|
=== bjf_ is now known as bjf |
|
=== slangase` is now known as slangasek |
|
=== rcj` is now known as rcj |
|
[06:21] <dholbach> good morning |
|
[07:08] <fgimenez> good morning |
|
[07:09] <mvo_> hey fgimenez, good morning |
|
[07:10] <fgimenez> hi mvo_ |
|
[07:10] <mvo_> no new good image yet that fully does the upgrade,rollback,upgrade dance, but except for this the most recent one should be good to test |
|
[07:11] <mvo_> and a new image is in the works that hopefully fixes upgrade/rollback/upgrade/rollback |
|
[07:11] <p_lorenz> hi - is there a way to redirect the snappy cli tool to an own server infrastructure? i work for a company which would like to host own snaps and system updates |
|
[07:22] <fgimenez> mvo_, ok thanks, i'll execute the tests against the latest one |
|
[07:32] <clobrano> good morning |
|
[07:32] <clobrano> I was looking for info about installing udev rules in snappy (other than using hw-assing). Any suggestion? :) |
|
[07:33] <clobrano> a link to any kind of documentation will be great |
|
[08:04] <Chipaca> mo'in |
|
[08:05] <Chipaca> clobrano: i don't think that exists; what's your use-case? (that is: what do you want to achieve?) |
|
[08:06] <clobrano> Chipaca, in our application, we often create symlink to device nodes that respect some regex |
|
[08:07] <clobrano> like matching vid/pid pair |
|
[08:08] <clobrano> I saw that the example here https://developer.ubuntu.com/en/snappy/guides/appliance-builder-guide-webcam/ that include a udev rule in the package.yaml, is there something already available for other udev rules? |
|
[08:10] <Chipaca> clobrano: I don't think there is, at the app level, something for that. I suspect what you want to do goes in a gadget snap, not an app snap |
|
[08:10] <Chipaca> clobrano: ogra_ probably knows more, but he was up very late fighting with some regressions in the next release :) |
|
[08:11] <clobrano> Chipaca: ok, do not worry :) you already gave me a starting point; I didn't know about "gadget snap" :D |
|
[08:22] <mvo_> fgimenez: r165+ should be ok, I did a 165->166->165->166 successfully now |
|
[08:23] <p_lorenz> Chipaca: thanks for your mail :) which nickname does Martín have in IRC? |
|
[08:23] <fgimenez> mvo_, ok thanks i'll try that now |
|
[08:23] <mvo_> thanks |
|
[08:23] <Chipaca> p_lorenz: beuno |
|
[08:23] * mvo_ tries a full upgrade instead of delta now too |
|
[08:24] <p_lorenz> Chipaca: thanks a bunch! |
|
[08:25] <Chipaca> p_lorenz: no worries |
|
[08:29] <zyga> mvo_: I'd like to add snapcraft --version |
|
[08:29] <zyga> mvo_: would you merge that? |
|
[08:31] <Chipaca> zyga: snapcraft, or snappy? |
|
[08:31] <mvo_> zyga: sure, unless sergio disagrees thats fine |
|
[08:31] <Chipaca> zyga: if snapcraft, ask sergiusens :) |
|
[08:31] <Chipaca> zyga: also probably not today |
|
[08:31] <zyga> Chipaca: snapcraft |
|
[08:31] <zyga> Chipaca: what's going on today? |
|
[08:31] <Chipaca> zyga: busy busy busy |
|
[08:31] <Chipaca> zyga: like yesterday, but more so |
|
[08:31] <mvo_> Chipaca: more so? *meeeh* |
|
[08:32] * ogra_ phears that |
|
[08:32] <Chipaca> mvo_: same pressure, more tiredness -> more mistakes -> moar busy |
|
[08:32] <ogra_> moaning :) |
|
[08:32] <zyga> Chipaca: I know the feeling |
|
[08:33] <ogra_> clobrano, there is no way to "ship" a udev rule, the rule you see in the example is generated by the hw-assign command |
|
[08:34] <mvo_> Chipaca: logical. flawlessly logical… |
|
[08:35] <Chipaca> mvo_: https://www.youtube.com/watch?v=THNPmhBl-8I&t=1m55s |
|
[08:35] * Chipaca links to the punchline of a very long joke |
|
[08:35] * Chipaca is obviously also not in tip-top form today :) |
|
[08:38] <mvo_> Chipaca: hehe |
|
[08:53] <clobrano> ogra_: ok, so is there a way to add custom udev rule? Or is it something yet to be added to snappy? |
|
[08:54] <JamesTait> Good morning all; happy Guacamole Day! 😃 |
|
[08:55] <ogra_> i dont think that is planned ... the hw-assign tool will be expanded and perhaps also changed ... and you can define a device (and a snap it uses) in the gadget (today it is called oem) snap |
|
[08:57] <clobrano> ogra_: uhm, I understand. Well, it will be great if hw-assign would be flexible enough to allow symlinks too :) |
|
[08:59] <ogra_> clobrano, worth filing a bug i guess :) so the implementers know about that |
|
[09:01] <clobrano> ogra_: I was more in the mood of contributing if possible :D, btw I will start with the feature request ;), Thank you, I understood that you all are very busy those days :) |
|
[09:02] <ogra_> clobrano, hw-assign is maintained by our security team, talk to tyhicks or jdstrand once they are around (i think both are in US timezones) |
|
[09:03] <ogra_> i agree it would be nice to be able to define a symlink name via hw-assign so you can use self picked device names in your snap |
|
[09:05] <clobrano> ogra_: yep, that was what I thought :) |
|
=== vrruiz_ is now known as rvr |
|
[09:39] <davmor2> hey guys in wily trying to create a snappy personal image is failing http://paste.ubuntu.com/12425581/ I'm assuming that the issue is it doesn't wait for everything to be downloaded before it starts the image creation process. |
|
[09:57] <ogra_> davmor2, it can do that in parallel ... what it creates is an empty image, then it copies stuff inside |
|
[09:57] <ogra_> so the order shouldnt matter |
|
[09:58] <ogra_> do you have enough free diskspace ? IIRC personal is quite big |
|
[09:59] <davmor2> ogra_: 756GB of free space and it installed fine on vivid only had this issue since I tried using it on wily |
|
[10:00] <ogra_> ah, u-d-f 0.30 landed yesterday ... that might have broken it ... try going back to 0.29 |
|
[10:01] <ogra_> ricmm, ^^^ :) |
|
[10:01] <ogra_> might be that the image needs some extra changes to work with the new u-d-f |
|
[10:02] <ogra_> blame sergio ;) |
|
[10:05] <ricmm> ah :) indeed |
|
[10:05] <ricmm> davmor2: sadly will need to wait for sergio to wake up |
|
[10:24] <ricmm> davmor2: wait that error is something else |
|
[10:24] <ricmm> davmor2: reboot your machine ;) |
|
[10:24] <ricmm> and try again |
|
[10:25] <davmor2> ricmm: will do after this meeting |
|
=== zbenjamin_ is now known as zbenjamin |
|
[11:48] <dholbach> sergiusens_, can we turn off the plainbox tests which need internet access during the build? |
|
[11:48] <dholbach> https://code.launchpad.net/~dholbach/snapcraft/1496363/+merge/271285 |
|
[11:58] <sergiusens_> dholbach, yeah, we need to make those autopkg ones |
|
[12:26] <Mikaela> dholbach: you might be interested in https://freenode.net/sasl |
|
[12:43] <dholbach> mh |
|
[12:48] <dholbach> sergiusens_, in which cases is the wiki plugin going to be used? |
|
[12:49] <sergiusens_> dholbach, when used in 'after' and the ref'ed part is not there |
|
[12:50] <dholbach> sergiusens_, ok... I was just surprised how often during the build wiki.u.c was trying to be opened: http://people.canonical.com/~dholbach/tmp/snapcraft_0.2_amd64.build |
|
[12:50] <sergiusens_> dholbach, or when a part is there but doesn't specify a 'type'; for which the full part definition will be pulled from the wiki and the local definition would overlay it (in python part_from_wiki.update(part_local) |
|
[12:50] <sergiusens_> dholbach, hmm, should be only once; maybe I can initialize it earlier |
|
[12:51] <dholbach> sergiusens_, shall I file a bug? |
|
[12:51] <sergiusens_> dholbach, sure |
|
[12:51] <dholbach> cool will do |
|
=== sergiusens_ is now known as sergiusens |
|
[12:52] <sergiusens> Chipaca, what is the python way for http://golang.org/pkg/sync/#Once.Do ? |
|
[12:52] <dholbach> https://bugs.launchpad.net/snapcraft/+bug/1496381 |
|
[12:52] <ubottu> Launchpad bug 1496381 in Snapcraft "wiki.u.c is opened a lot during the build" [Undecided,New] |
|
[12:54] <Chipaca> sergiusens: a "once" decorator |
|
[12:55] <sergiusens> Chipaca, of course :-) |
|
[12:58] <Chipaca> sergiusens: http://pastebin.ubuntu.com/12426544/ |
|
[13:03] <Chipaca> sergiusens: definitely not this: http://pastebin.ubuntu.com/12426580/ |
|
[13:03] <sergiusens> Chipaca, no, I prefer easily readable code |
|
[13:03] <Chipaca> :) |
|
[13:03] <sergiusens> thank you very much |
|
[13:03] <sergiusens> :-) |
|
[13:04] <sergiusens> Chipaca, the write less characters thing is only for play :-P |
|
[13:04] <Chipaca> sergiusens: disadvantage of the decorator is that the decorated function won't complain about bad number of args |
|
[13:04] <Chipaca> sergiusens: that is |
|
[13:04] <Chipaca> if f(a) |
|
[13:04] <Chipaca> and the first time you call f(1) |
|
[13:04] <Chipaca> and the second, f(1,2) |
|
[13:04] <Chipaca> the second will just work |
|
[13:05] <sergiusens> Chipaca, that's fine, it's an internal decorator |
|
[13:05] <Chipaca> k |
|
[13:06] <jdstrand> ogra_: fyi, I wouldn't say the security team is maintained by us-- we didn't even implement it :) we did have input to get something going, but others are refining it. the only thing we care about is that hw happens, somehow, in a separate step |
|
[13:07] * ogra_ notes down ... the security team isnt maintained by the security team |
|
[13:07] * ogra_ grins |
|
[13:07] <jdstrand> as for symlinks, that will be interesting because apparmor resolves symlinks (it must) |
|
[13:07] <jdstrand> haha |
|
[13:07] <jdstrand> s/the security team/hw-assign/ :) |
|
[13:07] <ogra_> you mean hw-assign i guess :) |
|
[13:08] <ogra_> jdstrand, well i think it might be convenient to allw hw-assign to have options like --symlink-name ... or --exec-script (which would be a script running under confinement inside your snap indeed) |
|
[13:08] <jdstrand> but actually, we are using cgroups for hw-assign currently |
|
[13:09] <ogra_> you dont generate the udev rule anymore ? |
|
[13:09] <jdstrand> I'm happy to review the mp |
|
[13:09] <jdstrand> there are different parts of it |
|
[13:09] <clobrano> jdstrand: hi, I was the one interested in symlinks :D, any hope to have the ability to create some? :) |
|
[13:09] <jdstrand> it has not changed since april |
|
[13:10] <jdstrand> oem snaps (gadgets) may provide udev rule snippets via the 'assign' yaml which can tag devices |
|
[13:10] <jdstrand> that tag is then used to add the device to a cgroup |
|
[13:12] <clobrano> jdstrand: yep, I saw something in the webcam example |
|
[13:12] <jdstrand> otoh, I don't see why snappy in this situation: 'snappy hw-assign foo /dev/symlink' couldn't resolve the symlink and tag the device in the same manner |
|
[13:13] <clobrano> interesting |
|
[13:13] <jdstrand> someone would 'just' need to implement that |
|
[13:13] <jdstrand> (fyi, you can read about the current implementation here: https://wiki.ubuntu.com/SecurityTeam/Specifications/SnappyConfinement#Cgroups) |
|
[13:14] <clobrano> how open is the project for contribution from outsiders? :) |
|
[13:14] <clobrano> I mean, I can try and do that? |
|
[13:14] <jdstrand> I suggest filing a bug. Note, how hw assignment works is bieing redone aiui, but I don't know who is doing it |
|
[13:14] <jdstrand> clobrano: it is completely open :) |
|
[13:15] <clobrano> Nice to hear that, I've already filed a bug (https://bugs.launchpad.net/snappy/+bug/1496319) and I think I'll have a look at that feature |
|
[13:15] <ubottu> Launchpad bug 1496319 in Snappy "Could not create custom udev rules" [Undecided,New] |
|
[13:15] <jdstrand> clobrano: I'm going to direct you to mvo_ for how to contribute processes, but essentially, check out a branch, do your work and then push it to launchpad and request an mp. perhaps a patch on the list would be accepted |
|
[13:15] <jdstrand> lp:snappy |
|
[13:16] <jdstrand> https://launchpad.net/snappy |
|
[13:16] <jdstrand> https://code.launchpad.net/~snappy-dev/snappy/snappy |
|
[13:16] <clobrano> perfect, thank you |
|
[13:16] <jdstrand> I'll comment in the bug |
|
[13:18] <clobrano> jdstrand: great |
|
[13:18] <dholbach> sergiusens, shall I file a separate bug for separating tests from build? (https://launchpadlibrarian.net/218099126/buildlog_ubuntu-wily-amd64.snapcraft_0.2-0~168~ubuntu15.10.1_BUILDING.txt.gz) |
|
[13:19] <sergiusens> dholbach, sure |
|
[13:19] <dholbach> will do |
|
[13:21] <jdstrand> clobrano: to get the full picture, you might also want to look at https://code.launchpad.net/~snappy-dev/ubuntu-core-launcher/trunk, but I don't think you'll need to change anything there (since it only looks at the tags to add things to the device cgroup-- so if snappy adds the right udev rule, it should all just work) |
|
[13:21] <dholbach> https://bugs.launchpad.net/snapcraft/+bug/1496392 |
|
[13:21] <ubottu> Launchpad bug 1496392 in Snapcraft "Separate tests from build" [Undecided,New] |
|
[13:22] <clobrano> jdstrand: ok, I'll have a look at it |
|
[13:46] <lool> mvo: just wanted to double-check if I got that right: the way to set network config for eth0 is to set ubuntu-core/network/interfaces -name: eth0 -content: /some/path/name to an ifupdown config? |
|
[13:49] <sergiusens> lool, just run snappy config ubuntu-core on the latest image and it will be clear ;-) |
|
[13:50] <lool> sergiusens: haha right |
|
[13:56] <lool> sergiusens: how would be transition to a non-ifupdown spec? |
|
[13:56] <sergiusens> lool, oh, you are the architect ;-) |
|
[13:56] <lool> sergiusens: tsss |
|
[13:56] <sergiusens> lool, I'm just the snapcraft software guy now ;-) |
|
[13:57] <lool> sergiusens: I'm mostly concerned about upgrades; we could keep compat with ifupdown for a while though |
|
[13:58] <sergiusens> Chipaca, https://code.launchpad.net/~sergiusens/snapcraft/1496381/+merge/271299 |
|
[13:59] <sergiusens> or dholbach ^ |
|
[14:07] <ricmm> lool: whats the problem? most things that configure network try to parse ifupdown first in case theres something they should avoid |
|
[14:08] <ricmm> we could also consider that network-admin caps can only be given to one snap |
|
[14:08] <ricmm> if ifupdown is being used, os-snap should have that assigned |
|
[14:08] <lool> ricmm: I wonder how we'd transition to something not defined by ifupdown snippets |
|
[14:08] <ricmm> if it doesnt have it, ifupdown is a no-op |
|
[14:08] <lool> ricmm: e.g. systemd-network |
|
[14:08] <lool> (for simple networking) |
|
[14:09] <ricmm> you mean a seamless translation from ifupdown files to *-network ? |
|
[14:09] <mvo_> lool: ups, sorry, missed the earlier question. yeah, tying ourself into ifup is a concern |
|
[14:09] <ricmm> well we are not tying ourseslves, we are exposing it as an option |
|
[14:10] <ricmm> we are just tying ourselves to provide the option |
|
[14:10] <zyga> https://code.launchpad.net/~zyga/snapcraft/gitignore/+merge/271313 |
|
[14:18] <mvo_> yay, r156 ready for armhf |
|
[14:22] <elopio> good morning team |
|
[14:22] <elopio> so what needs testing? |
|
[14:22] <elopio> fgimenez: ^ |
|
[14:22] <sergiusens> elopio, all the things |
|
[14:24] <fgimenez> elopio, more testing! even the tests need testing! :) how was your holiday |
|
[14:24] <elopio> it was gooood. |
|
[14:28] <dholbach> https://code.launchpad.net/~dholbach/snapcraft/1496392/+merge/271319 |
|
[14:29] <dholbach> ^ let me know if the test bits make sense like this and if the one test is expected to fail |
|
[15:33] <dholbach> sergiusens, replied |
|
[15:34] <sergiusens> @reviewlist |
|
[15:34] <nothal> https://code.launchpad.net/~snappy-dev/ubuntu-core-launcher/lp1496070/+merge/271167 | No reviews (less than a day old) |
|
[15:34] <nothal> https://code.launchpad.net/~kyrofa/webdm/autorestart_services/+merge/268120 | No reviews (32 days old) |
|
[15:34] <nothal> https://code.launchpad.net/~chipaca/snappy/precommit/+merge/271290 | No reviews (less than a day old) |
|
[15:34] <nothal> https://code.launchpad.net/~fgimenez/snappy/resize_test2/+merge/271131 | No reviews (less than a day old) |
|
[15:34] <nothal> https://code.launchpad.net/~fgimenez/snappy/udf-revision/+merge/271081 | No reviews (less than a day old) |
|
[15:34] <Chipaca> @dance for us |
|
[15:34] <nothal> Chipaca: No such command! |
|
[15:35] <sergiusens> dholbach, oh my; will fix |
|
[16:00] <elopio> sergiusens: have you seen this with the udf from tools-proposed? |
|
[16:00] <elopio> http://paste.ubuntu.com/12427656/ |
|
[16:01] <sergiusens> elopio, no |
|
[16:03] <elopio> sergiusens: just happened 2 out of 3 times here. I'll report a bug. |
|
[16:15] <elopio> fgimenez: ah, that makes sense. The comment says that the config is changed after the reboot. |
|
[16:15] <fgimenez> elopio, yep, that's for grub |
|
[16:15] <elopio> I was trying to be extra smart, and made a mess, as always. I can remove the check. |
|
[16:16] <fgimenez> elopio, but now failover tests are trying to modify kernel files on a, have you seen that? |
|
[16:17] <elopio> let me give it another try, because I am now wondering what I tested. |
|
[16:17] <elopio> fgimenez: no. |
|
[16:17] <fgimenez> elopio, ok give it a try |
|
[16:18] <elopio> fgimenez: so the change of Next is all wrong? |
|
[16:19] <fgimenez> elopio, i don't think so, it works fine on bbb, but maybe because of the problem that mvo_ mentioned with the delta upgrades we are seeing this issues now |
|
[16:21] <ogra_> sergiusens, elopio, davmor2 reported the same issue today (the kpartx one) |
|
[16:21] <ogra_> and i dont think he was using proposed ... that must be something older |
|
[16:22] <elopio> davmor2: link? |
|
[16:22] <davmor2> elopio: http://paste.ubuntu.com/12425581/ |
|
[16:23] <davmor2> seems to be working this time for what it is worth |
|
[16:23] <elopio> ogra_: the message is not the same. |
|
[16:23] <ogra_> yeah, i just noticed |
|
[16:23] <elopio> davmor2: can you please file a bug for that one? |
|
[16:23] <davmor2> elopio: I did an upgrade and rebooted at lunch time now it seems to be working |
|
[16:24] <elopio> davmor2: ok, if you happen to see it again, please let us know. |
|
[16:24] <ogra_> and send our greetings too |
|
[16:24] <ogra_> :P |
|
[16:26] <elopio> the typical joke from this side of the world would be: and what do I tell him if I don't see him? |
|
[16:27] <ogra_> :) |
|
[16:32] <slangasek> ricmm: did the fwupdate seeding for 15.10 happen today? |
|
[16:32] <ogra_> slangasek, cyphermox pinged me about it last night and then noticed it wasnt in the archive yet |
|
[16:32] <ogra_> is it now ? |
|
[16:33] <slangasek> ogra_: that's not correct, fwupdate has been in wily for weeks |
|
[16:33] <ricmm> ogra_: fwupdate should be in wily yea |
|
[16:33] <ricmm> ogra_: could you do that rq if it doesnt interfere with the other stuff? |
|
[16:34] <ogra_> oh |
|
[16:34] <ogra_> <cyphermox> disregard fwupd for now since it's not even in Debian. |
|
[16:34] * ogra_ notes thats a different package |
|
[16:34] <ogra_> i didnt catch that last night, sorry |
|
[16:34] <ogra_> do we actually have seeds ? |
|
[16:34] <ogra_> (up to now all changes for adding new packages were in livecd-rootfs) |
|
[16:37] <ogra_> i see https://code.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu-core.wily but are they actually used (we dont have a meta) |
|
[16:39] <ogra_> slangasek, added to the seed, not sure if i have to do any subsequent actions though since we dont have a meta |
|
[16:40] <slangasek> ogra_: the core seed is in the ubuntu.wily seed pod... |
|
[16:41] <slangasek> ogra_: the system-image seed. You've been hard-coding packages in livecd-rootfs?! ugh |
|
[16:41] <ogra_> slangasek, for 15.04, yes |
|
[16:41] <ogra_> i wouldnt know how i can retroactivly add anything to it |
|
[16:43] <slangasek> ogra_: ah because tasks |
|
[16:43] <ogra_> yeah |
|
[16:43] <ogra_> hmm, so the seed isnt actually the one i linked above ? |
|
[16:44] <slangasek> ogra_: no, it's not |
|
[16:45] <ogra_> eek ... |
|
[16:45] * ogra_ reverts the last commit then |
|
[16:45] <slangasek> ogra_: and yes, you're right, for 15.04 putting it in the seed doesn't help because task fields don't get regenerated |
|
[16:45] <ogra_> right |
|
[16:52] <elopio> fgimenez: I'm on kvm, and I see snappy_mode=try on /boot/grub/grubenv |
|
[16:54] <ricmm> mvo_: ^ |
|
[16:54] <ricmm> elopio: you probably hit the issue mvo just fixed |
|
[16:54] <ricmm> where the update process didnt finish correctly so flags werent set right |
|
[16:55] <elopio> ricmm: yes, that was fgimenez' suspicion. I was just giving it a try. |
|
[16:57] <Chipaca> elopio: bug 1496486 is also fixed by my branch |
|
[16:57] <ubottu> bug 1496486 in Snappy "trunk fails go vet: i18n/xgettext-go/main.go:93: unreachable code" [Undecided,In progress] https://launchpad.net/bugs/1496486 |
|
[16:57] <elopio> Chipaca: oh, sorry. |
|
[16:57] <elopio> Chipaca: which branch? |
|
[16:58] <elopio> you have like 3000, give or take. |
|
[17:00] <Chipaca> @reviewlist |
|
[17:00] <nothal> https://code.launchpad.net/~snappy-dev/ubuntu-core-launcher/lp1496070/+merge/271167 | No reviews (less than a day old) |
|
[17:00] <nothal> https://code.launchpad.net/~kyrofa/webdm/autorestart_services/+merge/268120 | No reviews (32 days old) |
|
[17:00] <nothal> https://code.launchpad.net/~elopio/snappy/fix1496486-go_vet_unreachable/+merge/271343 | No reviews (less than a day old) |
|
[17:00] <nothal> https://code.launchpad.net/~chipaca/snappy/precommit/+merge/271290 | No reviews (less than a day old) |
|
[17:00] <nothal> https://code.launchpad.net/~fgimenez/snappy/resize_test2/+merge/271131 | No reviews (less than a day old) |
|
[17:00] <Chipaca> elopio: ^ |
|
[17:00] <elopio> got it. |
|
[17:02] <Chipaca> ok, time for me to go make dinner |
|
[17:03] <mvo_> elopio: what scenario was that that you ended up with try mode? |
|
[17:04] <mvo_> elopio: and is that image/machine still available? the sudo systemctl status ubuntu-snappy.boot-ok would be good |
|
[17:04] <elopio> mvo_: I called snappy update. So I expected the try mode, and it worked. |
|
[17:06] <mvo_> elopio: aha and it worked, thats nice, I like that |
|
[17:06] <elopio> full of success :D |
|
[17:08] <mvo_> heh :) a bit worn down from all the testing/fixing actually |
|
[17:08] <fgimenez> elopio, mm was a it a delta upgrade? i've seen the issue from 168 to 169, for instance |
|
[17:10] <mvo_> I was able to reproduce from 168->170 on kvm/amd64 |
|
[17:10] <mvo_> (and I verified the fix there) |
|
[17:11] <elopio> fgimenez: 167 to 175. |
|
[17:12] <elopio> using snappy-from-branch from trunk. |
|
[17:16] <fgimenez> elopio, the last line of the errored update output should be "can not find file /writable/cache/assets/vmlinuz", are you trying on 15.04? |
|
[17:17] <elopio> fgimenez: with that I was just checking that my test for try mode was valid. |
|
[17:17] <elopio> I have a 15.04 run in progress, but with the tests from the 15.04 branch. |
|
[17:19] <fgimenez> elopio, ok that makes sense, then when the fix land we should expect that the mode is try, right? |
|
[17:19] <elopio> fgimenez: right. And if it fails, that's good, we are catching an error. |
|
[17:21] <fgimenez> elopio, ok remember to remove http://bazaar.launchpad.net/~fgimenez/snappy/1504_validation/view/head:/_integration-tests/testutils/partition/bootloader.go#L80 |
|
[17:21] <fgimenez> wait, i'll remove it |
|
[17:21] <elopio> mvo_: you also fixed this: https://bugs.launchpad.net/snappy/+bug/1496486 ^_^ |
|
[17:21] <ubottu> Launchpad bug 1496486 in Snappy "trunk fails go vet: i18n/xgettext-go/main.go:93: unreachable code" [Undecided,In progress] |
|
[17:22] <elopio> crazy times. |
|
[17:22] <mvo_> elopio: I did? excellent, I don't even remember |
|
[17:22] <elopio> yeah, we have three branches for it. |
|
[17:22] <mvo_> oh, I think I fixed and never proposed :/ ? |
|
[17:22] <mvo_> well, a good reason to land one now |
|
[17:24] <elopio> mvo_: I'm trying Chipaca's fix now. Not a big deal, you have a lot more big things going on. |
|
[17:25] * elopio just takes notes to support the slow release philosophy. |
|
[17:26] <mvo_> elopio: +10^100 from me for a slower release next time |
|
[17:26] <sergiusens> ogra_, elopio davmor2 i this is the case, it is most likely some open fds left over by snappy (rebuilt to use the latest) |
|
[17:27] <sergiusens> elopio, mvo_ fwiw, Chipaca has a branch that fixes that |
|
[17:28] <elopio> sergiusens: yup. We just found out. |
|
[17:30] <mvo_> we found 3 branches in total |
|
[17:30] <mvo_> :) |
|
[17:31] <elopio> that's how we roll. We will not just hit this release date, we will give you three fixes for everything! |
|
[17:31] <slangasek> ogra_: I see that fwupdate needs an MIR now ;) |
|
[17:31] <ogra_> how about faster release with less changes instead ? |
|
[17:32] <ogra_> slangasek, well, half of the stuff we ship needs a MIR :P ... once we are through that release chaos i'll take a look |
|
[17:32] <ogra_> (before EOW i hope) |
|
[17:35] <elopio> https://bugs.launchpad.net/snappy/+bug/1496515 |
|
[17:35] <ubottu> Launchpad bug 1496515 in Snappy "15.04 fails go vet: helpers/helpers.go:399: result of fmt.Errorf call not used" [Undecided,New] |
|
[17:35] <elopio> I think nobody has fixed this one yet ^ |
|
[17:35] * elopio goes for it. |
|
[21:09] <tedg> I really don't understand this nova build. Sometimes it seems fast. Now it's at 90m and 5.8GB of RAM. |
|
[21:10] <sergiusens> tedg, heh |
|
[21:18] <sergiusens> tedg, can you add back return self.handle_source_options() to python2-project? |
|
[21:18] <sergiusens> tedg, how's the shebang issue treating you? |
|
[21:19] <tedg> sergiusens: Not well, trying to seperate out the build and install steps but it's not clear how the various parameters interact. |
|
[21:19] <tedg> sergiusens: Now trying with the default "build" directory to see if that will work. |
|
[21:19] <sergiusens> tedg, there's a build_scripts target as well fwiw |
|
[21:22] <tedg> sergiusens: Hmm, yeah, I haven't given up on this yet :-) |
|
[21:22] * tedg has high hopes, he has apple pie, in the sky! |
|
[21:26] <atXyc0> hello, I am in the process of dd the RPi2 image |
|
[21:26] <atXyc0> it seems rather large |
|
[21:27] <atXyc0> Ctrl-t inidcates im not even 1/5 of the way done |
|
[21:28] <atXyc0> is that an ext4 partition for 4gb sdcard? |
|
[21:32] <ogra_> no, it is an SD card image with multiple partitions |
|
[21:32] <asac> atXyc0: should be 4 partitions with 3 being ext4 and one vfat for boot |
|
[21:33] <ogra_> asac, btw for the next release we should shrink writable to 10MB or so ... now that it can auto-grow |
|
[21:33] <ogra_> saves image size |
|
[21:33] <atXyc0> asac: that makes sense. do i need to expand the root partition to fill my 8gb sdcard? |
|
[21:34] <ogra_> if you wait 2 days you can get the new image that will auto resize on first boot |
|
[21:34] <ogra_> hmm, actually ... |
|
[21:34] <ogra_> just dd it and do the upgrade once the new image is out |
|
[21:34] <ogra_> it should then grow to full size automatically |
|
[21:35] <atXyc0> the dd is taking quite a long time |
|
[21:36] <ogra_> yeah, 4GB are a lot |
|
[21:36] <ogra_> for me it takes about 10-15min usually |
|
[21:37] <atXyc0> 30 minutes so far on a MBP |
|
[21:38] <atXyc0> 939524096 bytes |
|
[21:38] <asac> thats quite a long time |
|
[21:38] <ogra_> that doesnt look right |
|
[21:39] <asac> slow SD card or slow bus/writer |
|
[21:39] <asac> otherwise something fishy |
|
[21:39] <ogra_> whats the dd line you used ? |
|
[21:46] <atXyc0> dd if=ubuntu-15.04-snappy-armhf-rpi2.img if=/dev/disk2 bs=64m |
|
[21:47] <atXyc0> ogra_ ^ |
|
[21:48] <atXyc0> sorry, of for the dev line |
|
[21:48] <ogra_> disk2 ? |
|
[21:48] <atXyc0> im on osx |
|
[21:48] <ogra_> you actually have a device named like that ? |
|
[21:48] <ogra_> oh |
|
[21:48] <atXyc0> dd works the same as gnu |
|
[21:48] <ogra_> yeah |
|
[21:49] <atXyc0> the bs type is different |
|
[21:49] <ogra_> well, if it doesnt finish soon i'd try a smaller blocksize ... like 4M or so |
|
[21:49] <ogra_> it should definitely not take that long |
|
[21:51] <atXyc0> not that it matters, what was the blocksize that the image was dd off of |
|
[21:52] <ogra_> hmm, no idea what blocksize ubuntu-device-flash actually uses when creating the image ... sergiusens ? |
|
[21:54] <atXyc0_> sorry d/c |
|
[21:54] <sergiusens> ogra_, based on code from ubi provided by cj watson |
|
[21:55] <ogra_> sergiusens, so you dont know :) |
|
[21:55] <ogra_> but i'd guess some typical disk blocksize |
|
[22:01] <sergiusens> ogra_, don't remember, no |
|
[22:01] <sergiusens> and otp |
|
[22:13] <atXyc0_> 502245 bytes/second seems slow |
|
[22:13] <atXyc0_> i wonder if I got a crappy sdcard with this RPi2 |
|
[22:14] <atXyc0_> i restarted with bs=4m |
|
=== atXyc0_ is now known as atXyc0 |
|
[22:41] <asac> atXyc0: most likely bad sd... i had a similar slow one once, just trashed it and new one was on steroids |
|
|