*** sdake has joined #tripleo | 00:20 | |
*** dmellado has quit IRC | 00:21 | |
*** sdake_ has quit IRC | 00:24 | |
*** radez is now known as radez_g0n3 | 00:27 | |
*** dmellado has joined #tripleo | 00:29 | |
*** Marga_ has joined #tripleo | 00:36 | |
*** Marga_ has quit IRC | 00:41 | |
*** weshay has joined #tripleo | 01:05 | |
*** MasterPiece has quit IRC | 01:13 | |
*** Marga_ has joined #tripleo | 01:37 | |
*** Marga_ has quit IRC | 01:42 | |
*** Marga_ has joined #tripleo | 01:53 | |
*** tzumainn has joined #tripleo | 02:03 | |
*** Marga_ has quit IRC | 02:15 | |
*** dmellado has quit IRC | 02:22 | |
greghaynes | ok, I had to do ugly, ugly things | 02:34 |
---|---|---|
greghaynes | but I have coopted our break= in order to allow negative testing | 02:34 |
*** weshay has quit IRC | 02:37 | |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Initial element tests https://review.openstack.org/174681 | 02:38 |
greghaynes | SpamapS: ^ | 02:38 |
*** dmellado has joined #tripleo | 02:42 | |
*** tzumainn has quit IRC | 02:51 | |
*** tserong has quit IRC | 02:52 | |
*** vinkman has quit IRC | 02:53 | |
*** yamahata has quit IRC | 02:56 | |
*** untriaged-bot has joined #tripleo | 03:00 | |
untriaged-bot | Untriaged bugs so far: | 03:00 |
untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1445161 | 03:00 |
openstack | Launchpad bug 1445161 in diskimage-builder "setfiles doesn't work on all distros" [Undecided,New] | 03:00 |
*** untriaged-bot has quit IRC | 03:00 | |
*** tserong has joined #tripleo | 03:09 | |
*** tserong has joined #tripleo | 03:09 | |
*** eghobo has quit IRC | 03:12 | |
*** vinkman has joined #tripleo | 03:24 | |
*** pradeep has joined #tripleo | 03:28 | |
*** sdake_ has joined #tripleo | 03:34 | |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Initial element tests https://review.openstack.org/174681 | 03:38 |
*** sdake has quit IRC | 03:38 | |
*** ishant has joined #tripleo | 03:45 | |
*** spzala has quit IRC | 03:45 | |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Initial element tests https://review.openstack.org/174681 | 03:56 |
*** eghobo has joined #tripleo | 04:09 | |
*** eghobo_ has joined #tripleo | 04:12 | |
*** eghobo has quit IRC | 04:15 | |
*** dsneddon has quit IRC | 04:24 | |
openstackgerrit | Merged stackforge/kolla: Fix validate-all-yaml.sh to check for files with yml extension https://review.openstack.org/173168 | 04:30 |
*** yog__ has joined #tripleo | 04:48 | |
openstackgerrit | Merged stackforge/kolla: Disables firewalld for Heat-deployed Fedora instances https://review.openstack.org/174628 | 04:49 |
openstackgerrit | Merged stackforge/kolla: Adds Mariadb-client to the Heat template https://review.openstack.org/174629 | 04:52 |
*** hewbrocca has quit IRC | 05:17 | |
*** openstackgerrit has quit IRC | 05:21 | |
*** openstackgerrit has joined #tripleo | 05:21 | |
*** hewbrocca has joined #tripleo | 05:36 | |
*** eghobo_ has quit IRC | 05:37 | |
*** eghobo has joined #tripleo | 05:39 | |
*** pppingme has joined #tripleo | 05:55 | |
*** mmagr has joined #tripleo | 05:58 | |
*** aufi has joined #tripleo | 06:03 | |
*** masco has joined #tripleo | 06:20 | |
*** bnemec has quit IRC | 06:32 | |
*** adrianopetrich has joined #tripleo | 06:33 | |
*** pradeep has quit IRC | 06:36 | |
*** yamahata has joined #tripleo | 06:38 | |
*** jprovazn has joined #tripleo | 06:40 | |
*** eghobo has quit IRC | 06:41 | |
*** mmagr has quit IRC | 06:52 | |
*** pradeep has joined #tripleo | 06:55 | |
*** pradeep has quit IRC | 07:00 | |
*** zz_avozza is now known as avozza | 07:04 | |
*** sdake has joined #tripleo | 07:06 | |
*** mmagr has joined #tripleo | 07:07 | |
*** sdake_ has quit IRC | 07:09 | |
*** jcoufal has joined #tripleo | 07:10 | |
*** dtantsur|afk is now known as dtantsur | 07:13 | |
*** pradeep has joined #tripleo | 07:20 | |
*** ifarkas has joined #tripleo | 07:28 | |
*** mandre_ is now known as mandre | 07:30 | |
*** chlong has quit IRC | 07:33 | |
*** yamahata has quit IRC | 07:34 | |
*** gfidente has joined #tripleo | 07:43 | |
*** jistr has joined #tripleo | 07:50 | |
*** dtantsur is now known as dtantsur|brb | 07:51 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: puppet: add missing ::glance include on controller https://review.openstack.org/174382 | 07:58 |
*** dasm has quit IRC | 08:03 | |
*** dasm has joined #tripleo | 08:06 | |
*** dasm has quit IRC | 08:14 | |
*** dasm has joined #tripleo | 08:14 | |
*** lucasagomes has joined #tripleo | 08:19 | |
*** derekh has joined #tripleo | 08:22 | |
*** sdake has quit IRC | 08:28 | |
*** bvandenh has joined #tripleo | 08:33 | |
*** mcornea has joined #tripleo | 08:40 | |
*** jprovazn has quit IRC | 08:43 | |
*** jprovazn has joined #tripleo | 08:45 | |
*** GheRivero has quit IRC | 08:45 | |
*** Goneri has joined #tripleo | 08:49 | |
*** pcaruana has joined #tripleo | 08:50 | |
*** jprovazn has quit IRC | 08:54 | |
*** jprovazn has joined #tripleo | 08:57 | |
*** GheRivero has joined #tripleo | 09:00 | |
*** untriaged-bot has joined #tripleo | 09:00 | |
untriaged-bot | Untriaged bugs so far: | 09:00 |
untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1445161 | 09:00 |
openstack | Launchpad bug 1445161 in diskimage-builder "setfiles doesn't work on all distros" [Undecided,New] | 09:00 |
*** untriaged-bot has quit IRC | 09:00 | |
*** vinkman has quit IRC | 09:16 | |
*** athomas has joined #tripleo | 09:28 | |
*** athomas has quit IRC | 09:30 | |
*** athomas has joined #tripleo | 09:30 | |
*** jtomasek has quit IRC | 09:33 | |
*** jtomasek has joined #tripleo | 09:36 | |
jistr | gfidente: haha we have a new kind of problem :P pcsd automatically replicates the created resources to other nodes. If all nodes attempt to create them, there can be a race and some nodes will fail to create resource because it already exists, causing the puppet run to fail even though nothing is wrong in practice. So if we want pcsd enabled, we need to restrict resource creation to the bootstrap node. We can check with HA folks if we really | 09:36 |
jistr | need pcsd, but my bet is that we do, to follow some common HA architecture... | 09:36 |
*** avozza is now known as zz_avozza | 09:39 | |
gfidente | jistr, ack, I remember puppet-pacemaker had indeed a boolean to specify which node was the master | 09:44 |
gfidente | on the other hand, I see the VIP/haproxy resources created simultaneously on all nodes are doing fine | 09:45 |
gfidente | because we don't have pcsd yet | 09:45 |
jistr | gfidente: exactly, they're doing fine until you enable pcsd... | 09:45 |
gfidente | so looks like either we try to use it and adapt things accordingly, or live without it | 09:45 |
gfidente | I am not particularily happy of having to pick a master though, we can assume that being the bootstrap node | 09:46 |
gfidente | but when/if it goes down or is deleted things might get suboptimal | 09:46 |
jistr | gfidente: yeah that's what i just did, used the bootstrap node only to create the VIPs, and i have successful stack create | 09:46 |
gfidente | jistr, think this needs to be discussed with ha guys indeed | 09:47 |
jistr | yeah | 09:47 |
gfidente | they might point benefits/drawbacks of the two approaches | 09:47 |
*** akrivoka has quit IRC | 09:49 | |
*** akrivoka has joined #tripleo | 09:50 | |
*** adrianopetrich has quit IRC | 09:50 | |
*** vinkman has joined #tripleo | 09:52 | |
*** pelix has joined #tripleo | 09:56 | |
*** zz_avozza is now known as avozza | 10:07 | |
*** yog__ has quit IRC | 10:20 | |
*** dtantsur|brb is now known as dtantsur | 10:35 | |
*** mark-os has joined #tripleo | 10:41 | |
mark-os | Just wondering if it's possible to revert back to the original Horizon dashboard after enabling Tuskar UI and if so how would I go about doing this? | 10:42 |
*** yuanying has quit IRC | 10:43 | |
mark-os | I know the likes of _10_admin.py , _20_project.py and _30_identity.py were overwritten, is there any way to restore them back to the way they originally were | 10:44 |
akrivoka | mark-os: just delete the _50_tuskar.py file | 10:44 |
mark-os | cheers for the tip akrivoka | 10:46 |
*** adrianopetrich has joined #tripleo | 10:48 | |
*** Marga_ has joined #tripleo | 11:03 | |
*** Marga_ has quit IRC | 11:03 | |
*** Marga_ has joined #tripleo | 11:04 | |
*** rhallisey has joined #tripleo | 11:18 | |
*** ishant has quit IRC | 11:22 | |
*** pradeep has quit IRC | 11:25 | |
*** trown|outttypeww is now known as trown | 11:32 | |
*** athomas has quit IRC | 11:35 | |
*** prad has quit IRC | 11:36 | |
*** athomas has joined #tripleo | 11:38 | |
*** lblanchard has joined #tripleo | 11:41 | |
*** julim has joined #tripleo | 11:46 | |
*** dprince has joined #tripleo | 11:47 | |
*** thrash|g0ne is now known as thrash | 11:48 | |
*** prad has joined #tripleo | 11:51 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Update Delorean repo url https://review.openstack.org/174833 | 11:52 |
openstackgerrit | James Slagle proposed openstack/tripleo-puppet-elements: Use the split out -api and -registry glance packages https://review.openstack.org/174834 | 11:53 |
*** prad has quit IRC | 11:54 | |
*** adrianopetrich has quit IRC | 11:57 | |
*** adrianopetrich has joined #tripleo | 11:58 | |
*** masco has quit IRC | 12:01 | |
*** jprovazn has quit IRC | 12:07 | |
slagle | those 2 patches are co-dependent ^^. not sure how to roll that out | 12:13 |
slagle | dprince: fyi ^ | 12:13 |
*** weshay has joined #tripleo | 12:13 | |
*** lucasagomes is now known as lucas-hungry | 12:13 | |
dprince | slagle: you could cherry pick the tripleo-image-elements patch in your DELOREAN_URL tripleo-ci patch? | 12:15 |
dprince | slagle: Why is the RDO maintainer breaking us? | 12:16 |
dprince | slagle: I thought the feedback loop here would be to have the package change... not us | 12:16 |
dprince | slagle: FWIW neutron just did a bunch of package split outs (for all the AAS packages) and the maintainer was super conserned about not breaking end users in this regard. To the extent that we make cyclical package deps (it actually broke delorean bootstrapping). | 12:17 |
*** rbrady has quit IRC | 12:18 | |
dprince | slagle: I was reluctant to let it slide in Delorean because I hate those sorts of deps... but not breaking end users is important too. In the Neutron case it was the project that actually broke everyone anyway, not our packages. | 12:18 |
slagle | well, the glance packaging folks are much less concerned | 12:18 |
dprince | slagle: but for Glance... There isn't anything new there architecturally right? We just decided to split packages all of a sudden | 12:19 |
slagle | in fact, we could say not concerned at all | 12:19 |
slagle | this already all blew up on the puppet side as well | 12:19 |
dprince | slagle: right, someone (perhaps you) asked me this yesterday and my reply was like... we should just undo the package fix | 12:19 |
dprince | slagle: that is what reverts are for | 12:19 |
slagle | https://bugs.launchpad.net/puppet-glance/+bug/1444974 | 12:20 |
openstack | Launchpad bug 1444974 in puppet-glance "openstack-glance which provides glance-manage not getting installed on red hat platforms" [Critical,Won't fix] - Assigned to Emilien Macchi (emilienm) | 12:20 |
dprince | slagle: I know about the glance fix. Regardless of the packaging issue what Emilien proposed there was fine on its own. (essentially the 'include :glance' is a standard thing we do in puppet | 12:21 |
*** jayg|g0n3 is now known as jayg | 12:21 | |
*** rbrady has joined #tripleo | 12:21 | |
*** rbrady has joined #tripleo | 12:21 | |
dprince | slagle: but as for this package change I think I'm gonna say lets revert it | 12:21 |
dprince | slagle: so this doesn't fix it, https://github.com/openstack-packages/glance/commit/c8c25e78b723f0aac5cba464d1acae70a2686e79 | 12:23 |
slagle | dprince: nope, that was reverted! | 12:24 |
slagle | https://review.gerrithub.io/#/c/230356 | 12:24 |
slagle | b/c everyone wanted to wax all philisophical about why puppet should and shouldnt do certain things | 12:24 |
slagle | meanwhile, no one can install glance | 12:24 |
dprince | slagle: well, I just argued this same issue with Neutron (and I let them win) | 12:24 |
dprince | slagle: I'll re-propose the revert | 12:25 |
slagle | all i originally wanted was to see puppet-glance install openstack-glance, like it always had been | 12:26 |
dprince | slagle: right, that is what we should care about most I think | 12:26 |
slagle | but with the split out and the backwards Requires so that -api and -reg still get pulled in, that wasn't "pure" I guess | 12:26 |
slagle | backwards compatible Requires I mean | 12:26 |
dprince | slagle: https://review.gerrithub.io/230453 | 12:33 |
dprince | slagle: there, we should just revert it all | 12:33 |
slagle | dprince: i linked to the puppet-glance revert that would be required as well | 12:36 |
slagle | and an opm revert would also be needed | 12:36 |
dprince | slagle: this is such a mess | 12:37 |
slagle | yes! | 12:37 |
dprince | slagle: all we needed was a little pushback and all this would have been avoided | 12:37 |
slagle | well i disagree with that | 12:37 |
dprince | slagle: okay, then how would you avoid this? | 12:38 |
slagle | the pushback was there, I asked them to fix it | 12:39 |
slagle | so i dunno? | 12:40 |
slagle | maybe educate people about not making backwards incompatible packaging changes | 12:40 |
dprince | where is the puppet-glance fix? | 12:40 |
slagle | there is no puppet-glance fix, it got closed wontfix | 12:41 |
slagle | they decided to "fix" it on the packaging side | 12:41 |
EmilienM | slagle: isn't packaging issue solved? | 12:41 |
dprince | slagle: so if I land this Delorean fix does that resolve your issues today? | 12:41 |
*** rlandy has joined #tripleo | 12:42 | |
dprince | slagle: we can argue about the right thing to do on a list somewhere, or the original BZ for that matter | 12:42 |
slagle | dprince: i dont *need *that landed | 12:42 |
slagle | just trying to update tripleo so we can move forward without breaking | 12:42 |
dprince | slagle: I'm thinking that is the wrong reaction here though. If we do that we are ignoring the fact that others may hit these same packaging issues. | 12:44 |
dprince | slagle: seems like the right thing to do is fix the package | 12:44 |
openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Update error message for build-docker-images https://review.openstack.org/171822 | 12:45 |
slagle | dprince: i agree with you here | 12:49 |
slagle | i didnt push back hard enough in the interest of just actually getting things working again | 12:49 |
slagle | given the packaging change had already landed, and the puppet-glance had already landed, it felt to me like moving forward as-is might be the easiest way to go | 12:50 |
*** tzumainn has joined #tripleo | 12:51 | |
dprince | slagle: yeah, well that is what reverts are for. Now we just need 2 of them :) | 12:52 |
*** morazi has quit IRC | 12:54 | |
EmilienM | dprince: I can revert that one, if the packaging one is reverted too. | 12:56 |
dprince | EmilienM: I'm tempted to say this is the right thing to do. But I'd like to at least try chatting with people first | 12:57 |
EmilienM | dprince: ok. I just want to help , let me know | 12:58 |
*** dmellado has quit IRC | 13:02 | |
openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Refresh description for swift/cinder/ceph storage nodes https://review.openstack.org/174858 | 13:03 |
*** regebro has joined #tripleo | 13:03 | |
openstackgerrit | Merged openstack/tuskar-ui: Show required service configuration parameters as required https://review.openstack.org/166235 | 13:09 |
*** lucas-hungry is now known as lucasagomes | 13:09 | |
*** jprovazn has joined #tripleo | 13:13 | |
*** adrianopetrich has quit IRC | 13:14 | |
dprince | slagle: this requires element changes too. it will break our seed as well in CI | 13:24 |
*** radez_g0n3 is now known as radez | 13:24 | |
*** bvandenh has quit IRC | 13:26 | |
slagle | good point | 13:27 |
dprince | slagle: I'm not "fixing" it there BTW | 13:28 |
dprince | slagle: just sayin | 13:28 |
EmilienM | dprince: do you want me to restore my puppet fix in the meantime? https://review.openstack.org/#/c/174375/ | 13:28 |
slagle | well now we're nitpicking on the revert about rpm macros | 13:28 |
dprince | EmilienM: no, I think we are going to fix the packaging side (for real this time) | 13:29 |
EmilienM | dprince: ack | 13:30 |
slagle | EmilienM: dprince : we will need to rever the original puppet-glance change | 13:32 |
slagle | revert | 13:32 |
slagle | https://review.openstack.org/#/c/172440/ | 13:32 |
slagle | assuming the packaging revert goes in | 13:32 |
EmilienM | slagle: ok, I'm doing it | 13:35 |
*** prad has joined #tripleo | 13:35 | |
*** pradeep has joined #tripleo | 13:35 | |
*** Marga_ has quit IRC | 13:35 | |
*** rhallisey has quit IRC | 13:35 | |
dprince | slagle: since I'm spending my morning on glance packages... :) | 13:35 |
dprince | slagle: my *only* issue with the split was that the previous openstack-glance (meta) package no longer installed both the API and the Registry | 13:36 |
dprince | slagle: so long as we still did that I'd be totally fine with the changes I think | 13:36 |
dprince | slagle: would you concur? | 13:36 |
slagle | dprince: yes. you would have to convince the packaging folks about that. there the ones that balked at the "correctness" | 13:37 |
*** rhallisey has joined #tripleo | 13:37 | |
*** pblaho_ has quit IRC | 13:37 | |
dprince | slagle: okay. So let me land just that as a fix. And you can't balk at something which breaks end users anyways | 13:37 |
*** pblaho has joined #tripleo | 13:41 | |
*** morazi has joined #tripleo | 13:48 | |
*** pblaho has quit IRC | 13:49 | |
*** pblaho has joined #tripleo | 13:51 | |
*** bnemec has joined #tripleo | 14:03 | |
*** BadCub_Hiding is now known as BadCub | 14:06 | |
*** sdake has joined #tripleo | 14:12 | |
*** sdake_ has joined #tripleo | 14:47 | |
*** sdake has quit IRC | 14:50 | |
shardy | dprince: re https://review.openstack.org/#/c/174150/ | 14:52 |
shardy | I think we need a different pattern, where we don't apply any config by default, so there's no need to wait for an empty config script | 14:53 |
shardy | https://review.openstack.org/#/c/172050/2/controller-post.yaml | 14:53 |
shardy | dprince: something similar to that, where we just create an empty nested stack as a noop default | 14:54 |
*** avozza is now known as zz_avozza | 14:54 | |
shardy | Or, have the StructuredDeployment able to go directly to CREATE_COMPLETE if it's passed no config | 14:54 |
shardy | I'll work an alternate example up | 14:55 |
dprince | shardy: sure, totally fine with that | 14:55 |
dprince | shardy: My goal was to have the validations take no extra time | 14:56 |
shardy | dprince: Yeah, hopefully I can find an approach which does that, but also plays nice with breakpoints | 14:56 |
shardy | another layer of abstraction will probably do it ;) | 14:57 |
dprince | shardy: hopefully not too complex though | 14:57 |
*** prad has quit IRC | 14:57 | |
dprince | slagle: I just sent the openstack-glance revert it. The way I see it we can sort out another split patch next week that doesn't break people. | 14:58 |
shardy | dprince: No, it should be pretty simple I think | 14:58 |
*** adrianopetrich has joined #tripleo | 15:00 | |
dprince | spredzy, EmilienM: just a heads up... shardy is working on another iteration of https://review.openstack.org/#/c/174150/2 | 15:00 |
dprince | shardy: btw, I hit a really weird issue in trying to use get_file: get_param.... | 15:01 |
*** prad has joined #tripleo | 15:01 | |
*** yamahata has joined #tripleo | 15:01 | |
dprince | shardy: it couldn't find the file | 15:01 |
dprince | shardy: essentially what I was tyring to do in patchset 1 (of my patch) | 15:01 |
dprince | shardy: so I removed it, but I may file a bug on it because it should have worked I think | 15:02 |
*** jprovazn has quit IRC | 15:02 | |
*** Marga_ has joined #tripleo | 15:03 | |
*** radez is now known as radez_g0n3 | 15:03 | |
shardy | dprince: I'm not sure we can make that work | 15:04 |
shardy | dprince: because the get_file is run in python-heatclient, before the template is parsed | 15:04 |
dprince | shardy: oh, is it a lazy load thing | 15:04 |
shardy | so we have to pass all the files to heat from the client, then resolve all the template references | 15:05 |
dprince | shardy: oh, right :( | 15:05 |
shardy | to fix it we'd have to make heatclient capable of fully resolving the template | 15:05 |
dprince | shardy: so then perhaps we need to fail fast in heatclient on this | 15:05 |
shardy | maybe there's a way to do it via a call to template_validate, then do the get_file | 15:05 |
dprince | shardy: I get it. Just something that seems valid to try and the error was unclear | 15:05 |
shardy | dprince: yeah, definitely, please raise a bug and we'll figure out if we can fix it, or fail better | 15:06 |
*** morazi_ has joined #tripleo | 15:06 | |
*** dsneddon has joined #tripleo | 15:06 | |
*** radez_g0n3 is now known as radez | 15:06 | |
*** morazi has quit IRC | 15:08 | |
*** jistr is now known as jistr|mtg | 15:16 | |
slagle | dprince: wfm. | 15:16 |
*** morazi_ is now known as morazi | 15:21 | |
*** asparks has joined #tripleo | 15:24 | |
asparks | ok, this is getting strange and annoying, dozer1 is off again | 15:24 |
asparks | sorry wrong place | 15:24 |
gfidente | EmilienM, jistr|mtg do you have any opinion over either include puppet-pacemaker in puppet-openstack OR add pcsd to puppet-corosync? | 15:24 |
EmilienM | gfidente: what is puppet-openstack ? | 15:27 |
EmilienM | gfidente: we have this: https://github.com/stackforge/puppet-openstack_extras/tree/master/manifests/pacemaker | 15:27 |
EmilienM | that relies on puppetlabs-corosync | 15:27 |
*** dprince has quit IRC | 15:28 | |
*** hewbrocca has quit IRC | 15:32 | |
*** pdb has quit IRC | 15:32 | |
*** pdb has joined #tripleo | 15:33 | |
*** zz_avozza is now known as avozza | 15:34 | |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Fix image size to fit filesystem journal https://review.openstack.org/169166 | 15:36 |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Add functional smoke test for disk-image-create https://review.openstack.org/169167 | 15:36 |
*** dtantsur is now known as dtantsur|afk | 15:37 | |
*** asparks has left #tripleo | 15:37 | |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Initial element tests https://review.openstack.org/174681 | 15:37 |
*** jistr|mtg is now known as jistr | 15:44 | |
shardy | EmilienM: Hey, FYI I've updated https://review.openstack.org/#/c/162109/ | 15:57 |
EmilienM | shardy: will look | 15:58 |
*** alop has joined #tripleo | 15:58 | |
shardy | EmilienM: it requires two in-progress heatclient patches, and https://review.openstack.org/161927 | 15:58 |
shardy | my testing indicates it all works fine except if you let an update time out, then we hit a problem with the heatclient hook-poll behavior | 15:59 |
shardy | that's related to a heat API bug I'm looking into | 15:59 |
shardy | it should be good enough for folks to start trying now though, if you're interested | 15:59 |
EmilienM | shardy: that what I was waiting to continue | 16:00 |
*** kbyrne has quit IRC | 16:00 | |
*** dprince has joined #tripleo | 16:02 | |
*** mark-os has quit IRC | 16:04 | |
*** lblanchard has quit IRC | 16:06 | |
*** yamahata has quit IRC | 16:08 | |
*** mmagr has quit IRC | 16:08 | |
*** trown is now known as trown|lunch | 16:08 | |
*** pcaruana has quit IRC | 16:10 | |
*** aufi has quit IRC | 16:11 | |
*** lblanchard has joined #tripleo | 16:14 | |
*** subscope has joined #tripleo | 16:20 | |
gfidente | jistr, you around? | 16:27 |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: Initial element tests https://review.openstack.org/174681 | 16:27 |
jistr | gfidente: yah | 16:27 |
gfidente | so CI is currently using puppet-corosync and setting up the one node cluster | 16:28 |
gfidente | jistr, what I would do is 1) add puppet-pacemaker 2) switch the cluster config to pacemaker module 3) remove puppet-corosync | 16:29 |
gfidente | which we should be able to do without breakage | 16:29 |
gfidente | makese sene? | 16:29 |
*** yamahata has joined #tripleo | 16:31 | |
jistr | gfidente: yes | 16:31 |
*** stendulker has joined #tripleo | 16:31 | |
stendulker | bnemec: Hi | 16:33 |
stendulker | bnemec: Can you please look at followig patches when you get time. | 16:34 |
*** jcoufal has quit IRC | 16:34 | |
stendulker | bnemec: https://review.openstack.org/#/c/172715/ | 16:34 |
stendulker | bnemec: https://review.openstack.org/#/c/172721/ | 16:35 |
*** sdake has joined #tripleo | 16:35 | |
stendulker | bnemec: Thank you. | 16:35 |
*** derekh has quit IRC | 16:37 | |
greghaynes | huh. | 16:38 |
*** sdake_ has quit IRC | 16:39 | |
openstackgerrit | Steven Dake proposed stackforge/kolla: WIP: Configure neutron for network topology changes https://review.openstack.org/174953 | 16:39 |
greghaynes | ah, theres a comment answering my question - how do you test it | 16:39 |
openstackgerrit | Giulio Fidente proposed openstack/tripleo-puppet-elements: Add puppet-pacemaker module to puppet-modules https://review.openstack.org/174954 | 16:39 |
gfidente | dprince, ^^ sorry about that but ... looks like we're switching to puppet-pacemaker and dropping puppetlabs-corosync instead | 16:40 |
*** Marga_ has quit IRC | 16:43 | |
dprince | gfidente: fine with me | 16:44 |
dprince | gfidente: this was initially what confused me, but its all good | 16:44 |
dprince | gfidente: I'm no pacemaker expert | 16:45 |
gfidente | dprince, neither am I :P | 16:46 |
EmilienM | I think not much people are. | 16:47 |
*** stendulker has quit IRC | 16:48 | |
*** jistr has quit IRC | 16:48 | |
dprince | EmilienM: +2 on that change. Once ci is happy send away! | 16:48 |
EmilienM | dprince: what's the status of HA ci jobs? | 16:49 |
dprince | EmilienM: I'll poke infra again to get us progress there | 16:49 |
dprince | EmilienM: waiting on spredzy's patch first | 16:49 |
dprince | gfidente, EmilienM: I'll push a patch to switch tripleo-ci's HA job to be puppet (in tripleo-ci) | 16:51 |
gfidente | there is one from spredzy | 16:52 |
dprince | gfidente: not the infra one, in tripleo-ci | 16:52 |
dprince | gfidente: did I miss something? | 16:52 |
gfidente | dprince, ah sorry | 16:52 |
gfidente | nope I misunderstood, there is no change up for tripleo-ci | 16:53 |
gfidente | not that I know of at least | 16:54 |
dprince | gfidente: Yeah. I don't see it. | 16:54 |
*** pblaho has quit IRC | 16:54 | |
dprince | gfidente: 3 controllers you think? (like the old HA job) | 16:54 |
dprince | gfidente: I'm going to generify the puppet stuff a bit too | 16:54 |
dprince | gfidente: or is 2 controllers enough | 16:55 |
* dprince thinks 2 controllers is fine | 16:55 | |
gfidente | dprince, we need three in the real world scenario because 2 can't make a quorum | 16:57 |
dprince | gfidente: good point, 3 it is | 16:57 |
gfidente | dprince, but I am not sure if for the sake of CI having 2 is so terrible | 16:57 |
dprince | gfidente: well. it likely effects whether or not a ceph job is separate (with a split out Ceph node) | 16:58 |
dprince | gfidente: it *may* effect that. | 16:58 |
gfidente | dprince, have we ever had any periodic job for tripleo? | 16:58 |
gfidente | dprince, would it make sense to test ha in a periodic job only? | 16:59 |
dprince | gfidente: not that I know of. We've talked about it several times... Probably not a good case for HA as breaking it is likely easy to do | 16:59 |
*** ifarkas has quit IRC | 17:00 | |
gfidente | I think for ceph we might attach it to some non-ha deployment because the cluster at the ceph level is not in charge to pacemaker anyway but managed by ceph itself | 17:01 |
gfidente | unless we get to test things like migration or inject failure into the ha scenario | 17:02 |
*** prad has quit IRC | 17:02 | |
*** alop has quit IRC | 17:02 | |
dprince | gfidente: yep, The CI on ceph is what is driving my build-images stuff BTW. If CI works this weekend I may push a patch for that too | 17:03 |
dprince | gfidente: heat env customizations too... | 17:03 |
*** lucasagomes is now known as lucas-beer | 17:04 | |
*** prad has joined #tripleo | 17:05 | |
gfidente | so taking ceph on a side, for the ha job, I think safest thing is start with 3 and check if we can make it work with two | 17:05 |
*** prad has quit IRC | 17:06 | |
gfidente | rabbit is the single thing which I think won't accept connections until there is a quorum | 17:06 |
*** akrivoka has quit IRC | 17:07 | |
dprince | gfidente: I think all that is needed is for you to fix this: https://review.openstack.org/#/c/173263/2 | 17:08 |
dprince | gfidente: way TOCI_JOB_TYPE is split up already works with the puppet and the HA stuff | 17:09 |
dprince | gfidente: only point I had was that USE_MERGEPY should be the default, with only the precise-overcloud job being the exception | 17:09 |
*** prad has joined #tripleo | 17:09 | |
dprince | gfidente: if you make that change then +2 we are good to go for the HA job I think | 17:10 |
gfidente | dprince, yeah I couldn't find an easy logic to achieve that though | 17:10 |
gfidente | how do I match the precise-overcloud job only there? | 17:10 |
gfidente | either precise, overcloud or nonha match also other jobs | 17:11 |
gfidente | unless we set it for 'precise' , given the only other precise job is undercloud ? | 17:12 |
dprince | gfidente: yeah, this is a bit fancy isn't it | 17:12 |
dprince | gfidente: forget it. Lets just go with your fix as it | 17:12 |
gfidente | well 'precise' seems to do it | 17:13 |
gfidente | oh well | 17:13 |
gfidente | we can add a suffix to the job name | 17:13 |
gfidente | which would make more sense to me actually | 17:13 |
gfidente | and match based on that | 17:13 |
dprince | gfidente: precise would do it | 17:14 |
gfidente | dprince, ack | 17:15 |
slagle | dprince: hah, i was wondering how tripleo was even passing considering the old deloren pin and the updated puppet-glance | 17:18 |
slagle | dprince: it was this: https://review.openstack.org/#/c/174382 | 17:18 |
slagle | you'd already worked around the initial issue | 17:18 |
openstackgerrit | Giulio Fidente proposed openstack-infra/tripleo-ci: Use merge.py only for overcloud-precise-nonha job https://review.openstack.org/173263 | 17:18 |
dprince | slagle: right, that confused me as well. That change is correct regardless of this opentack-glance package revert (the revert of the split out packages) | 17:21 |
dprince | slagle: so we want to keep that regardless | 17:21 |
dprince | gfidente: nice, thanks | 17:22 |
dprince | slagle: can you review https://review.openstack.org/#/c/173263/3 | 17:22 |
dprince | slagle: if infra lands Yanis's patch to make fedora20-ha work this afternoon I'd like to have things inline | 17:23 |
dprince | slagle: still need to wait for the CI passes... | 17:23 |
slagle | sure, lgtm | 17:24 |
dprince | gfidente: so hey, build-images... | 17:26 |
dprince | gfidente: I want a load-images features as well that will load any images built via build-images. This could be part of build-images but I'm inclined to make it a separate script. | 17:27 |
dprince | gfidente: thoughts? | 17:27 |
dprince | Fedora20 HA job is now +A https://review.openstack.org/#/c/173255/ | 17:28 |
dprince | EmilienM: ^^^ | 17:28 |
gfidente | it is something more than existing load-image ? | 17:29 |
dprince | gfidente: it needs to loop over images from a YAML file. It would actually just call load-image | 17:29 |
dprince | gfidente: a simple python script to call load-image in sequence | 17:29 |
dprince | gfidente: I could probably "jq it up" and make load-image do it though too | 17:30 |
* dprince hates jq sometimes though | 17:30 | |
gfidente | I think the job will fail though until we get rabbit/galera under control with pacemaker or the keepalived based submissions | 17:31 |
dprince | gfidente: doesn't matter, the pressure is now on you sir :) | 17:31 |
gfidente | dprince, think bnemec would only plusvote something python based :) | 17:31 |
gfidente | and he's better than me at arguing as well | 17:32 |
dprince | gfidente: just figured I'd ask. Will push something soon... | 17:32 |
* bnemec likes Python things | 17:32 | |
dprince | bnemec: you are in the right project then I think | 17:33 |
* dprince should probably look for a new job though | 17:33 | |
bnemec | dprince: But tripleo is all bash. :-/ | 17:33 |
dprince | bnemec: right well you've got oslo for that too | 17:34 |
dprince | bnemec: small doses of non-Python won't kill you | 17:34 |
bnemec | Yeah, the split has been more large doses of bash, tiny doses of Python once in a while lately though. | 17:35 |
bnemec | And puppet, which I'm still reserving judgment on. | 17:36 |
dprince | bnemec: good man | 17:36 |
gfidente | bnemec, but if you had to compare with elements (bash) ... | 17:36 |
EmilienM | dprince: w00t | 17:42 |
*** Goneri has quit IRC | 17:43 | |
*** gfidente is now known as gfidente|afk | 17:45 | |
*** prad has quit IRC | 17:45 | |
*** prad has joined #tripleo | 17:46 | |
*** athomas has quit IRC | 17:48 | |
*** trown|lunch is now known as trown | 17:50 | |
*** mcornea has quit IRC | 17:50 | |
*** Marga_ has joined #tripleo | 17:55 | |
*** radez is now known as radez_g0n3 | 17:56 | |
*** radez_g0n3 is now known as radez | 17:57 | |
*** alop has joined #tripleo | 18:20 | |
openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Perform basic setup of Pacemaker cluster using puppet-pacemaker https://review.openstack.org/174994 | 18:20 |
dprince | shardy: your comments about devtest_overcloud_images.sh here, https://review.openstack.org/#/c/173645/ | 18:21 |
dprince | shardy: I mostly copied those as-is out of devtest_overcloud.sh | 18:21 |
dprince | shardy: So I suppose I'd rather save most of the edits to another patch | 18:21 |
*** avozza is now known as zz_avozza | 18:34 | |
*** lucas-beer has quit IRC | 18:44 | |
*** lblanchard has quit IRC | 18:46 | |
*** lblanchard has joined #tripleo | 18:50 | |
*** zz_avozza is now known as avozza | 18:53 | |
*** pelix has quit IRC | 19:07 | |
mordred | dprince, gfidente|afk: btw - I have an ansible role that knows how to build one or more images using dib and then upload them to one or more clouds - all driven by a yaml file of course | 19:12 |
mordred | dprince, gfidente|afk: it needs maybe 5 minutes of me cleaning a personal assumption out of it because I'm the only consumer at the moment ... but if it's useful to you for the thing you were talking about above, I'm happy to do that | 19:13 |
openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Perform basic setup of Pacemaker cluster using puppet-pacemaker https://review.openstack.org/174994 | 19:15 |
*** bnemec is now known as beekneemech | 19:22 | |
*** radez is now known as radez_g0n3 | 19:22 | |
*** yamahata has quit IRC | 19:31 | |
dprince | mordred: that would mean an ansible requirement for TripleO devtest | 19:32 |
dprince | mordred: I'm interested | 19:32 |
mordred | dprince: https://github.com/emonty/ansible-build-image | 19:32 |
dprince | mordred: but I'm not sure everyone would want to make tripleo devtest require ansible always... | 19:33 |
mordred | dprince: https://review.openstack.org/#/c/140840/12/tools/test-dib-upload.yaml <-- there is an example of using it | 19:33 |
mordred | dprince: both have slightly too much hardcoded right now - but that's just because I've been hacking on them | 19:34 |
dprince | mordred: I sort of like a separation between the building and the loading I think | 19:34 |
mordred | dprince: and yah - I can imagine folks may not want that depend :) | 19:34 |
dprince | mordred: both tools can share the same YAML format though | 19:35 |
mordred | yup | 19:35 |
mordred | I should probably split those into a build and an upload tbh | 19:35 |
dprince | mordred: no surprise our config file formats aren't too different though. See comment here: https://review.openstack.org/#/c/173644/3/scripts/build-images,cm | 19:35 |
mordred | dprince: hah. nice | 19:35 |
mordred | yeah - there's really not that many different ways to express this data | 19:36 |
greghaynes | Eventually if you use the nodepool stuff youll get the build farming aspect too | 19:36 |
greghaynes | which would be nice | 19:36 |
greghaynes | obviously a way down the road thing though | 19:36 |
mordred | yup | 19:36 |
greghaynes | although if youre going the 1 image route thats less of an issue | 19:36 |
dprince | greghaynes: some in our camp are interested in 1 image. My goal is let people cut it like the see it | 19:37 |
greghaynes | :) | 19:37 |
dprince | greghaynes: I like very minimal images myself, and install at deploy time | 19:37 |
greghaynes | infra style | 19:38 |
*** Marga_ has quit IRC | 19:38 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Use merge.py only for overcloud-precise-nonha job https://review.openstack.org/173263 | 19:39 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Remove DIB cherry-pick (package updating, merged) https://review.openstack.org/173728 | 19:40 |
*** subscope has quit IRC | 19:40 | |
openstackgerrit | Merged openstack/tripleo-incubator: Don't hard code the use-ephemeral element https://review.openstack.org/172209 | 19:41 |
*** funzo has quit IRC | 19:53 | |
openstackgerrit | Dan Prince proposed openstack/tripleo-incubator: build-images: drive DIB via YAML config file https://review.openstack.org/173644 | 19:54 |
openstackgerrit | Dan Prince proposed openstack/tripleo-incubator: devtest_overcloud.sh: split out image building https://review.openstack.org/173645 | 19:54 |
openstackgerrit | Dan Prince proposed openstack/tripleo-incubator: load-images: load images via YAML config file https://review.openstack.org/175028 | 19:54 |
openstackgerrit | Dmitry Teselkin proposed openstack/diskimage-builder: Allow dib to be lauched from venv https://review.openstack.org/168002 | 19:55 |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Add support for Glance RBD backend https://review.openstack.org/172032 | 19:56 |
*** jdob has quit IRC | 19:57 | |
*** Marga_ has joined #tripleo | 20:09 | |
*** funzo has joined #tripleo | 20:09 | |
*** gfidente|afk has quit IRC | 20:09 | |
*** Marga_ has quit IRC | 20:13 | |
openstackgerrit | Merged stackforge/kolla: Neutron doa in current form https://review.openstack.org/173627 | 20:16 |
*** dprince has quit IRC | 20:16 | |
*** morazi has quit IRC | 20:16 | |
*** bvandenh has joined #tripleo | 20:21 | |
openstackgerrit | Merged stackforge/kolla: Add workaround to work with cirros image https://review.openstack.org/174158 | 20:21 |
*** openstackgerrit has quit IRC | 20:22 | |
*** openstackgerrit has joined #tripleo | 20:22 | |
*** trown is now known as trown|outttypeww | 20:29 | |
*** jayg is now known as jayg|g0n3 | 20:35 | |
*** sdake_ has joined #tripleo | 20:56 | |
*** bvandenh has quit IRC | 20:59 | |
*** jpeeler has quit IRC | 20:59 | |
*** jpeeler has joined #tripleo | 20:59 | |
*** jpeeler has quit IRC | 20:59 | |
*** jpeeler has joined #tripleo | 20:59 | |
*** sdake has quit IRC | 20:59 | |
*** untriaged-bot has joined #tripleo | 21:00 | |
untriaged-bot | No untriaged bugs so far! \o/ | 21:00 |
*** untriaged-bot has quit IRC | 21:00 | |
*** avozza is now known as zz_avozza | 21:04 | |
-openstackstatus- NOTICE: Gerrit will be unavailable between 22:00 and 23:59 UTC for project renames and a database update. | 21:05 | |
*** Marga_ has joined #tripleo | 21:08 | |
*** BadCub is now known as Party_Bear | 21:17 | |
*** lblanchard has quit IRC | 21:18 | |
*** shardy has quit IRC | 21:20 | |
*** openstackgerrit has quit IRC | 21:23 | |
*** openstackgerrit has joined #tripleo | 21:23 | |
*** sdake has joined #tripleo | 21:25 | |
*** sdake_ has quit IRC | 21:29 | |
-openstackstatus- NOTICE: Gerrit is unavailable until 23:59 UTC for project renames and a database update. | 22:05 | |
-openstackstatus- NOTICE: Gerrit is unavailable until 23:59 UTC for project renames and a database update. | 22:09 | |
*** ChanServ changes topic to "Gerrit is unavailable until 23:59 UTC for project renames and a database update." | 22:09 | |
*** regebro has quit IRC | 22:16 | |
*** sdake_ has joined #tripleo | 22:24 | |
*** sdake has quit IRC | 22:27 | |
*** daneyon_ has joined #tripleo | 22:28 | |
*** rhallisey has quit IRC | 22:30 | |
*** daneyon has quit IRC | 22:31 | |
*** daneyon has joined #tripleo | 22:32 | |
*** daneyon_ has quit IRC | 22:32 | |
*** rlandy has quit IRC | 22:44 | |
*** sdake has joined #tripleo | 22:46 | |
*** sdake_ has quit IRC | 22:50 | |
*** ChanServ changes topic to "Do http://fpaste.org/211124/ to work around https://bugzilla.redhat.com/show_bug.cgi?id=1200043 | Deploying OpenStack Using OpenStack | Summit proposed topics: https://etherpad.openstack.org/p/tripleo-liberty-proposed-sessions | https://wiki.openstack.org/wiki/TripleO" | 23:07 | |
-openstackstatus- NOTICE: Gerrit is available again. | 23:07 | |
*** rhallisey has joined #tripleo | 23:13 | |
*** sdake_ has joined #tripleo | 23:14 | |
*** funzo has quit IRC | 23:15 | |
*** sdake has quit IRC | 23:17 | |
*** sdake has joined #tripleo | 23:32 | |
*** panda has quit IRC | 23:36 | |
*** sdake_ has quit IRC | 23:36 | |
*** panda has joined #tripleo | 23:36 | |
*** wfoster has quit IRC | 23:39 | |
*** alop has quit IRC | 23:39 | |
*** Marga_ has quit IRC | 23:50 | |
*** weshay has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!