IRC logs for #baserock for Tuesday, 2015-04-28

*** rdale_ has joined #baserock02:31
*** rdale has quit IRC02:34
*** franred has quit IRC02:55
*** fay_ has quit IRC02:55
*** tiagogomes_ has quit IRC02:55
*** flatmush has quit IRC02:55
*** sambishop has quit IRC02:55
*** petefoth has quit IRC02:55
*** nowster has quit IRC02:55
*** paulw has quit IRC02:55
*** sambishop has joined #baserock02:55
*** franred has joined #baserock02:55
*** paulw has joined #baserock02:55
*** fay_ has joined #baserock02:56
*** nowster has joined #baserock02:56
*** petefoth has joined #baserock02:56
*** flatmush has joined #baserock02:57
*** tiagogomes has joined #baserock02:57
*** zoli__ has joined #baserock04:23
*** zoli__ has quit IRC04:33
*** zoli__ has joined #baserock04:58
*** rdale_ has quit IRC06:39
*** ssam2 has joined #baserock06:52
*** ChanServ sets mode: +v ssam206:52
ssam2Logins on wiki.baserock.org should be working again07:17
petefothssam2: when I try to login using my codethink google login I get '400. That’s an error.07:20
petefothOpenID auth request contains an unregistered domain: http://wiki.baserock.org/ikiwiki.cgi'07:20
petefothI can et in with my baserock.org OPenID login though :)07:21
ssam2ok, i'll report that Google Login has broken07:24
ssam2do you know if logging in with a Google account worked for you between 10th April and yesterday?07:25
petefothssam2: I'm not sure google login *ever* worked. I think I've always had to use a Baserock OpenId07:25
ssam2ah, right.07:26
ssam2I'm sure I've seen some Google account IDs in the commit logs07:26
petefothAnd I think I have beenm logged in contuously over the period you mention07:26
*** a1exhughe5 has joined #baserock07:27
petefothIndeed, and checking thos out, one of them looks like me. I'm not helping am I? :)07:27
*** Albert_ has joined #baserock07:29
petefothNow I'm waking up, it is the Baserock StoryBoard instance not w.b.o which requires the Baserock OpenId - sorry for the confuison07:30
ssam2so you can log into wiki.baserock.org with a Google Account?07:30
petefothssam2: back to wher we came in - no. I get the error reported above. But I had been logged in with that account for the past several weeks, until I logged out juts now to see if I could login, whcih I can't.07:32
ssam2ah, ok :)07:33
ssam2I/ve emailed Branchable about that07:33
petefothssam2: FWIW I *can* login to yakking,branchabale.com with the Codethink Google ID07:35
ssam2Right. I think it's something that they changed to fix an issue triggered by installing the SSL certifuicate07:36
ssam2certificate07:36
*** rdale has joined #baserock07:48
*** bashrc has joined #baserock08:05
KinnisonGoogle login is definitely broken08:06
KinnisonBecause Google have withdrawn OpenID support08:06
petefothKinnison: but I can login to yakking using Google ID08:06
Kinnisonpetefoth: Odd, they were withdrawing it08:06
Kinnisonhttps://developers.google.com/identity/protocols/OpenID2Migration08:07
KinnisonPerhaps they've updated ikiwiki to use oath authentication08:07
petefothI had to join Twitter to use http://www.swapmyvote.uk/. Which is disappointing08:07
KinnisonI dread to think how many idiots will use that08:09
KinnisonGiven you can say "My preferred party is UKIP" a non-zero number is expected08:09
KinnisonHowever, this is perhaps not the forum for UK political discussions08:09
* Kinnison apologises08:09
*** edcragg has joined #baserock08:10
petefoth:)  http://voteswap.org/ would only let me swap with a Labour voter which doesn't work in my constituency08:10
* petefoth has juts realised where he is posting, and apologises too!08:10
*** jonathanmaw has joined #baserock08:19
ssam2i've asked if Branchable can remove 'Google Account' as an option, if it doesn't work any more08:20
*** gary_perkins has joined #baserock08:20
*** mauricemoss_ has joined #baserock08:43
*** CTtpollard has joined #baserock08:44
ssam2parted needs its .gitmodules file updating to point to upstream:gnulib instead of the real GNUlib URL08:58
ssam2I'm wondering if we can somehow block the Mason systems at baserock.org from connecting to any Git server except git.baserock.org, so that we catch this in CI in future08:59
ssam2OpenStack security groups don't allow subtractive rules so it's a bit hard to reason about how I'd do it that way08:59
rjekssam2: I was having a thought along these lines just the other day (should lorry or whatnot rewrite git submodules somehow and automatically lorry them too)08:59
ssam2i don't like that we have to do it manually, but I haven't though of a better way to do it09:00
ssam2if Lorry did it, we have a mirroring tool that needs to create commits, which seems a bit weird09:01
ssam2but it does seem the right place to do it, at the same time09:01
ssam2the other option would be to keep .gitmodules as-is, but have Morph rewrite it when it performed an operation on a commit. but that doesn't sound great either09:02
rjekAlso all the SHAs would be different.09:03
ssam2yes09:03
rjekIt's all a bit sticky09:03
rjekI do not like that if upstream sneaks in a git submodule and we don't notice that we then don't have all the sources in your trove for your system, and that upstream essentially make us contact arbitrary external servers.09:03
rjekSo perhaps having mason have local firewall rules that prevent a build from accessing the internet is a good idea anyway09:04
rjekAs you'll notice pretty quick that something very non-deterministic is happening.09:04
ssam2it does seem so, actually, yes09:04
ssam2Mason needs to allow *incoming* HTTP connections to serve its status page, though09:05
* rjek wonders if you can have iptables rules apply to a specific process tree; if you could you could just deny anything running under morph build access to the internet, and thus also spot this on your dev machine.09:05
ssam2it's the Morph process itself that checks out the submodules, so that wouldn't help here09:06
rjekIt wouldn't help that specific problem, no, but it would stop builds themselves contacting the internet.09:06
ssam2they struggle already, in practice, due to lack of /etc/resolv.conf09:07
rjekheh09:07
ssam2but it's not perfect09:07
franredI think the preference would be to explicitly add the submodules information to the chunk and not allow morph to deal with the submodules09:07
ssam2so morph ignores .gitmodules completely, and does whatever the chunk morph tells it. That sounds like a really good solution09:08
rjek+109:09
franredssam2, yes, with this we can also fix the problems when one chunk requires another source code to compile and it is not a submodule - so compiling from multiple sources09:10
rjekYou can have iptables rules apply to specific uid and gid (owner module).  Perhaps we could have a nointernet group which morph setgroups() its children into.09:10
rjekHmm; does POSIX allow for a user to /leave/ a group without being root?09:10
* rjek supposes a container is a possibility, too09:12
ssam2morph's staging area already is a 'container' in some sense of the word09:13
ssam2it would be good to align its behaviour and the terms we use with other Linux container implementations09:14
richard_mawrjek: are we only concerned about firewalling at build time to prevent builds attempting to fetch submodules or something else?09:19
richard_mawrjek: as we already deny network access to builds09:19
rjekBoth.09:20
rjekHow do we deny network access to builds, OOI?09:21
Kinnisonlinux-user-chroot creates a network namespace in which there is only local-loopback09:21
*** Krin has joined #baserock09:22
rjekRight.09:22
ssam2franred, rjek: hopefully a good summary of the submodules tasks: https://storyboard.baserock.org/#!/story/3809:22
* rjek nods ssam2 09:23
Kinnisonssam2: still not got the baserock cert on storyboard :(09:23
Kinnisonssam2: What does one do if one disagrees with the proposed approach?09:24
richard_mawKinnison: technically --unshare-net leaves it a lo and a sit0, but its lo device is unconfigured, and does not have the capabilities to configure it, so even local loopback networking isn't possible09:25
Kinnisonrichard_maw: Oh09:25
Kinnisonrichard_maw: Eww09:25
rjekThat might be needed to run some projects' tests?09:27
richard_mawyep, morph included09:27
* richard_maw isn't sure what happens if a socket is bound to 0.0.0.0 when there's no configured addresses09:28
* Kinnison discovers he can log into storyboard and post a comment09:35
* Kinnison does so09:35
ssam2Kinnison: don't know. discuss it on mailing list if nothing else09:45
Kinnisonssam2: I commented on storyvboard09:45
*** flatmush has quit IRC09:46
*** flatmush1 has joined #baserock09:46
*** simonh_ has joined #baserock09:54
*** mauricemoss_ has quit IRC09:54
*** flatmush1 is now known as flatmush10:02
radiofreeanyone know why "Command failed: sh -c 'scripts/config --set-val NOUVEAU_DEBUG_DEFAULT=3'" ?10:05
radiofreeerm.. would fail?10:05
Kinnison        --set-val option value10:06
Kinnison                             Set option to value10:06
Kinnisons/=/ /10:06
Kinnisonin your line10:06
radiofreei think it's because of the "="10:06
radiofreeyeah ".config:4223:warning: symbol value '3=' invalid for NOUVEAU_DEBUG_DEFAULT"10:07
radiofreeKinnison: yep, thanks10:07
radiofreejjardon: are you going to be submitting your "move mako to python-common" patch? i need it for mesa10:10
ratmice__hmm, not sure if/how capsicum can isolate networking (doubly so for the linux port), but in theory it'd be possible to whitelist specific ports on specific hosts to be connected via some capsicum enabled tcpip stack, this is kind of how the networking works in capros more or less10:11
ratmice__that said trying to do isolation for the rest of processes morph runs would be involved :/10:13
ratmice__and I don't think anyone would be eager to rely on an unmerged kernel fork :)10:14
jjardonradiofree: yep, I was about to finish the weston system build to test it10:21
radiofreemason still red :\10:33
* pedroalvarez goes to check mason10:33
pedroalvarezradiofree: currently building qemu10:34
radiofree14 hours into a build10:35
pedroalvareznote that the performance of our server has been affected by some IO problems of the hosting provider10:40
rjekradiofree: You should send your builds into the future to be compiled by Skynet.10:40
pedroalvarezall should be sorted now10:40
radiofreewhat system in the CI has qemu?10:43
jjardonradiofree: https://gerrit.baserock.org/47310:44
radiofree+110:44
jjardonthis reminds me a new version of qemu has been just released :)10:45
jjardonradiofree: openstack10:45
radiofreeah yeah, course10:46
radiofreehttp://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/definitions.git/tree/strata/virtualization/qemu.morph?id=1e38ff99eb9a005e0867c34b63232f6cb2a566cd10:47
radiofreei can't remember the last time i tried to compile qemu, but doesn't it default to building everything?10:47
radiofreeit does..10:51
radiofreei can't imagine things like qemu-system-m68k are going to be too useful10:51
rjekDebian don't support M68K anymore; where else are Amiga users going to go for their Linux fix if not us?10:54
*** radiofree has quit IRC10:55
*** radiofree has joined #baserock10:56
*** radiofree has quit IRC10:56
*** radiofree has joined #baserock10:56
* Kinnison hides10:57
radiofreemaybe i should add UAE to baserock, and then we can standardise our desktop environment to be workbench10:57
rjekhaha10:58
* rjek suggests ArcEm10:58
jjardonZara: hi, what was the compiler error you had?11:02
mwilliams_ctjjardon: in some of the Makefiles, it runs $(CC) -D_BSD_SOURCE -Wall -Werror -std=c99 -o $@ $^ $(LDFLAGS)11:04
mwilliams_ctonly BSD_SOURCE is deprecated, so we changed that to DEFAULT_SOURCE11:04
Zarajjardon: http://paste.baserock.org/ixucakuyut11:05
jjardonah, ok then11:07
* radiofree remembers that hack he was going to do to not build llvm on arm11:09
radiofreeare we using it for something other than mesa these days?11:09
radiofreeapparently not11:10
jjardonradiofree: AFAIK only mesa11:10
jjardonradiofree: BTW, Did you have time to try gcc 5.1 on ARM in the end?11:11
pedroalvarezradiofree: good point, maybe we can disable some things in qemu11:12
radiofreepedroalvarez: you can set a target-list, set that to only the things you'll need11:13
jjardonmmm, seems the ci is trying to deploy the system before its build:11:13
jjardonERROR: Deployment failed as system is not yet built.11:14
jjardonPlease ensure the system is built before deployment.11:14
pedroalvarezjjardon: we have seen that in the past. We haven't investigated yet11:15
jjardonpedroalvarez: any idea how did we fix the problem in the past?11:21
pedroalvarezjjardon: it just fails once, and then the next run it just works11:21
pedroalvarezso, we never fixed anything.11:22
pedroalvarezmaybe this problem is likely to happen more often now that we have more sytems in the ci11:23
jjardonah, ok; cross fingers then :)11:26
ssam2what we need to do is get the V2 Mason working11:28
ssam2I want it to provide the same interface as the V1 Mason though as well as the Gerrit integration11:28
Zaraweird issue: if I run morph branch [whatever], then try to build a system, it builds. if I run morph branch [whatever], git checkout [branch], and try to build a system, the build fails. why would this happen?11:29
ssam2because as long as people can commit directly to 'master', we still need to test 'master' as well11:29
pedroalvarezssam2: I agree, I've been thinking about forking the current mason before we change it to V211:30
ssam2I don't think it'd be too hard to write a timer unit that triggers a Zuul build and puts the result into a static page11:31
ssam2that way we could merge V2 Mason without losing V1 Mason11:31
pedroalvarezssam2: yes, that would be simpler11:32
ssam2Zara: hard to answer without knowing what error you get11:37
straycatdamn it gerrit needs to order things sensibly11:38
straycator give me a way to do that11:38
straycat</rant>11:38
* straycat sets his default screen to be the oldscreen11:41
ratmice__so, yeah it doesn't seem like capsicum does have any solution to limiting bind/connect other than disallowing it, so you'd (likely) need to modify those programs you want to have networking to accept a connected fd11:42
ratmice__https://lists.cam.ac.uk/pipermail/cl-capsicum-discuss/2013-July/msg00001.html11:42
jonathanmawhrm, undoing my changes to 'tar' still shows the same problem I had before - http://paste.baserock.org/amuxabicog11:42
Zarassam2: logs: http://paste.baserock.org/onahiratet11:42
ratmice__in theory that might not be too difficult given the limited number of programs you actually want to have networking from within the build :)11:43
radiofreejjardon: you have +2 on your mako patch, can you merge it?11:45
jjardonradiofree: done11:46
Zara(further investigation reveals that this component fails to build if built straight from morph checkout, but not if built straight from morph branch)11:48
ssam2zara: definitely indicates some sort of bug in Morph11:49
ssam2it shouldn't ever exit with a backtrace under normal circumstances11:49
ssam2'morph --verbose' seems to make Morph give /less/ output than 'morph', now11:52
ssam2maybe it's unrelated11:53
radiofreejjardon: ta11:54
Zaragit log says my morph was last updated mar 25, so maybe it's a bit out of date or buggy11:54
jjardonif someone have some time, it would really helpful if you can look at (already have a +1):  https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:rework_cross_systems11:55
jjardonmwilliams_ct: if you have time, you can review that ^ first , so building a cross tarball for armv5 will be much faster11:57
mwilliams_ctWill look at it. not sure I've got the expertise to give a +1, but shall see!11:58
jjardonmwilliams_ct: ok, thanks!11:59
ssam2https://download.baserock.org doesn't work, it asks for a username and password12:04
ssam2http:// does work, but it's a bit sucky12:04
ssam2(that https:// does not)12:04
radiofreethe certificate is also invalid12:06
Zara(I'm going to work around whatever's going on by avoiding morph checkout (and git checkout) for now, and sticking to morph branch.)12:11
*** paulw has quit IRC12:20
*** paulw has joined #baserock12:20
*** fay_ has quit IRC12:30
*** SotK_ is now known as SotK12:39
*** fay_ has joined #baserock12:46
*** mariaderidder has joined #baserock12:47
rdalein glibc.morph there are two entries for '- artifact: glibc-libs' at the top - i'm not sure of the significance of that and how it differs from putting everything in a single entry12:53
* pedroalvarez has sent another version of https://gerrit.baserock.org/#/q/topic:baserock/pedroalvarez/pxeboot-fixes12:58
pedroalvarezrdale: looks like we lost a comment: http://git.baserock.org/cgi-bin/cgit.cgi/delta/eglibc2.git/commit/eglibc.morph?h=baserock/2.15-build-essential&id=df0258044f321990eadd647e03095a48ad04c1a813:02
*** juergbi has quit IRC13:02
rdalepedroalvarez: thanks, i would rather have the comment back, as i would have never worked it out otherwise13:04
pedroalvarezrdale: I agree. It shouldn't trigger any rebuild.13:05
rdalei'm doing a change to add '/etc/ld.so.conf' to the glibc-libs metadata, and so i could put the comment back when i do that patch13:06
*** juergbi has joined #baserock13:07
*** lachlanmackenzie has joined #baserock13:18
radiofreei managed to create baserock/jetson/4.0 on one repo (nouveau), however when i try to do it http://git.baserock.org/cgi-bin/cgit.cgi/delta/linux-firmware.git/ i get an error13:22
radiofree ! [remote rejected] baserock/jetson/4.0 -> baserock/jetson/4.0 (failed to update ref)13:22
radiofreeerror: unable to resolve reference refs/heads/baserock/jetson/4.0: Not a directory13:22
richard_mawradiofree: is there a baserock/jetson branch in there currently?13:23
radiofreerichard_maw: yep13:23
richard_mawradiofree: then you can't have baserock/jetson/4.0 because of a git implementation detail that can't be fixed any more13:23
radiofreeheh, ok, i'll make it jetson-4.0 then13:23
pedroalvarezradiofree: jjardon was trying to set an standard for branch naming13:31
pedroalvarezsomething like: baserock/`git describe`-<some topic>13:31
radiofreepass...13:32
radiofreei could namespace everything under baserock/tegra actually13:33
paulsherwoodfranred: did you break mason, by any chance?13:35
pedroalvarezpaulsherwood: mason is still building13:35
paulsherwoodoh?13:35
pedroalvarezand the patch to put the genivi systems in the CI hasnt been merged yet!13:37
radiofreewhat's it up to now?13:37
pedroalvarezwe need to:13:37
pedroalvareza) Make our CI faster13:37
pedroalvarez  or13:37
pedroalvarezb) Make our CI give some feedback so the people can see that is still building13:37
pedroalvarezradiofree: pep*13:38
pedroalvarezpep813:38
pedroalvarezI think I need some help with context managers13:44
radiofreepedroalvarez: so i guess it's nearly done?13:45
pedroalvarezradiofree: yes13:45
pedroalvarezso, is this the right way of doing stacked context managers? http://paste.baserock.org/tiwocaciko.diff13:57
tiagogomesis the cache server working?13:58
radiofreeit was yesterday13:58
tiagogomesyesterday is too distant13:59
pedroalvarezshould be13:59
pedroalvarezMason is using it to build13:59
pedroalvarezand is building13:59
*** a1exhughe5 has quit IRC14:02
Kinnisonpedroalvarez: essentially yes, that is the way to stack 'em14:05
*** mariaderidder has quit IRC14:08
*** a1exhughe5 has joined #baserock14:17
richard_mawpedroalvarez: yep, but get rid of the space in `os.path.join (tftproot`14:18
jjardonpedroalvarez: is there a cache server for armv7lhf hardware?14:21
pedroalvarezjjardon: no atm14:22
jjardonpedroalvarez: ok, thanks14:22
pdarI was just booting a baserock instance on my baserock openstack and got the following error when booting for the first time. http://paste.baserock.org/jofasojigi14:26
pdarwhen I rebooted the instane I'd just deployed it worked fine.14:27
*** Krin has quit IRC14:33
*** Krin has joined #baserock14:34
pedroalvarezpdar: odd14:42
pedroalvarez:.14:42
pedroalvarez:/14:42
pedroalvarezok, I think that this patch is ready now: https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:baserock/pedroalvarez/pxeboot-fixes14:42
*** a1exhughe5 has quit IRC14:43
jjardonHi, this patch has already a +2, but I want to ask before apply it: does anyone has any big reason against it? https://gerrit.baserock.org/#/c/268/14:55
perrylhi, https://gerrit.baserock.org/#/c/408/ has had +2 and no changes apart from rebasing off master, is it fine to be merged?15:03
perryli'm not sure i like that rebasing atop master sends a completely new patch set to the series rather than checking for content changes (as opposed to just SHA1 changes) and responding appropriately...15:04
ssam2the SHA1 includes the history, so it has to change if you rebase15:05
pedroalvarezIt has a comment from straycat saying that he was going to do the merge15:05
ssam2it's annoying though, I agree15:05
perryli suppose it's a small price to pay15:06
pedroalvarezIf you do a diff using version 6 and 7 of the patch, you see that there are no changes: https://gerrit.baserock.org/#/c/408/6..715:06
pedroalvarezOnly the commit message: https://gerrit.baserock.org/#/c/408/6..7//COMMIT_MSG15:07
perrylfor content i think it's the same between patch 5 and 7, except the parent commit and commitdate in the commit msg15:08
jonathanmawrichard_maw: can I have a response to my linux-pam question - installing pam libs to /lib is in line with what systemd currently does - should I alter systemd or stick with normal systemd behaviour?15:11
*** mariaderidder has joined #baserock15:12
pedroalvarezperryl: I think it can be merged15:14
perryl\o/15:14
pedroalvarezI also think my patch-set can be merged as well :P15:14
ssam2jjardon: seems nobody has a big reason :)15:15
straycatwoah15:15
straycatcan you please not rush the merge of a branch i'm actively testing15:15
straycatthat is somewhat irritating15:15
pedroalvarezI wasn't going to, I was going to suggest that your input was needed15:16
perrylstraycat: is that referring to patch 408? given the message on april 22nd i assumed it was fine to merge and that it had been tested, apologies15:16
ssam2straycat: it may be helpful if you can set a -1 on the patch while you are testing it. it's very difficult to keep track of who is testing what15:17
straycathrm, problem is they disappear when new patchsets get pushed15:19
ssam2true....15:19
perrylit would be present in the history as a reminder however15:19
ssam2a -2 doesn't, if you want to be heavy handed15:19
perrylthat's how i found it already received +2 days earlier15:20
straycati just pushed new stuff to that topic, https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/morph+branch:master+topic:lauren/distbuild-status it's just fixups to be applied, it's been tested but let me know if anything looks wrong15:26
straycatthere's a couple of small issues, but they can be addressed later i think15:28
straycatnotably i'd like a GraphProgress message15:29
straycatunless it's so quick now that it's not worth it15:29
perrylstraycat: is BuildStarted always called in this case?15:32
perryli.e. for distbuild-start and distbuild processes?15:33
straycatyes15:34
perrylahh, is build_started only set for InitiatorStart proceses?15:35
* perryl does a full diff and sees that it is15:36
perrylstraycat: the changes you've sent look fine15:36
* straycat nods slightly and searches for catnip15:40
*** robtaylor has quit IRC15:44
*** robtaylor has joined #baserock15:44
jjardonfranred: FYI, Ive replied to your question in https://gerrit.baserock.org/#/c/187/15:45
*** zoli__ has quit IRC15:45
franredjjardon, replied back ;-)15:50
jjardonfranred: I implied you +1 https://gerrit.baserock.org/#/c/268/ then?15:52
franredjjardon, no, I vote 0, Im not opposed to be in but I not entirely sure we want it as it is15:53
jjardonfranred: yeahl, https://gerrit.baserock.org/#/c/268/ is the "proper" solution I was talking in the bug15:55
franredjjardon, yeah, that looks fine to me15:59
jjardonfranred: ok, I will push and rework the patch :)16:04
*** Albert_ has quit IRC16:05
ssam2jjardon: as soon as essential-files is merged we can add /etc/inputrc to it !! :)16:05
ssam2http://www.linuxfromscratch.org/lfs/view/6.2/chapter07/inputrc.html16:06
ssam2and the delete key will finally work16:06
jjardonssam2: exactly! And default PATH to /usr/bin;/usr/lib , and bash autocompletion by default and .... :)16:07
pedroalvarezmwilliams_ct: hey, were you looking at this? https://storyboard.baserock.org/#!/story/2 Isn't it working already?16:09
mwilliams_ctpedroalvarez: ages ago, and only for a short period of time. I noticed earlier it was still marked as doing so I marked it  todo and couldnt figure out how to de-assign16:10
pedroalvarezmwilliams_ct: just asking because I don't understand the problem16:12
pedroalvarezit just worked for me16:12
mwilliams_ctpedroalvarez: not over here16:12
mwilliams_ctI'm getting the "perhaps it is on your local network." error now16:12
pedroalvarez:/16:12
pedroalvarezodd error16:13
pedroalvarezmwilliams_ct: I was just curious, thanks for the info16:13
mwilliams_ctpedroalvarez: next time I have some upstream time I will take a look at it should it still exist16:13
mwilliams_ctsounds a fun one to chase down16:13
pedroalvarez:)16:14
* pedroalvarez sees the author "Storyboard Monkey" and starts laughing16:17
pedroalvarezjust because I remember why it exists :)16:17
*** zoli__ has joined #baserock16:22
straycatssam2, perryl, we're good to merge these distbuild changes now yes?16:22
perrylstraycat: i believe so16:23
ssam2fine by me16:25
*** Krin has quit IRC16:26
straycathrm, i'm a little bothered by the merge conflicts, maybe i'd better rebase and test then merge16:27
jonathanmawquick question: am I invisible?16:32
jmacsNo16:32
jonathanmawthank goodness16:32
jjardonmmm, trying to build a system of current definitions master and morph is building stage2-glibc :( is the cache server down?16:33
nowsteryes16:34
nowsterct-mcr-1 has been down all day16:35
straycatwhat is ct-mcr-1 ?16:35
nowsterour internal cache server16:35
straycati have no idea what you're talking about16:35
jmacsI think he means Codethink's internal cache server16:36
nowsterjmacs: yes indeed16:36
*** mariaderidder has quit IRC16:36
*** ssam2 has quit IRC16:36
pedroalvarezjjardon: the cache server is not down. It  has been building the whole day a sha from yesterday, and we have just merged some changes in build-essential16:40
nowsterdifferent cache server, then?16:41
pedroalvarezthe one true cache server: cache.baserock.org16:41
nowsterperhaps jjardon's morph.conf is pointing at the internal one16:42
radiofreepedroalvarez: are we nearly there yet?16:50
radiofreeeverything needs to be rebuilt because of http://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/definitions.git/commit/?id=1d5156078cd8475cd777f1f90127e22f6d340a2916:50
pedroalvarezradiofree: green! https://mason-x86-64.baserock.org/16:50
radiofree:D16:50
radiofreewhy didn't it do anything between 21:46:40 yesterday and 9am today?16:51
jjardonpedroalvarez: yeah, true, forgot about that16:51
pedroalvarezthis is the consequence of adding more systems to the CI16:51
pedroalvarezslower feedback, and slower cache popultaion16:51
radiofreeit's only the openstack-system-x86_64 that's been added in this round16:52
radiofreepedroalvarez: is mason building again? from HEAD?16:52
pedroalvareznote that I'm not against adding more systems to it16:52
radiofreebecause that's going to take all night16:52
pedroalvarezradiofree: yes16:53
pedroalvarezaround 60000 seconds16:53
radiofreedo you often estimate hours in seconds?16:54
jjardonit would be nice to have a result (red/gree) for every system, instead wait to build all the systems defined in ci.morph. Not sure how difficult to implement that would be though16:54
* radiofree fears pedroalvarez has been spending too much time with mason16:54
pedroalvarezradiofree: :)16:55
tlsahttps://www.google.co.uk/search?q=%2825279+%2B+43158%29s+in+hours16:55
straycatnice work tlsa :p16:56
radiofree"yeah i'll meet you in about 54 hundred seconds"16:56
jonathanmaw"I'll be there in 20 moments"16:58
bashrc"units" on the commandline16:59
*** jonathanmaw has quit IRC16:59
nowsterI'll be there in three millifortnights17:01
pedroalvarezANNOUNCEMENT: I'll be restarting gerrit in an hour to install download-commands plugin17:04
pedroalvarez18:00 UTC, right?17:04
radiofreeyeah17:05
straycat18:06 :p17:05
radiofree3240 seconds17:06
pedroalvarezhahah17:06
*** bashrc has quit IRC17:06
straycathrm, something's up with this now it's been rebased against master17:07
pedroalvarezgerrit says that the pluging has been enabled, maybe is not needed restarting gerrit17:07
*** edcragg has quit IRC17:09
pedroalvarezoh, seems that gertty is working :)17:10
pedroalvarezANNOUNCEMENT: gerrit restart operation aborted17:11
*** lachlanmackenzie has quit IRC17:15
pedroalvarezrichard_maw, straycat: IIRC you were interested on this. Gertty is working now17:21
straycatahh cool17:22
pedroalvarezlooks interesting17:24
straycati've just found an issue with distbuild, so i suggest we hold off merging the cancel/status/start commands for now17:24
pedroalvarezA comment in the patch would be enough17:25
pedroalvarezmaybe even a -217:26
pedroalvarezsome reviewers are giving +1 to patches that have some -1's and even -2's :/17:26
straycati've put a -2 on them17:29
straycatanyway, will have to figure this out tomorrow17:29
*** simonh_ has quit IRC17:29
pedroalvarezok, something is broken in master of defniitions now17:43
straycatwe don't prioritise messages in distbuild and we can send a lot of build log messages, so perhaps our command responses are getting held up by other messages17:44
* pedroalvarez wonders why mason fails but his local build succeded17:45
straycatit might be worth having a separate queue for control/status messages17:45
*** zoli__ has quit IRC18:07
pedroalvarezI've cleaned a bit mason, removed cache, etc18:09
pedroalvarezI hope that fixes the error, because it should be working18:09
*** gary_perkins has quit IRC18:45
*** zoli__ has joined #baserock19:25
*** zoli__ has quit IRC19:32
*** zoli__ has joined #baserock19:44
*** zoli__ has quit IRC19:53
*** De|ta has quit IRC21:39
*** De|ta has joined #baserock21:40
*** wdutch has quit IRC21:47
*** wdutch has joined #baserock21:54
*** zoli__ has joined #baserock22:53
*** zoli__ has quit IRC22:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!