ePierre | hi there! | 06:10 |
---|---|---|
ePierre | I am trying Snappy on a KVM virtual machine, and I would like to have more free space in the image HDD in order to install more snaps. How can I do that? I tried to follow this guide: http://itsignals.cascadia.com.au/?p=28 but when I start gparted live, I have a kernel panic message | 06:11 |
=== chihchun_afk is now known as chihchun | ||
dholbach | good morning | 07:38 |
zyga | good morning, hey dholbach :) | 07:38 |
dholbach | hey zyga | 07:39 |
ePierre | morning zyga :) | 07:45 |
ePierre | and dholbach :) | 07:45 |
ePierre | zyga, you may have an idea about this: | 07:45 |
ePierre | I am trying Snappy on a KVM virtual machine, and I would like to have more free space in the image HDD in order to install more snaps. How can I do that? I tried to follow this guide: http://itsignals.cascadia.com.au/?p=28 but when I start gparted live, I have a kernel panic message | 07:45 |
dholbach | hi ePierre | 07:46 |
zyga | ePierre: just make a bigger disk | 07:57 |
zyga | ePierre: then snappy expands on first boot | 07:57 |
zyga | ePierre: my recommendation is to use -snapshot all the time | 07:57 |
zyga | ePierre: so each reboot gives you identical environment | 07:58 |
zyga | ePierre: then script any post-boot activities, | 07:58 |
zyga | ePierre: this is great for testing | 07:58 |
ePierre | zyga, let me try this | 08:01 |
ePierre | zyga, hm so I followed everything up to step 8 of that page: http://itsignals.cascadia.com.au/?p=28 ; which supposedely makes my .img file much bigger | 08:05 |
ePierre | zyga, yet when I boot I still have 0 free space | 08:05 |
fgimenez | good morning | 08:06 |
ePierre | Filesystem Size Used Avail Use% Mounted on | 08:06 |
ePierre | /dev/sda6 471M 461M 0 100% /home | 08:06 |
ePierre | zyga, to follow up, my disk is bigger, but snappy won't expand automagically | 08:42 |
ePierre | /dev/sda6 6569984 19427327 12857344 6.1G Linux filesystem | 08:42 |
zyga | ePierre: try latest images, I think this happens automatically now (it did for me) | 08:48 |
JamesTait | Good morning all; happy Thursday, and happy Wright Brothers Day! ✈ | 09:41 |
fgimenez | Chipaca, hello! i'm getting http://paste.ubuntu.com/14072304/ in 15.04/edge, don't know if you are the right person to ask about this | 10:15 |
Chipaca | fgimenez: that's interesting | 10:16 |
Chipaca | fgimenez: what happens if you disable autopilot using config? | 10:16 |
fgimenez | Chipaca, sorry, how can i do that? | 10:17 |
Chipaca | fgimenez: rolling: echo 'config: {ubuntu-core: {autoupdate: off}}' | sudo snappy config ubuntu-core - | 10:17 |
Chipaca | fgimenez: 15.04: echo 'config: {ubuntu-core: {autopilot: off}}' | sudo snappy config ubuntu-core - | 10:18 |
Chipaca | fgimenez: in fact | 10:18 |
Chipaca | fgimenez: echo 'config: {ubuntu-core: {autopilot: off, autoupdate: off}}' | sudo snappy config ubuntu-core - | grep auto | 10:18 |
Chipaca | fgimenez: should work for both | 10:18 |
Chipaca | for now :-) | 10:18 |
Chipaca | (we're going to break this on rolling at some point in the new year) | 10:19 |
fgimenez | Chipaca, "echo 'config: {ubuntu-core: {autopilot: off}}' | sudo snappy config ubuntu-core -" works (config doesn't give the "invalid unit status error" afterwards) but i'm not able to reenable it http://paste.ubuntu.com/14072341/ | 10:22 |
Chipaca | fgimenez: echo 'config: {ubuntu-core: {autopilot: off, autoupdate: on}}' | sudo snappy config ubuntu-core - | grep auto | 10:23 |
Chipaca | fgimenez: or, try mask/unmask instead of enable/disable if you need to do it through systemctl | 10:23 |
Chipaca | fgimenez: (fwiw i think this is new, and if the change broke stuff you should blame mvo) | 10:23 |
Chipaca | fgimenez: (mostly because blaming mvo is always fun) | 10:24 |
Chipaca | ahm | 10:24 |
Chipaca | fgimenez: i meant to change both 'off's to 'on's, not just one, fwiw | 10:24 |
Chipaca | elopio: wrt https://github.com/ubuntu-core/snappy/pull/232 | 10:25 |
Chipaca | elopio: i don't care whether you use a url or python to make sure the json is valid, but one of the two needs to happen | 10:25 |
mvo | Chipaca, fgimenez: hm, I think you now need to mask/unmask | 10:27 |
Chipaca | mvo: the 'invalid unit status' thing is probably a bug though | 10:28 |
fgimenez | Chipaca, mvo, ok thanks, to continue using systemctl it seems that "sudo systemctl disable snappy-autopilot.timer" is now equivalent to "sudo systemctl disable snappy-autopilot.timer && sudo systemctl mask snappy-autopilot.timer", right? | 10:28 |
Chipaca | mvo: makes me think some people will see breakage on upgrade to this | 10:28 |
Chipaca | fgimenez: no, just mask | 10:29 |
mvo | Chipaca: oh, yeah, for upgrades it is indeed a bug. well, maybe, I think part of the problem is that a disable will not survie a reboot because it gets copied up again :) | 10:29 |
fgimenez | Chipaca, mvo ok, just mask works great, thanks! | 10:32 |
Chipaca | mvo: good point :) | 10:32 |
mvo | Chipaca: sorry for being a smartass, I'm just a bit tired this morning | 10:34 |
Chipaca | mvo: your transgression has been duly noted | 10:36 |
stevebiscuit|awa | Chipaca, zyga, elopio: I've noted the decision to not separate the REST-wrangling client code | 10:50 |
Chipaca | stevebiscuit|awa: *for now* :-) yes | 10:50 |
=== stevebiscuit|awa is now known as stevebiscuit | ||
stevebiscuit | Chipaca: if others want to contribute to fleshing out the client lib then I'm all for it :) | 10:52 |
stevebiscuit | Chipaca: ripping out all the snappy code from webdm will take a while but it'll be worth it | 10:53 |
Chipaca | stevebiscuit: i'll take another pass at your branches in a bit | 10:53 |
Chipaca | --- there are a few others ahead of yours in my pile, is all :-) | 10:53 |
stevebiscuit | Chipaca: appreciated! | 10:53 |
=== retrack is now known as Guest58165 | ||
Chipaca | stevebiscuit: you need two +1s to land on master though | 11:43 |
Chipaca | stevebiscuit: (just in case you weren't aware) | 11:43 |
Chipaca | ah! you already have them | 11:44 |
Chipaca | go go go | 11:44 |
Chipaca | merged | 11:44 |
stevebiscuit | Chipaca: gentleman and a scholar, ta! | 11:44 |
* Chipaca straightens his monocle | 11:45 | |
Chipaca | indeed | 11:45 |
stevebiscuit | :D | 11:45 |
=== retrack is now known as Guest58144 | ||
stevebiscuit | so the concept of a port for a snap service is going to go away? | 12:44 |
stevebiscuit | "fyi TODO we have removed the concept of ports." << from code review by Sergio | 12:46 |
Chipaca | stevebiscuit: i think sergio misused the word 'concept' there | 12:46 |
Chipaca | the *concept* didn't go away | 12:46 |
Chipaca | the plan is to expose it in a different way, probably | 12:46 |
Chipaca | and the underlying problem of actually doing something with the intent expressed in the ports declaration is still present | 12:47 |
Chipaca | but the change should make it feasible to do that something | 12:47 |
Chipaca | we're talking post-16.04 stuff now | 12:47 |
stevebiscuit | Chipaca: ok, so the data will still be available | 12:48 |
pedronis | it's going to be trough capabilities afaiu | 12:50 |
Chipaca | stevebiscuit: if i remember correctly, it'll be an attribute of a capability | 12:50 |
Chipaca | stevebiscuit: but that's a zyga question i think | 12:50 |
Chipaca | he was paying more attention than i | 12:50 |
Chipaca | (because he's the caps guy) | 12:51 |
stevebiscuit | ah, so the service will need the capability to listen on a specif port, makes sense | 12:51 |
=== davidcalle_ is now known as davidcalle | ||
=== davidcalle__ is now known as davidcalle_ | ||
jdstrand | beuno, pindonga: hey, I'm getting a 503 when trying to install hello-world in a vm | 14:39 |
beuno | jdstrand, yes | 14:39 |
jdstrand | hello-world failed to install: received an unexpected http response code (503) when trying to download https://public.apps.ubuntu.com/anon/download/canonical/hello-world.canonical/hello-world.canonical_1.0.18_all.snap | 14:39 |
beuno | prodstack going through some heavy turbulance atm | 14:39 |
jdstrand | ok | 14:39 |
kyrofa | elopio, I'm sorry I completely lost track of time | 14:40 |
kyrofa | elopio, do you have power today? :P | 14:40 |
=== frecel_ is now known as frecel | ||
frecel | can someone tell me what oem snaps are supposed to be? | 16:09 |
ogra_ | currently they define the hardware and ship all bootloader related bits | 16:14 |
frecel | ogra_: can I put a kernel in an oem snap? | 16:52 |
ogra_ | no | 16:53 |
ogra_ | today you need a device tarball ... | 16:53 |
ogra_ | soon this will all change ... | 16:53 |
ogra_ | and you need a kernel snap ... | 16:53 |
ogra_ | the oem snap is going away and the bootloader will go into that kernel snap ... | 16:53 |
ogra_ | the HW description moves to a so-called gadget snap then | 16:54 |
frecel | oh interesting | 16:54 |
ogra_ | today you still need a device tarball and an oem snap though | 16:54 |
frecel | I'm trying to get an image for intel edison | 16:54 |
frecel | ogra_: is there a documentation on what a proper oem snap is supposed to look like? | 16:56 |
frecel | or should I just look at the pi2 snap and try to figure it out | 16:56 |
ogra_ | https://developer.ubuntu.com/en/snappy/guides/oem/ | 16:59 |
ogra_ | http://bazaar.launchpad.net/~snappy-dev/snappy-hub/snappy-systems/files | 17:00 |
ogra_ | have fun :) | 17:00 |
kyrofa | jdstrand, mvo my take on a fix for bug #1466234: https://github.com/ubuntu-core/snappy/pull/264 | 17:08 |
ubottu | bug 1466234 in ubuntu-snappy (Ubuntu) "Apparmor denial for access to SNAP_APP_USER_DATA_PATH as root" [Critical,Triaged] https://launchpad.net/bugs/1466234 | 17:08 |
kyrofa | Well, as far as snappy itself goes anyway | 17:09 |
frecel | ogra_: I think https://developer.ubuntu.com/en/snappy/guides/porting/ should link to that oem guide | 17:14 |
ogra_ | true | 17:15 |
jdstrand | mvo: hi! | 17:26 |
jdstrand | mvo: can you think of why my last comment in https://github.com/ubuntu-core/snappy/pull/264 is happening? | 17:27 |
jdstrand | mvo: ubuntu-core/rolling/edge doesn't have the newest ubuntu-core-security-* packages installed | 17:27 |
jdstrand | mvo: it has 16.04.2, which is from Nov 12, but 16.04.7 is available in xenial | 17:29 |
mvo | jdstrand: is that maybe a xenial vs xenial-proposed thing? without having looked at it | 17:32 |
jdstrand | mvno. 16.04.7 is in xenial release | 17:32 |
jdstrand | err | 17:32 |
jdstrand | mvo: no 16.04.7 is in xenial release | 17:33 |
jdstrand | mvo: since Dec 2 | 17:33 |
mvo | hmmm | 17:33 |
ogra_ | http://cdimage.ubuntu.com/ubuntu-core/daily-preinstalled/current/xenial-preinstalled-core-amd64.manifest | 17:33 |
ogra_ | i see 16.04.7 | 17:33 |
jdstrand | well that's weird | 17:34 |
=== xnox is now known as xnox_2016 | ||
jdstrand | my vm doesn't have it | 17:34 |
jdstrand | and it claims it is up to date | 17:34 |
jdstrand | mvo: ok, seems to be a local thing | 17:34 |
* jdstrand blows it away and regenerates | 17:34 | |
ogra_ | jdstrand, and it actually landed on dec 3rd in the image http://people.canonical.com/~ogra/core-image-stats/20151203.changes | 17:36 |
mvo | jdstrand: scary - the buildlog also has it | 17:36 |
ogra_ | another delta generation issue ? | 17:37 |
mvo | it might be :/ | 17:37 |
=== dpm is now known as dpm-afk | ||
=== joc_ is now known as joc|away | ||
=== joc|away is now known as joc | ||
mine_field | what is this? o_O https://uappexplorer.com/app/canonical-linux-pc.canonical | 19:16 |
ogra_ | well, what the description says :) | 19:24 |
mine_field | no description :D | 19:25 |
mine_field | i understand nothing :( | 19:25 |
mine_field | but is very late so i must sleep now :P | 19:25 |
ogra_ | it is the new kernel snap | 19:26 |
mine_field | oh 4.3.0-2-2 | 19:26 |
mine_field | got it :d | 19:26 |
ogra_ | (currently snappy still uses the system-image technology from the phone for upgrades ... this will change ... all parts of the OS will become snap packages) | 19:26 |
mine_field | skynet 0.1 got it :D | 19:27 |
mine_field | thanks! good night all! ogra_ | 19:27 |
Rob1507 | Hi there. Can anyone help me with something now? | 19:53 |
kyrofa | jdstrand, I just made a new VM off of 15.04 stable and I'm still getting apparmor denials to /root/apps. Did I misunderstand you when I thought that would be fixed? | 19:59 |
kyrofa | Rob1507, ask away! | 20:06 |
Rob1507 | kyrofa, I want to snap NodeJS app but I get error | 20:07 |
Rob1507 | Failed doing build for webchat: Command '['/bin/sh', '/tmp/tmp36flb68n', 'npm', 'install', '-g']' returned non-zero exit status 2 | 20:07 |
kyrofa | Rob1507, using snapcraft? | 20:08 |
Rob1507 | kyrofa, yes | 20:08 |
kyrofa | Rob1507, can you put the code somewhere so I can take a look? And make a pastebin of the entire output from snapcraft | 20:09 |
kyrofa | Rob1507, also, what version of ubuntu are you running? | 20:09 |
jdstrand | kyrofa: is that for an existing app? | 20:09 |
Rob1507 | 14.04 LTS | 20:09 |
kyrofa | jdstrand, no just a test case (roscore running as a service) | 20:09 |
kyrofa | jdstrand, but I can tar it up if you want it | 20:10 |
Rob1507 | kyrofa, I am snapping one of the examples - shout. Send code or not needed? | 20:10 |
kyrofa | jdstrand, or simplify it | 20:10 |
jdstrand | kyrofa: what is the output of dpkg -l|grep ubuntu-core-security-apparmor | 20:10 |
kyrofa | Rob1507, you're saying a snapcraft example isn't snapping correctly? | 20:11 |
kyrofa | jdstrand, ubuntu-core-security-apparmor 15.04.12~ppa14 | 20:11 |
Rob1507 | http://pastebin.com/ev5Esfv4 | 20:11 |
Rob1507 | kyrofa, this is output | 20:11 |
kyrofa | jdstrand, just downloaded the 15.04 stable image and ran snappy update followed by a reboot, no more updates available | 20:11 |
jdstrand | kyrofa: that should have it. did you install your snap prior to snappy update/reboot? (that is what I was trying to ask before) | 20:12 |
kyrofa | jdstrand, oh I see. And no, I updated and rebooted before I tried installing anything | 20:12 |
jdstrand | kyrofa: what is the output of grep HOME /usr/share/apparmor/easyprof/templates/ubuntu-core/15.04/default? | 20:13 |
kyrofa | jdstrand, http://pastebin.ubuntu.com/14077787/ | 20:13 |
kyrofa | Rob1507, are you using the snappy-tools ppa to get snapcraft on trusty? | 20:14 |
jdstrand | ok, that's good | 20:15 |
jdstrand | kyrofa: can you paste the denial? | 20:15 |
kyrofa | jdstrand, http://pastebin.ubuntu.com/14077817/ | 20:16 |
kyrofa | jdstrand, hmm | 20:16 |
kyrofa | jdstrand, actually that looks like a snappy problem huh | 20:16 |
kyrofa | jdstrand, it's trying to create /root/apps | 20:16 |
jdstrand | kyrofa: indeed | 20:17 |
kyrofa | jdstrand, I'm sorry, I just saw a denial on /root and came to you, I should have investigated further | 20:17 |
kyrofa | jdstrand, so my PR should probably fix that, eh? | 20:17 |
kyrofa | jdstrand, hmm... wonder how to best do that in a systemd conf | 20:17 |
jdstrand | it should, but I thought the launcher was already doing that | 20:18 |
jdstrand | oh, this is in a service, not a binary? | 20:18 |
kyrofa | jdstrand, yes a service | 20:18 |
jdstrand | hmmm | 20:18 |
jdstrand | I'm not sure the best way to handle that | 20:19 |
jdstrand | kyrofa: I think most would probably say that a service should be using SNAP_APP_DATA_PATH and not SNAP_APP_USER_DATA_PATH | 20:19 |
kyrofa | jdstrand, I agree, but that's not always possible | 20:19 |
kyrofa | jdstrand, not to mention having an environment variable point to a dir that doesn't exist seems wrong anyway | 20:20 |
kyrofa | jdstrand, some apps just log to HOME/something, and the person making the .snap can't always change that | 20:21 |
jdstrand | it does | 20:21 |
jdstrand | does it make sense for SNAP_APP_USER_DATA_PATH to point to SNAP_APP_DATA_PATH? | 20:21 |
jdstrand | HOME/something is weird for a service though | 20:21 |
kyrofa | jdstrand, agreed, in the case of a service, I think that would be fine. That's more of an mvo question though | 20:22 |
jdstrand | yes | 20:22 |
kyrofa | jdstrand, okay thanks for trouble-shooting with me :) | 20:23 |
jdstrand | np | 20:25 |
dPragmatist | snapcraft init | 20:34 |
kyrofa | dPragmatist, that may not result in what you'd hoped in this window | 20:36 |
dPragmatist | ha, too many open windows kyrofa | 20:40 |
kyrofa | dPragmatist, haha, it's happened to all of us | 20:41 |
kyrofa | mvo, are you around? It might be a bit late for you | 20:43 |
=== retrack is now known as Guest35077 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!