/srv/irclogs.ubuntu.com/2017/04/21/#cloud-init.txt

=== sambetts|afk is now known as sambetts
=== shardy is now known as shardy_lunch
=== shardy_lunch is now known as shardy
=== shardy is now known as shardy_afk
smoserpowersj, so obviously my general inability to keep up with things is the cause of the problem.13:50
smoserbut https://code.launchpad.net/~syed1/cloud-init/+git/cloud-init/+merge/32202413:50
smoserlinks to a bunch of failures that dont exist13:51
smoseranything we could do about that ?13:51
powersjLooking13:51
powersj1. Easiest is to trigger a rebuild or 2. I can up the amount of runs we keep even higher. That doesn't guarantee this won't happen again though.13:52
smosereven the rebuild button doesnt work :)13:52
powersjO.o13:52
smosercause it points to the same build13:53
smoserwhich had the context of which branch was built and such13:53
powersjhmmm13:54
powersjincreasing the number of jobs we keep may not even help without increasing it greatly. With 25 we only have 10 days worth13:58
powersjI did manually kick a job off though13:58
smoserits kind of pointless though really. i wonder we could reasonably post at least the failure portion of it to the log14:02
powersjyeah14:17
=== shardy_afk is now known as shardy
smoserblackboxsw, you were looking for a bitesize bug14:33
smoseri dont have a bug (you can open one)14:33
smoser we use yaml14:33
smoserhttp://stackoverflow.com/questions/27743711/can-i-speedup-yaml14:33
smoserwe dont use it extensively, and its even possible that loading the c code could have overhead for the small bits of yaml that we load...14:33
smoserbut i've wanted to look at that14:33
smoserthat make sense ?14:33
blackboxswsmoser, sure reading that now.14:34
blackboxswyeah that makes sense14:36
blackboxswok will file a bug about it as I test out whether there is a good performance imact.14:36
blackboxsw*impact14:36
smoserbug 114:57
rharpersmoser: is there such a think as a daily-ppa but for fedora/centos cloud-init rpms ?17:48
rharperlarsks: ^^17:49
smoserhttps://build.opensuse.org/ is the closest thing that i'm aware of17:49
rharpermaybe a lxd centos/fedora container calling build-rpm ?17:51
powersjhttps://gist.github.com/smoser/19e65095b342e98fd4d6466e4d4fa1ac17:52
powersjI've been using that locally to prototype running unittests and build tests for centos17:52
powersjI think I made only a few slight changes17:52
rharperpowersj: nice!17:53
smoserrharper, but honestly.. the suse build service migth be useful.17:55
smoseri think it is basically the equivalent of ppas17:55
smoserso we could have daily builds for rpm17:55
smoserthat anyone could easily use17:55
rharperfor fedora/centos  ?17:55
rharperI assumed it was suse specific17:55
powersjopen build service does other distros17:55
rharpercool17:57
=== sambetts is now known as sambetts|afk
smoserrharper, all the touches to datasources/*.rst there...19:49
smoserthose just make targets for links ?19:49
rharpersmoser: I think so, I might need more context20:04
rharpersmoser: ah those touches were for anchors20:05
rharperyou're talking about the netconfig update20:05
rharperyou have to put a header in those files to allow link refs to work20:05
rharpersmoser: in particular, I wanted to document which network config was supported in which datasources20:06
rharperand in the network doc link to that datasource documentation page20:06
rharpersince I was in the few that support network config; I wanted to touch all of them in case someone else wanted to use the refs later on20:07
smoserrharper, yeah, i paste failed.20:10
smosermeant to include link20:10
rharperno worries, does the above make sense?20:17
=== rangerpb is now known as rangerpbzzzz
smoseryeah,  that sfine20:23
* smoser is out20:47
smoserrharper, i commented on your mp20:47
smoserit does look really good  thank you.20:47
rharperok20:47
rharpersure20:48

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!