [15:58] sergiusens: trade ya https://code.launchpad.net/~chipaca/snappy/no-empty-channel-in-update/+merge/275731 [15:58] elopio: and then try to set it again [15:58] ppisati: awesome. [15:59] ppisati: so, how do we tell the bbb oem package to load that module? [16:00] elopio: if the correct trigger is not built into the kernel [16:00] elopio: you can't do that at boot time [16:00] elopio: i'm checking one thing, hold on [16:00] huh, so we need it in the kernel snap instead. [16:01] elopio: we need a kernel with tha option built-in [16:02] elopio: i checked vivid, and we had the same config back then [16:02] elopio: so it's not a regression [16:02] elopio: if you care to open an lp bug, i'll prepare a config patch and send it [16:03] ppisati: sure. in lp:snappy? or in a project for the kernel? [16:04] elopio: just a normal lp bug, we are using the same kernel (ubuntu vs snappy) [16:07] ppisati: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1510165 [16:07] Launchpad bug 1510165 in linux (Ubuntu) "heartbeat module is not loaded" [Undecided,New] [16:08] thank you! [16:10] Chipaca, fair trade [16:22] lool: just let you know i found out why my web app did not run. Seems your sample code use to define a short app name but I use original pom.xml has no the section then lead the web app full name contain version and SNAPSHOT postfix. I can see my webapp run well with right url with long name now. Thanks your sample as good referance. :) [17:20] all rightie... I call it a day - see you all tomorrow! [17:21] sergiusens: mvo: can you kick a new build of u-d-f so it picks up a new snappy? [17:21] mvo: if you have a release checklist, can you add building a new u-d-f and a new webdm to it? [17:23] mvo: if you don't have a release checklist, http://www.amazon.co.uk/dp/1846683149 [17:40] sergiusens: any idea what are we missing here? https://code.launchpad.net/~stephen-stewart/webdm/use-correct-version-for-post-css-bemlinter/+merge/272467/comments/696732/+download [17:43] shuduo: awesome [17:43] shuduo: we can adjust the sample if it's not good enough [17:44] shuduo: the idea is to keep it simple and robust, yet do something relatively advanced [17:44] elopio: i just saw your comment wrt licensing [17:44] elopio: and i know why you're seeing what you're seeing [17:44] Chipaca: and you have a magic solution? [17:44] elopio: yes, yes I do. [17:44] elopio: software is magic, right? [17:45] yes it is. [17:47] elopio: do you have an easy way, from the integration tests, to give the test an actual terminal? [17:48] elopio: if you do, do it. If you don't, do this instead: http://pastebin.ubuntu.com/12971806/ [17:49] Chipaca: I have no idea how to give it a terminal, but I can search. [17:50] elopio: probably easier to include the above diff in your branch [17:50] yes, probably. [17:51] thanks Chipaca. I'll try to write the test with that. [17:51] I'm also having a problem sending y to the stdin, and I'm not sure that fixes it. [17:51] elopio: echo y | your-test ? [17:52] * Chipaca fights down the urge to mention yes `yes` [17:52] Chipaca: I haven't tried that, just writing the the stdin pipe. /me tries [17:53] elopio: ah! [17:53] elopio: let me write it in go for you, gimme a bit [17:53] elopio, phantomJS fails TypeError: 'undefined' is not an object (evaluating 'Backbone.Marionette.ItemView') [17:56] sergiusens: that I know. But how come it is returning undefined? [17:57] elopio, I don't know, beowulf who isn't here should know [17:57] I don't know much about JS much less supporting tools [17:58] sergiusens: I've pinged him. [18:01] getting a 'connection to Mir server failed' error when trying to start clocks example [18:01] elopio: http://paste.ubuntu.com/12971957/ [18:02] snappyc: is mir running and creating its socket? [18:02] snappyc: I think it creates it as /tmp/mir_socket or something like that [18:02] snappyc: you need to make it use that socket [18:07] thanks chipaca. how do i check the log? [18:07] snappyc: mir's logs would be under "snappy service logs mir" [18:08] snappyc: is that the log you mean? [18:08] yeah. how do i get to it? /var/log/... ? [18:09] or is it somewhere else? [18:09] snappyc: snappy service logs mir? [18:10] snappyc, run what Chipaca mentions [18:10] yeh, that's an actual command [18:10] snappy service logs mir [18:11] mvo: we should make snappy a snap [18:12] how to execute ur script? [18:12] :{ [18:14] sergiusens: beowulf is on parental leave. [18:14] $snappy service logs mir --returns -- unknown command "service". Please specify one command of: booted, build, config, ... [18:14] Chipaca: es, I think we should [18:15] snappyc: your snappy is quite old [18:15] snappyc: update that, first [18:15] snappyc: "snappy update" should do that for you [18:15] ubuntu-personal/rolling/edge [18:16] personal! is that supported in any way right now? [18:16] * Chipaca has no idea [18:16] lol. :{ [18:20] the personal target is not supported [18:20] the mir snap should not be installable on snappy personal either [18:20] snappyc: what are you trying to do? [18:20] mir snap installed. [18:20] sergiusens: i just assumed it was from the mir snap, it might be the system mir on personal? (dunno) (because, see: dunno) [18:21] * sergiusens doesn't want to discuss mir dev as he is not the right person [18:22] * sergiusens uses his remaining mana to invoke kgunn [18:22] snappyc: snappy ubuntu personal is still a way in the future, fwiw [18:22] snappyc: i don't know what you're trying to do, but you're probably better off using snappy ubuntu-core [18:22] snappyc: are there instructions out there that point at -personal, or did you do that yourself? [18:23] mvo, Chipaca elopio are you already on xenial? [18:23] sergiusens: nop. [18:23] * sergiusens updated to wily yesterday and is considering going all in [18:23] sergiusens: updating now [18:24] just doing it myself [18:24] sergiusens: not yet, but soon :-D [18:24] sergiusens: at least on my main box [18:24] snappyc: ah. Good job on figuring out the bits :) [18:24] snappyc: but yeah, that's very much not ready for anything right now [18:25] snappyc: it only gets updated when people want to take a poke to see how things would look [18:28] Chipaca: can i install snappy desktop on snappy core? [18:29] snappyc: there isn't a "snappy desktop" [18:29] :{ [18:30] snappyc: ? [18:38] Chipaca: any way to export installed app? [18:38] *snap [18:39] snappyc: you're going to have to explain that one to me, i'm afraid [18:41] :{ [18:41] um [18:41] how is ubuntu-core updated? [18:41] can i update it manually? [18:41] I tried to change the time on my snappy system with the command "sudo date 102613392015.00" but the new date will not stick [18:42] am I doing something wrong? [18:43] snappyc: depends what you mean by "manually". "snappy update" updates the system, and you can run that command yourself [18:43] snappyc: it's also run automatically, so you don't have to [18:44] chipaca: snappy update doesnt work. so i need to update manually [18:44] snappyc: on your snappy personal? [18:44] yeah [18:44] how do i download ubuntu-core.snap? [18:44] wiggleworm: $ sudo date --set="Sun Apr 04 17:43:26 UTC 2015" [18:44] snappyc: you can't update from personal to core, they are different things [18:44] personal has core [18:45] Chipaca: is there a ubuntu-core.snap i can download? [18:47] snappyc: I don't think so, no. [18:47] snappyc: the base system is not (yet) itself a snap [18:47] :{ [18:48] * kgunn reads scrollback [18:48] snappyc: and i'm not sure in what sense "personal has core", but whatever you're trying to do, it probably won't work [18:48] snappyc: what are you running this on? a vm? [18:48] yep [18:48] snappyc: so why not download the ubuntu-core image? [18:49] i did.. but if i can update ubuntu-personal... id like to stick with it [18:50] snappyc: why do you need ubuntu-personal ? [18:50] no reason [18:52] snappyc: sorry that did not work either (time change) [18:53] Chipaca: thx for the help [18:54] wiggleworm: when you say "will not stick" [18:54] wiggleworm: what do you mean? [18:55] Chipaca: stick as in - after I issue the change date command I check the date [18:55] time and its not changed [18:55] wiggleworm: ok, so, two things [18:56] wiggleworm: if you don't say --set or -s, date will just interpret the date you give it and print it out [18:56] afaik at least [18:56] wiggleworm: date --help might help [18:58] command that I am running "sudo date --set="Sun Apr 04 17:43:36 UTC 2015" [18:58] not the actual date, just an example [18:59] Chipaca: thanks, I tried date --help first - still no luck [18:59] wiggleworm: also, note that snappy runs a time sync daemon [18:59] wiggleworm: so maybe use timedatectl [18:59] Chipaca: what is that? time sync daemon [19:00] wiggleworm: systemd-timesyncd.service [19:00] wiggleworm: on a full ubuntu system, man systemd-timesyncd [19:00] wiggleworm: also man systemd-timedated [19:01] Chipaca: Sorry I meant what does it do - I am guessing it syncs with some time server? [19:02] wiggleworm: yes [19:10] wiggleworm: although i think by default no server is configured, so it'll take them from dhcp if given [19:10] wiggleworm: in any case, to set the time, timedatectl [19:11] wiggleworm: thank you for asking this, btw, because i didn't know :) [19:12] Chipaca: thank you - I just maned the timedatectl and I see how I should do it. I also figured out that I needed to change the time zone - :) after that my time looked right [19:12] wiggleworm: hah [19:13] wiggleworm: that's one you set via snappy, fwiw [19:13] defaults to etc/utc [19:55] mvo: is there an easy way to get the tools ppa rebuild for x? [19:59] Chipaca: all of it? I don't think so, may I ask for the use-case? [20:01] mvo: dunno about "all of it" :) [20:01] mvo: i mean, i'm moving to x [20:01] mvo: should i worry about that ppa? :) [20:05] Chipaca: do we have ppa builds for X already? [20:05] mvo: no [20:05] Chipaca: but sorry, let me reply properly [20:06] Chipaca: I think its fine to not worry about the X ppa for now, we just push everything into X proper [20:06] problem solved :) [20:06] mvo: \o/ [20:07] If I deploy a snap with a service and a binary (two separate files), is the there something special I need to do to let the service run the binary? [20:07] because currently /apps/bin is not in the PATH of the service [20:08] tasdomas: you don't want to run things from /apps/bin from inside a service [20:08] tasdomas: double-confinement won't work :) [20:08] tasdomas: instead, do [20:09] tasdomas: $SNAP_APP_PATH/bin/whatever [20:09] Chipaca, thanks! [20:09] np [20:17] augh [20:18] ogra_: looks like we messed something up :( [20:26] Chipaca: whats broken? [20:26] mvo: replied to jdstrand's mail in reply to your announcement [20:26] mvo: not a regression though :) [20:34] Chipaca: hrm, hrm, I need to look why the writable path change did not land, this is terrible [20:35] mvo: no, it's not terrible [20:35] mvo: it's just bad [20:35] mvo: relax a little :) [20:36] mvo: or maybe i just read too much panic in your "this is terrible" [20:38] Chipaca: /me relaxes [20:38] Chipaca: well, maybe its too stronly worded, sorry for that [20:38] mvo: i mean, yes it's bad, but it's not a regression and has an easy fix [20:39] yikes [20:39] I didn't mean to stress people out [20:39] * jdstrand just viewed it as a feature that didn't land [20:39] hence the email rather than ping [20:39] jdstrand: yeah, its fine, no worries [20:40] sorry for using too strong language [20:40] its just that I want things to be perfect :) [20:40] well, if you typed it, you might be feeling it [20:40] if it they are not "its terrible" [20:40] * jdstrand hugs mvo :) [20:41] * mvo hugs jdstrand and Chipaca [20:42] * Chipaca hugs everybody [20:42] * jdstrand hugs Chipaca :) [20:45] * mvo will backport the fix tomorrow and calls it a day [21:23] should I be using snappy and possibly docker? I have a small project which I was going to use a raspberry pi to capture images from an ip camera and rsyncing them to a remote server. Or should I just go for 14.04