Monday, 2015-06-08

*** sdake_ has joined #tripleo00:06
*** noslzzp_ has quit IRC00:09
*** sdake has quit IRC00:10
*** noslzzp has joined #tripleo00:10
*** sdake has joined #tripleo00:32
*** nijaba has joined #tripleo00:33
*** nijaba has quit IRC00:33
*** nijaba has joined #tripleo00:33
*** sdake_ has quit IRC00:35
*** sdake_ has joined #tripleo01:06
*** sdake__ has joined #tripleo01:09
*** sdake has quit IRC01:09
*** sdake__ is now known as sdake01:09
*** sdake_ has quit IRC01:13
*** jpeeler has joined #tripleo01:26
*** sdake_ has joined #tripleo01:43
*** sdake has quit IRC01:46
*** sdake has joined #tripleo01:54
*** sdake_ has quit IRC01:58
*** sthillma has joined #tripleo02:10
*** sthillma_ has joined #tripleo02:11
*** sthillma has quit IRC02:15
*** sthillma_ is now known as sthillma02:15
*** sdake_ has joined #tripleo02:25
*** sdake has quit IRC02:29
*** sthillma has quit IRC02:30
*** weshay has quit IRC02:30
*** weshay has joined #tripleo02:32
*** davidlenwell has quit IRC02:33
openstackgerritMartin AndrĂ© proposed openstack/diskimage-builder: Fix init system detection on Docker images  https://review.openstack.org/15278502:35
*** sseago has joined #tripleo02:40
*** sdake has joined #tripleo02:44
*** sdake_ has quit IRC02:48
*** julim has joined #tripleo02:50
*** untriaged-bot has joined #tripleo03:00
untriaged-botUntriaged bugs so far:03:00
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146161503:00
openstackLaunchpad bug 1461615 in diskimage-builder "Yum vars directory doesn't exist" [Undecided,In progress] - Assigned to Abel Lopez (al592b)03:00
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146163103:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/145946503:01
openstackLaunchpad bug 1461631 in diskimage-builder "New Fedora 22 cloud images fail to build" [Undecided,New]03:01
openstackLaunchpad bug 1459465 in tripleo "devtest fails to build ramdisk on Fedora or CentOS due to Yum multilib error" [Undecided,Triaged]03:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/146198403:01
openstackLaunchpad bug 1461984 in tripleo "VM element: /usr/sbin/grub2-install: error: will not proceed with blocklists." [Undecided,New]03:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146160103:01
openstackLaunchpad bug 1461601 in diskimage-builder "CentOS 6 not resizing" [Undecided,In progress] - Assigned to Abel Lopez (al592b)03:01
*** untriaged-bot has quit IRC03:01
*** cody-somerville- has joined #tripleo03:20
*** dasm has quit IRC03:27
*** shadower has quit IRC03:27
*** tserong has quit IRC03:27
*** mestery_afk has quit IRC03:27
*** dasm has joined #tripleo03:27
*** tserong has joined #tripleo03:28
*** tserong has joined #tripleo03:28
*** mestery has joined #tripleo03:28
*** julim has quit IRC03:31
*** noslzzp has quit IRC03:43
*** davidlenwell has joined #tripleo03:55
*** ChanServ sets mode: +v davidlenwell03:56
*** jang has joined #tripleo04:13
*** jang1 has joined #tripleo04:16
*** weshay has quit IRC04:21
*** weshay has joined #tripleo04:29
*** dasm has quit IRC04:54
*** masco has joined #tripleo05:02
*** lsmola has joined #tripleo05:36
*** matbu has joined #tripleo05:56
*** matbu has left #tripleo05:57
*** jcoufal has joined #tripleo05:57
*** tteggel has quit IRC06:04
*** dasm|afk has quit IRC06:04
*** zoli is now known as zoli|afk06:10
*** dasm has joined #tripleo06:12
dasmo/06:12
*** ukalifon has joined #tripleo06:15
*** zoli|afk is now known as zoli06:22
*** jprovazn has joined #tripleo06:43
*** dtantsur|afk is now known as dtantsur06:44
*** mordred has quit IRC06:46
*** athomas has joined #tripleo07:18
*** bvandenh_ has joined #tripleo07:20
spredzymorning all07:21
*** yog__ has joined #tripleo07:24
*** lazy_prince has joined #tripleo07:29
*** paramite has joined #tripleo07:31
*** mcornea has joined #tripleo07:38
*** Goneri has joined #tripleo07:40
*** pino|work has joined #tripleo07:46
*** gfidente has joined #tripleo07:51
gfidentemorning07:51
*** zoli is now known as zoliXXL07:51
*** jistr has joined #tripleo07:52
*** lucasagomes has joined #tripleo08:05
*** davidlenwell has quit IRC08:06
*** davidlenwell has joined #tripleo08:09
openstackgerritMerged openstack/tripleo-heat-templates: Update mongodb implementation to reflect master  https://review.openstack.org/18562808:10
*** dguerri` is now known as dguerri08:11
*** yuanying has quit IRC08:13
*** derekh has joined #tripleo08:14
*** panda has joined #tripleo08:14
openstackgerritMerged openstack/tripleo-incubator: Ensure ext-net is defined as a flat network  https://review.openstack.org/18834108:14
*** davidlenwell has quit IRC08:23
*** lxsli has joined #tripleo08:24
*** lxsli has left #tripleo08:24
*** aufi has joined #tripleo08:28
*** davidlenwell has joined #tripleo08:30
*** regebro has joined #tripleo08:31
*** yog__ has quit IRC08:33
*** shadower has joined #tripleo08:34
*** yog__ has joined #tripleo08:35
*** aufi has quit IRC08:40
*** al has quit IRC08:40
*** al has joined #tripleo08:41
*** pblaho has joined #tripleo08:45
*** lazy_prince has quit IRC08:51
*** yog__ has quit IRC08:57
*** untriaged-bot has joined #tripleo09:00
untriaged-botUntriaged bugs so far:09:00
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146161509:00
openstackLaunchpad bug 1461615 in diskimage-builder "Yum vars directory doesn't exist" [Undecided,In progress] - Assigned to Abel Lopez (al592b)09:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146163109:01
openstackLaunchpad bug 1461631 in diskimage-builder "New Fedora 22 cloud images fail to build" [Undecided,New]09:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/145946509:01
openstackLaunchpad bug 1459465 in tripleo "devtest fails to build ramdisk on Fedora or CentOS due to Yum multilib error" [Undecided,Triaged]09:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/146198409:01
openstackLaunchpad bug 1461984 in tripleo "VM element: /usr/sbin/grub2-install: error: will not proceed with blocklists." [Undecided,New]09:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146160109:01
openstackLaunchpad bug 1461601 in diskimage-builder "CentOS 6 not resizing" [Undecided,In progress] - Assigned to Abel Lopez (al592b)09:01
*** untriaged-bot has quit IRC09:01
*** dtantsur is now known as dtantsur|bbl09:03
*** akrivoka has joined #tripleo09:07
*** yog__ has joined #tripleo09:09
*** epheo has joined #tripleo09:22
*** gothicmindfood has joined #tripleo09:23
*** stendulker has joined #tripleo09:30
*** hewbrocca has quit IRC09:31
openstackgerritThomas Herve proposed openstack/os-collect-config: Add support for zaqar transport  https://review.openstack.org/18588209:43
*** kbyrne has quit IRC09:45
*** kbyrne has joined #tripleo09:48
*** pelix has joined #tripleo09:55
*** sdake has quit IRC10:00
*** hewbrocca has joined #tripleo10:00
*** hewbrocca has quit IRC10:05
*** tteggel has joined #tripleo10:08
*** openstackgerrit has quit IRC10:09
*** hewbrocca has joined #tripleo10:09
*** openstackgerrit has joined #tripleo10:09
*** tteggel has quit IRC10:12
*** tteggel has joined #tripleo10:13
*** dtantsur|bbl is now known as dtantsur10:14
jprovaznHi, anyone knows how we deal with multi-line strings in heat-templates if new lines have to be preserved? I have an issue where newlines are replaced with spaces in an ssh key: http://paste.openstack.org/show/273836/10:16
jprovaznwhich seems to be because heatclient replaces "|" with quotes10:16
jprovaznshardy: ^ maybe you might know?10:16
*** paramite is now known as paramite|afk10:17
*** bvandenh_ has quit IRC10:17
jprovaznFWIW KeystoneCACertificate which uses multiline cert is broken in templates ATM too10:18
jprovazn(I'll create a bug for this, but I wonder if there is a known solution)10:19
*** shardy_ has joined #tripleo10:19
*** shardy has quit IRC10:21
jprovaznshardy_: hi, "shardy number 2" :), I pasted a heat related question just before you joined, have you ever hit this one?10:24
jprovazn<jprovazn> Hi, anyone knows how we deal with multi-line strings in heat-templates if new lines have to be preserved? I have an issue where newlines are replaced with spaces in an ssh key: http://paste.openstack.org/show/273836/10:24
jprovazn<jprovazn> which seems to be because heatclient replaces "|" with quotes10:24
*** shardy_ has quit IRC10:25
*** shardy has joined #tripleo10:25
*** ChanServ sets mode: +v davidlenwell10:58
*** wuhg has joined #tripleo11:08
*** bvandenh_ has joined #tripleo11:10
*** thrash|wknd is now known as thrash11:10
openstackgerritThomas Herve proposed openstack/os-collect-config: Add support for zaqar transport  https://review.openstack.org/18588211:14
*** paramite|afk is now known as paramite11:17
hewbroccajistr: didn't you just do something with multi-line strings in heat templates?11:17
jistrit's not yet merged but yeah11:17
* jistr reads back11:17
*** dguerri is now known as dguerri`11:19
jistrjprovazn: all looks fine to me except the last line. I'm not sure if hiera always does this on command line or there's some problem. Do you write that out to some file? are the newlines replaced there too?11:20
*** wfoster has joined #tripleo11:20
jistrjprovazn: the /etc/puppet/hieradata/compute.yaml from the fpaste looks ok to me tbh11:20
jistrjprovazn: we already have multiline bits there and they work (but they should work better when this is merged https://review.openstack.org/#/c/188772/)11:21
jistrugh CI is down11:30
jistrFedora is broken11:30
jistrsame thing when i try it on my desktop Error: nothing provides python-ipaddress needed by python-cryptography-0.9-1.fc21.x86_6411:31
jistrhttp://logs.openstack.org/72/188772/1/check-tripleo/check-tripleo-ironic-overcloud-f21puppet-nonha/20b3258/console.html#_2015-06-08_08_24_36_45711:31
jprovazn"<jistr> jprovazn: we already have multiline" - can you send me a link to an existing working example, because for example the existing KeystoneCACertificate is broken in the same way11:35
jprovaznjistr: I *think* the problem is that "|" is replaced by quotes - if I replace qutes back with pipe in the hieradata file, it works as expected11:36
jistrjprovazn: ok, i think KeystoneCACertificate was what i had in mind actually. Then this is the fix indeed https://review.openstack.org/#/c/188772/11:36
*** rpodolyaka has joined #tripleo11:36
jprovaznjistr: great, thanks - I'll try that11:37
jistrjprovazn: thanks, please let me know if it fixed the problem for you11:37
*** stendulker has quit IRC11:40
*** lucasagomes is now known as lucas-hungry11:44
*** masco has quit IRC11:47
*** kodiakfff has left #tripleo11:49
* jistr bugging folks at #fedora about the broken package11:52
jprovaznjistr: your patch fixes the issue, thanks11:54
jistrjprovazn: awesome, thanks :)11:55
jistrjprovazn: could you give a review of that patch when you have a minute, please?11:57
jprovaznjistr: I'm doing it right now ;)11:57
jistrthx :)11:57
jistrit already passed non-ha and ha jobs, the ceph job still failing on capacity issues it seems, and when i did a recheck it turned all red because of the bad package11:58
*** u_glide has left #tripleo12:02
jistrhttps://bugzilla.redhat.com/show_bug.cgi?id=122929512:09
openstackbugzilla.redhat.com bug 1229295 in python-cryptography "python-cryptography fails to install - unsatisfied dependency python-ipaddress" [Urgent,New] - Assigned to mcepl12:09
*** jayg|g0n3 is now known as jayg12:15
*** noslzzp has joined #tripleo12:16
*** rhallisey has joined #tripleo12:20
*** rhallisey is now known as rhallisey|pto12:23
*** morazi has joined #tripleo12:26
*** morazi has quit IRC12:32
*** masco has joined #tripleo12:32
*** morazi has joined #tripleo12:32
*** trown|outttypeww is now known as trown12:34
*** lazy_prince has joined #tripleo12:34
*** shardy_ has joined #tripleo12:36
*** lucas-hungry is now known as lucasagomes12:38
*** shardy has quit IRC12:38
*** pdb has quit IRC12:41
*** pdb has joined #tripleo12:42
*** shardy_ has quit IRC12:42
*** shardy has joined #tripleo12:43
*** paramite is now known as paramite|afk12:43
openstackgerritRyan Hefner proposed openstack/tripleo-heat-templates: Enable Manila integration  https://review.openstack.org/18813712:44
*** prad has joined #tripleo12:45
*** dguerri` is now known as dguerri12:46
*** lazy_prince has quit IRC12:50
*** mordred has joined #tripleo12:54
*** akrivoka has quit IRC12:54
*** ukalifon2 has joined #tripleo12:55
*** ukalifon has quit IRC12:55
*** MasterPiece has joined #tripleo12:56
*** paramite|afk is now known as paramite13:02
*** akrivoka has joined #tripleo13:06
*** bvandenh_ has quit IRC13:07
*** dprince has joined #tripleo13:09
*** radez_g0n3 is now known as radez13:09
*** sseago has left #tripleo13:10
*** dguerri is now known as dguerri`13:12
*** panda has quit IRC13:15
*** panda has joined #tripleo13:16
*** bvandenh_ has joined #tripleo13:21
*** dguerri` is now known as dguerri13:24
*** spzala has joined #tripleo13:25
*** dguerri has left #tripleo13:27
*** akrivoka has quit IRC13:27
*** akrivoka has joined #tripleo13:28
openstackgerritRyan Hefner proposed openstack/tripleo-heat-templates: Add NetApp integration to Manila  https://review.openstack.org/18813813:32
*** akrivoka has quit IRC13:35
*** julim has joined #tripleo13:36
*** akrivoka has joined #tripleo13:41
*** sdake has joined #tripleo13:48
*** gfidente has quit IRC13:52
*** lblanchard has joined #tripleo13:57
*** sdake_ has joined #tripleo14:00
openstackgerritDan Prince proposed openstack/os-collect-config: shorter sleeps if metadata changes are detected  https://review.openstack.org/18902614:03
*** sdake has quit IRC14:04
*** BadCub-Hiding is now known as BadCub14:05
openstackgerritJan Provaznik proposed openstack/tripleo-heat-templates: Config & deployments to update overcloud packages  https://review.openstack.org/17997414:08
dprinceshardy: https://review.openstack.org/#/c/189026/ this one gets it for me14:09
*** morazi has quit IRC14:10
hewbroccadprince: morning!14:10
dprinceshardy: I have a slight preference for no config options where possible...14:10
dprincehewbrocca: morning14:10
shardydprince: Hey, totally fine by me, I can abandon my patch14:11
shardydprince: do you see measurable improvements in the end-to-end time?14:12
shardyI found it inconclusive in my VM based setup14:12
dprinceshardy: I haven't time those yet... I mean we are probably talking about a few minutes max here14:12
dprinceshardy: on something in CI which takes over an hour. So only a slight increase. But for local testing shaving off 2-3 minutes for something that takes 15 is quite noticable14:13
dprinceshardy: I'd say leave your patch posted for now. If people like having an extra config option I think perhaps we should go w/ yours14:14
dprinceshardy: sorry... I looked but apparently not carefully enough when I posted this on Saturday14:14
shardydprince: Agreed, would be good to confirm it actually makes a difference though - in my tests it looked like the quick poll occ already does generally avoided the sleep14:14
shardyhttps://github.com/openstack/os-collect-config/blob/master/os_collect_config/collect.py#L27214:16
dprinceshardy: Yeah. There is that too. The one count went to 16 though.14:16
shardye.g when we do reexec_self we poll right after the signal anyway14:16
shardyso unless heat is really loaded, the new metadata should be there14:16
shardyobviously that's a bit racy though14:16
shardyIt'd be good to somehow profile the time spent sleeping vs processing config while the create is in progress14:17
shardyMaybe we could use osprofiler14:18
dprinceshardy: yeah. A systemd-analyze here would be super useful14:19
dprinceshardy: osprofiler might work14:19
*** panda is now known as panda|afk14:19
*** morazi has joined #tripleo14:19
dprinceshardy: if it doesn't make a huge difference in local or CI wall times though I may not be as in favor of the idea14:20
*** noslzzp_ has joined #tripleo14:20
*** noslzzp_ has quit IRC14:21
shardydprince: Yeah I did some rechecks on https://review.openstack.org/#/c/185674/ intending to flip the default and compare, but the variation was too much14:21
shardyI re-tested locally a few times and wasn't sure we were getting much, if any benefit14:22
dprinceshardy: overcloud only?14:22
*** noslzzp has quit IRC14:22
dprinceshardy: or all of devtest?14:22
shardydprince: Yeah, just overcloud14:22
*** bvandenh_ has quit IRC14:22
shardyAfter that I kinda went off the idea a bit due to the extra polling load on heat14:23
shardyIf we can quantify any improvement then I'm fine with it though14:24
shardythat was also why I went for an iburst approach vs rising rate multiplier14:24
shardyIf there isn't any new data say 1 or 2 seconds after you send a signal, chances are there isn't any14:24
dprinceshardy: good to know. I give more weight to local testing numbers in this case due to the variability of the general CI wall time14:25
shardydprince: I might patch my local image to write to a file every time occ sleeps14:25
shardythen compare that to quantify the sleep/configuring ratio14:25
dprinceshardy: does enabling debug logging not help you do the same?14:26
shardydprince: Oh, hmm, yes I guess it does :D14:26
shardygrep ftw!14:26
shardywe can compare the log timestamps with the heat resource events14:27
openstackgerritMerged openstack/tripleo-heat-templates: Config & deployments to update overcloud packages  https://review.openstack.org/17997414:28
shardydprince: hey we discovered an interesting bug last week - it's impossible to turn on Debug via a stack-update14:29
shardybecause we update the hieradata which doesn't trigger the manifiests to be reapplied14:29
shardyI'm still pondering the best way to fix it14:30
shardyreally we need all of the derived puppet-manifest applying Deployment resources to reference a hash of the data, or something14:30
dprinceshardy: hmmm. it should trigger an update I think.14:32
dprinceshardy: perhaps I've pre-optimized something like: http://git.openstack.org/cgit/openstack/tripleo-heat-templates/tree/puppet/controller-post-puppet.yaml#n2614:32
shardydprince: the heat hook won't reapply a Deployment which hasn't changed, and only the data (not the manifest) has changed?14:32
dprinceshardy: oh, right. That14:32
dprinceshardy: yeah, so following stevebakers initial design for these hooks we are probably in violation of some of the original ides.14:33
shardydprince: kinda, yeah - there's a hidden dependency14:34
dprinceshardy: I think the original intention was that all "inputs" would be associated with the software deployments and get passed in via FACTOR_ (or Hiera) variables.14:34
shardydprince: exactly14:34
dprinceshardy: now, we normalized our Hiera a bit... to share things14:34
dprinceshardy: which is great, but we've introduced some update bugs14:34
dprinceshardy: I think we should just always run puppet on an update14:36
dprinceshardy: puppet is good about knowing what services need restarted on hiera updates itself14:36
shardydprince: Yeah, you may be right, or at least allow a flag which does that14:37
dprinceshardy: yes, let me have a look the hook scripts14:37
dprinceshardy: I think all we need is a new config option to force the update on select software deployments....14:37
dprinceshardy: given this may effect more than just puppet it will probably need to be off by default though14:38
*** masco has quit IRC14:40
jistrjprovazn: something like this https://github.com/stackforge/puppet-ceilometer/blob/master/manifests/init.pp#L13914:42
jistrjprovazn: looks like nova doesn't have that atm, or at least i didn't find it when grepping puppet-nova14:43
jprovaznjistr: thanks14:44
jistrjprovazn: here's a doc for the user resource type, the 'shell' parameter is what you want14:44
jistrhttps://docs.puppetlabs.com/references/stable/type.html#user14:44
jprovaznI was searching for it just now, thanks :)14:45
*** NobodyCam has joined #tripleo14:46
dprinceshardy: so option to skip this check (called force-deploy?) http://git.openstack.org/cgit/openstack/heat-templates/tree/hot/software-config/elements/heat-config/os-refresh-config/configure.d/55-heat-config#n9514:47
dprinceshardy: notice the log warning allready there :)14:48
dprinceshardy: and then we update our software deployments in TripleO to just use that....14:48
shardydprince: Yeah, maybe we use a special input to the deployment e.g "deploy_always: True" or something14:51
shardyhttps://github.com/openstack/heat/blob/master/heat/engine/resources/openstack/heat/software_deployment.py#L10214:51
shardythat would align with the existing "special" inputs created internally14:51
dprinceshardy: I think it goes where we already have enable_hiera/enable_factor: http://git.openstack.org/cgit/openstack/tripleo-heat-templates/tree/puppet/controller-post-puppet.yaml#n4414:52
dprinceshardy: so "options" right?14:52
shardydprince: aha, yup that works14:53
openstackgerritMerged openstack/tripleo-puppet-elements: Include NetApp Cinder Dependencies in Images  https://review.openstack.org/18640714:54
shardydprince: one problem though is we lose the feedback to heat if the puppet apply fails14:55
dprinceshardy: how is that?14:55
shardyFor that we need the template level dependency to be expressed so the Deployment gets replaced on update14:55
shardydprince: Heat thinks the deployment is COMPLETE, so it ignores any signals14:56
*** lazy_prince has joined #tripleo14:57
dprinceshardy: well that is just trickery14:57
shardydprince: well, you can't traverse the graph if stuff moves to complete, then may revert to incomplete asynchronously14:58
shardydprince: may have to chat with stevebaker about this when he wakes up14:58
dprinceshardy: agree, I can give him a poke about it if you want since it'll be later for you14:58
dprinceshardy: or we could start an email thread14:59
* dprince likes IRC better for this one14:59
*** untriaged-bot has joined #tripleo15:00
untriaged-botUntriaged bugs so far:15:00
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146161515:00
openstackLaunchpad bug 1461615 in diskimage-builder "Yum vars directory doesn't exist" [Undecided,In progress] - Assigned to Abel Lopez (al592b)15:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146163115:01
openstackLaunchpad bug 1461631 in diskimage-builder "New Fedora 22 cloud images fail to build" [Undecided,New]15:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/145946515:01
dprinceshardy: So perhaps an evil idea here... but I could probably wire up an output to all the hiera configurations we do...15:01
openstackLaunchpad bug 1459465 in tripleo "devtest fails to build ramdisk on Fedora or CentOS due to Yum multilib error" [Undecided,Triaged]15:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/146198415:01
openstackLaunchpad bug 1461984 in tripleo "VM element: /usr/sbin/grub2-install: error: will not proceed with blocklists." [Undecided,New]15:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146160115:01
openstackLaunchpad bug 1461601 in diskimage-builder "CentOS 6 not resizing" [Undecided,In progress] - Assigned to Abel Lopez (al592b)15:01
*** untriaged-bot has quit IRC15:01
dprinceshardy: and then pass those into other deployments that require them... thus changing their deployments on update15:01
dprinceshardy: I mean, we already depend_on them anyways... this just makes it more explicit15:01
*** dasm is now known as dasm|afk15:02
dprinceshardy: we wouldn't actually be using that data however... it is just there to "trip" Heat deployments to run on Hiera updates15:02
shardydprince: not evil, I was thinking along similar lines15:03
*** jprovazn has quit IRC15:03
dprinceshardy: no heat changes required for that, we could just do it15:03
shardydprince: then the hidden dependency (which is always evil in heat) becomes explicit, which should solve the problem :)15:03
dprinceshardy: perhaps it is also an opportunity to gut (re-write in python) the Hiera element I initially did15:03
dprinceshardy: which I'm not happy with anyways15:03
*** dsneddon has joined #tripleo15:04
*** spzala has quit IRC15:06
*** killer_prince has joined #tripleo15:06
*** spzala has joined #tripleo15:06
*** jcoufal has quit IRC15:08
*** lazy_prince has quit IRC15:09
shardydprince: sounds good, I was wondering about a hook which could write the data directly vs oac15:09
shardyI suppose it doesn't matter which tool tho15:09
dprinceshardy: we aren't have almost entirely moved away from oac15:09
dprinceshardy: It should go away entirely15:10
shardydprince: re tripping the update, there's even a digest function which could be used to calculate a hash of the list_joine'd data15:10
dprinceshardy: yeah, a hash of the hiera datafiles could be the ticket.15:10
shardye.g to make clear we're not passing the actual data in15:11
dprinceshardy: still, we'd need an output variable to provide it to Heat15:11
dprinceshardy: I think I like this approach best15:11
shardyhttp://docs.openstack.org/developer/heat/template_guide/hot_spec.html#digest15:12
shardyThe one and only function not currently used in t-h-t! ;D15:12
shardyOnly supports a string for value atm by the looks of it15:12
dprinceshardy: I noticed we just added UpdateIdentifier (stevebaker's work) to our t-h-t this weekend15:12
hewbroccaLOL15:13
dprinceshardy: I think this is exactly the same pattern here....15:13
dprinceshardy: just for hiera15:13
shardydprince: yup, it's basically the same pattern15:14
hewbroccaIf we had set out to write something that would exercise every conceivable Heat feature, I don't think we could have done much better than t-h-t15:14
dprincehewbrocca: well we aren't using custom resources yet, but shardy won't let me so that may be a good thing15:15
shardyhewbrocca: very true, TripleO has proven to be a very effective kitchen-sink heat functional tests ;)15:15
openstackgerritAna Krivokapic proposed openstack/tuskar-ui: Only fetch kernel and ramdisk images when needed  https://review.openstack.org/18933015:15
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Add support for isolating swift storage nets  https://review.openstack.org/18865515:15
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Add support for isolating block storage nets  https://review.openstack.org/18902515:15
*** dprince has quit IRC15:17
*** bvandenh_ has joined #tripleo15:18
*** andrearosa has joined #tripleo15:19
*** lazy_prince has joined #tripleo15:19
*** killer_prince has quit IRC15:22
*** killer_prince has joined #tripleo15:23
*** killer_prince has quit IRC15:24
*** lazy_prince has quit IRC15:25
*** lazy_prince has joined #tripleo15:26
openstackgerritMerged openstack/tripleo-heat-templates: Enable NetApp Backends in Cinder  https://review.openstack.org/18770015:31
*** killer_prince has joined #tripleo15:31
*** lazy_prince has quit IRC15:35
*** yog__ has quit IRC15:37
*** sdake_ has quit IRC15:42
*** sdake has joined #tripleo15:47
*** sdake_ has joined #tripleo15:48
*** sdake has quit IRC15:52
*** dtantsur is now known as dtantsur|afk15:54
*** cwolferh has joined #tripleo15:54
*** sdake_ is now known as sdake15:56
*** daneyon has joined #tripleo15:56
*** wuhg has quit IRC15:57
*** killer_prince is now known as lazy_prince15:58
*** devananda has joined #tripleo15:59
*** mcornea has quit IRC16:04
*** jistr has quit IRC16:10
*** Goneri has quit IRC16:11
*** daneyon_ has joined #tripleo16:12
*** daneyon has quit IRC16:14
*** yamahata has joined #tripleo16:15
openstackgerritMerged openstack/tuskar-ui: Only fetch kernel and ramdisk images when needed  https://review.openstack.org/18933016:17
*** ukalifon2 has quit IRC16:17
*** alop has joined #tripleo16:23
*** alop has quit IRC16:23
*** alop has joined #tripleo16:24
*** rhallisey|pto has quit IRC16:31
*** dprince has joined #tripleo16:31
*** rhallisey|pto has joined #tripleo16:43
*** bnemec has quit IRC16:45
*** morazi has quit IRC16:45
*** bnemec has joined #tripleo16:46
*** rlandy has joined #tripleo16:54
*** killer_prince has joined #tripleo16:56
*** lazy_prince has quit IRC17:00
*** lucasagomes is now known as lucas-dinner17:01
*** derekh has quit IRC17:01
*** hewbrocca has quit IRC17:04
*** daneyon has joined #tripleo17:05
*** SlickNik has joined #tripleo17:06
*** daneyon_ has quit IRC17:07
*** akrivoka has quit IRC17:10
*** panda|afk has quit IRC17:15
*** panda|afk has joined #tripleo17:16
*** SlickNik has left #tripleo17:17
*** killer_prince has quit IRC17:17
*** rhallisey|pto has quit IRC17:19
*** Marga_ has joined #tripleo17:23
*** paramite has quit IRC17:23
dprinceall tripleo Fedora seeds (with packages) fail with this error: https://bugs.launchpad.net/tripleo/+bug/146310717:25
openstackLaunchpad bug 1463107 in tripleo "Fedora seed fails to build due to python-ipaddress package requirement" [Critical,Triaged]17:25
*** Marga_ has quit IRC17:27
*** Marga_ has joined #tripleo17:27
*** trown is now known as trown|lunch17:29
*** pradk has joined #tripleo17:30
*** rhallisey|pto has joined #tripleo17:30
*** lsmola has quit IRC17:32
*** SlickNik has joined #tripleo17:39
*** rlandy has quit IRC17:42
*** daneyon_ has joined #tripleo17:45
*** daneyon has quit IRC17:48
*** pelix has quit IRC17:51
*** sthillma has joined #tripleo17:52
*** pino|work has quit IRC17:55
*** dontalton has joined #tripleo18:00
*** sdake has quit IRC18:02
*** sdake has joined #tripleo18:02
*** hewbrocca has joined #tripleo18:03
*** openstackgerrit has quit IRC18:09
*** openstackgerrit has joined #tripleo18:09
mordredgreghaynes: so, you said something about output format testing18:09
mordredgreghaynes: as it relates to the docker output patch18:09
mordredgreghaynes: can you point me towards that?18:09
greghaynesmordred: check out tests/image_output_formats.bash18:10
greghaynesin diskimage-builder18:10
mordredgreghaynes: btw - the docker root element is even cooler18:11
mordredgreghaynes: it's hella quick compared to debootstrap18:12
greghaynesnice!18:12
mordredgreghaynes: there are ... a few things that some of the low-level elements expect to be true organizationall which the docker element calls in to question18:12
mordredgreghaynes: given that it is a root element that does not know what distro it is installing18:12
greghayneshrm, it cant figure that out?18:13
mordredwell, it can18:13
mordredbut it can't use the existing infrastructure for doing so18:13
mordredit doesn't need to do a ton that depends on it - but, for instace, the docker ubuntu images dont' have python in them, which kinda breaks other elements - so it needs to be installed18:14
mordredbut we're not bootstrapped enough to be able to use install-packages to install python, because install-packages is in python :)18:14
*** Goneri has joined #tripleo18:14
greghaynesyou mean package-installs?18:15
greghaynesbecause I thought install-packages was bash18:15
*** rhallisey|pto has quit IRC18:15
mordredyah18:15
mordredwhichever one it is18:15
mordredit's also possible that I don't knw what I'm talking about :)18:15
*** rhallisey has joined #tripleo18:16
greghaynesyea, package-installs is probably it18:16
greghayneswe should make that element assert that python is installed18:16
greghaynesah, so youre tunneling through a DIB_DOCKER_IMAGE var18:17
mordredyah18:17
mordredalthough defaulting to using DISTRO_NAME and DIB_RELEASE if I can - since other things want those variables set too18:18
greghaynesmaybe for now we say if you set that you have to specify DISTRO_NAME and DIB_RELEASE18:18
mordredgreghaynes: do we have any way for an element to add more elements to the list of elements ... dynamically?18:18
greghaynesno18:18
mordredk18:18
greghayneswhyfor?18:19
mordredI think for now disk-image-create docker dpkg and disk-image-create docker yum is fine18:19
mordredto be able to have the root element look at the docker image and figure out whether you want the dpkg or the yum elements included18:19
greghaynesah18:19
greghaynesyea, this will get interesting18:20
mordred:)18:20
*** shardy_ has joined #tripleo18:21
greghaynesI could see a case for us doing a very very early run a script per element if it exists to see if we should add any additional dependencies18:22
*** shardy has quit IRC18:22
greghaynesas long as its before we get into the build process much18:22
greghaynesmordred: theres custom metadata for docker images right?18:24
greghaynesmaybe we can put deps and the DIB_RELEASE and DISTRO_NAME in the doker image metadata18:24
greghaynesand alternatively support specifying those explicitly18:25
*** shardy_ has quit IRC18:26
*** daneyon_ has quit IRC18:26
mordredneat idea18:26
*** shardy has joined #tripleo18:27
openstackgerritMonty Taylor proposed openstack/diskimage-builder: Add docker output support  https://review.openstack.org/18915818:27
openstackgerritMonty Taylor proposed openstack/diskimage-builder: Add base element for using docker as image base  https://review.openstack.org/18916518:27
mordredgreghaynes: added tests to the first one18:27
mordredlemme know if I'm doing that right18:27
*** trown|lunch is now known as trown18:29
greghaynesmordred: I dont think you can stack traps like that18:29
mordredpoo18:29
mordreddo I need to put both in the one trap then?18:29
greghaynesyep18:29
openstackgerritMonty Taylor proposed openstack/diskimage-builder: Add docker output support  https://review.openstack.org/18915818:29
greghayneslooks good18:31
mordred\o/18:31
mordredjeblair: ^^ because I know it's exactly what you've always wanted - I made a patch to allow the creation of docker images as an output format from dib18:31
jeblairmordred: now we can stack them in neat little piles and go on strike!18:32
*** rhallisey is now known as rhallisey|pto18:32
jeblair(it's possible i may have the wrong idea what containerization is about)18:32
mordredyup!18:33
mordredheh. nope. I think you've got it right18:33
mordredjeblair: amazingly enough, it works pretty well and is super easy to use to debug/verify whether or not content got put into an image like you thought it would18:34
mordredjeblair: if I didn't think we'd DDOS dockerhub, I'd suggest we publish a docker version of all of our images for nodepool :)18:35
jeblairmordred: cool; i'm going to switch to -infra for the next part of this convo18:35
bkeromordred: yesplz18:36
*** Marga_ has quit IRC18:40
*** Goneri has quit IRC18:44
*** Marga_ has joined #tripleo18:47
*** prad__ has joined #tripleo19:14
*** BadCub is now known as BadCub-Hiding19:20
*** prad__ has quit IRC19:23
*** pradk has quit IRC19:28
*** dprince has quit IRC19:39
*** MasterPiece has quit IRC19:42
*** Marga_ has quit IRC19:57
*** dprince has joined #tripleo20:01
*** jayg is now known as jayg|g0n320:11
*** dprince has quit IRC20:22
*** radez is now known as radez_g0n320:23
*** chuckC_ has joined #tripleo20:34
*** shardy_ has joined #tripleo20:34
*** shardy has quit IRC20:36
*** shardy_ has quit IRC20:40
*** shardy has joined #tripleo20:40
*** dsneddon has quit IRC20:45
*** spzala has quit IRC20:50
*** Marga_ has joined #tripleo20:54
openstackgerritDan Sneddon proposed openstack/os-net-config: Set primary interface on OVS bonds  https://review.openstack.org/18944720:55
*** dsneddon has joined #tripleo20:56
*** pradk has joined #tripleo20:57
*** Marga_ has quit IRC20:59
*** Marga_ has joined #tripleo20:59
*** untriaged-bot has joined #tripleo21:00
untriaged-botUntriaged bugs so far:21:00
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146161521:00
openstackLaunchpad bug 1461615 in diskimage-builder "Yum vars directory doesn't exist" [Undecided,In progress] - Assigned to Abel Lopez (al592b)21:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146160121:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/145946521:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/146309221:01
openstackLaunchpad bug 1461601 in diskimage-builder "CentOS 6 not resizing" [Undecided,In progress] - Assigned to Abel Lopez (al592b)21:01
untriaged-bothttps://bugs.launchpad.net/diskimage-builder/+bug/146163121:01
openstackLaunchpad bug 1459465 in tripleo "devtest fails to build ramdisk on Fedora or CentOS due to Yum multilib error" [Undecided,Triaged]21:01
openstackLaunchpad bug 1463092 in tripleo "hieradata change doesn't reapply manifest" [Undecided,Triaged]21:01
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/146198421:01
openstackLaunchpad bug 1461631 in diskimage-builder "New Fedora 22 cloud images fail to build" [Undecided,New]21:01
*** untriaged-bot has quit IRC21:01
openstackLaunchpad bug 1461984 in tripleo "VM element: /usr/sbin/grub2-install: error: will not proceed with blocklists." [Undecided,New]21:01
*** lblanchard has quit IRC21:08
openstackgerritDan Sneddon proposed openstack/os-net-config: Set primary interface on OVS bonds  https://review.openstack.org/18944721:16
*** julim has quit IRC21:16
openstackgerritOpenStack Proposal Bot proposed openstack/os-cloud-config: Updated from global requirements  https://review.openstack.org/18848321:19
*** sdake_ has joined #tripleo21:19
openstackgerritOpenStack Proposal Bot proposed openstack/tuskar: Updated from global requirements  https://review.openstack.org/18693921:21
*** SpamapS has quit IRC21:22
*** sdake has quit IRC21:23
openstackgerritRyan Brady proposed openstack/tuskar-ui: Adds network configuration  https://review.openstack.org/18947321:23
*** trown is now known as trown|outttypeww21:24
openstackgerritDan Sneddon proposed openstack/os-net-config: Set primary interface on OVS bonds  https://review.openstack.org/18944721:35
*** Marga_ has quit IRC21:37
*** lucas-dinner has quit IRC21:37
*** EmilienM is now known as EmilienM|afk21:41
openstackgerritRyan Brady proposed openstack/tuskar-ui: Adds network configuration  https://review.openstack.org/18947321:42
*** BadCub-Hiding is now known as BadCub21:47
*** dprince has joined #tripleo21:49
openstackgerritEmilien Macchi proposed stackforge/puppet-tripleo: Implement Advanced Firewalling support  https://review.openstack.org/18948421:49
*** lucas-dinner has joined #tripleo21:50
*** lucas-dinner has quit IRC21:50
*** SpamapS has joined #tripleo21:55
*** chuckC_ has left #tripleo21:56
openstackgerritRyan Brady proposed openstack/tuskar-ui: Adds network configuration  https://review.openstack.org/18947322:00
openstackgerritDan Sneddon proposed openstack/os-net-config: Set primary interface on OVS bonds  https://review.openstack.org/18944722:00
*** pradk has quit IRC22:08
*** spzala has joined #tripleo22:21
*** penick has joined #tripleo22:31
*** bnemec changes topic to "CI Broken due to DIFFERENT bug: https://bugs.launchpad.net/tripleo/+bug/1463107 | Deploying OpenStack Using OpenStack | https://wiki.openstack.org/wiki/TripleO"22:37
*** daneyon has joined #tripleo23:10
*** daneyon_ has joined #tripleo23:11
*** daneyon has quit IRC23:15
*** dprince has quit IRC23:18
*** EmilienM|afk is now known as EmilienM23:20
*** yuanying has joined #tripleo23:25
openstackgerritDan Sneddon proposed openstack/os-net-config: Set primary interface on OVS bonds  https://review.openstack.org/18944723:34
*** chlong has joined #tripleo23:36
*** dontalton has quit IRC23:45
*** weshay has quit IRC23:47
*** alop has quit IRC23:53

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