| *** dmsimard is now known as dmsimard|off | 04:10 | |
| *** hashar has joined #zuul | 06:20 | |
| *** jesusaur has quit IRC | 07:23 | |
| *** jesusaur has joined #zuul | 08:31 | |
| *** smyers has quit IRC | 09:03 | |
| *** smyers has joined #zuul | 09:14 | |
| *** electrofelix has joined #zuul | 09:18 | |
| odyssey4me | dmsimard|off tristanC tobiash ah, buildah facilitates building OCI images: https://github.com/projectatomic/buildah | 09:47 |
|---|---|---|
| odyssey4me | pabelanger dmsimard|off I do think it'd be super nice if we could move from d-i-b to something that has broader community backing... maybe https://github.com/systemd/mkosi will take off and become that thing everyone uses | 09:50 |
| dmellado | odyssey4me: that looks nice, I would back it up xD | 10:03 |
| dmellado | dmsimard|off: ping me when you're back online, pls | 10:20 |
| *** jkilpatr has joined #zuul | 10:48 | |
| *** dkranz has joined #zuul | 12:26 | |
| *** dkranz has quit IRC | 13:07 | |
| dmsimard|off | FYI I'm tagging 0.14.4 for release today, delta from 0.14.3 is only Ansible 2.4 support https://github.com/openstack/ara/compare/0.14.3...master | 13:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Sort output in generated legacy files https://review.openstack.org/507118 | 14:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Merge project pipeline definitions https://review.openstack.org/507163 | 14:47 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Don't double quote rsync options https://review.openstack.org/506802 | 14:59 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Sort output in generated legacy files https://review.openstack.org/507118 | 15:00 |
| *** dkranz has joined #zuul | 15:13 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Fix node request assignment order https://review.openstack.org/507182 | 15:56 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Fix node request assignment order https://review.openstack.org/507182 | 16:03 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Fix node request assignment order https://review.openstack.org/507182 | 16:39 |
| jlk | ooh today is migration day? | 16:40 |
| *** bhavik1 has joined #zuul | 16:44 | |
| mordred | jlk: hell yah baby | 16:46 |
| mordred | odyssey4me: youknow it's hard to take anything in https://github.com/systemd seriously right? | 16:47 |
| odyssey4me | mordred haha, it is the borg, it shall assimilate you | 16:48 |
| mordred | odyssey4me: that said - dib could certainly be taught to use mkosi to make the lowest level root.d element things | 16:48 |
| odyssey4me | the point I was trying to get across is that this wheel has been invented all over the place, and it'd be super nice if we could integrate with other communities instead of continuing to perpetuate more islands | 16:49 |
| mordred | odyssey4me: as it seems to use dnf --installroot, debootstrap, pacstrap and zypper for the initial creatoin - which is what the -minimal elementsof dib do | 16:49 |
| odyssey4me | ja, at the very least if that project takes off it might help reduce some code in d-i-b | 16:49 |
| clarkb | odyssey4me: except our island was there first... | 16:49 |
| mordred | odyssey4me: oh - totally. EXCEPT - when we wrote dib there were not tools that existed that did what it did | 16:49 |
| mordred | and what clarkb said - nobody has wanted to take the next step - everyone wants to rewrite from scatach. oh well | 16:50 |
| odyssey4me | for sure, we often invent good things - perhaps we should share them better ;) | 16:50 |
| mordred | yah | 16:50 |
| odyssey4me | well, we share well - but perhaps we should advocate what we do to other communities better | 16:50 |
| clarkb | mordred: makes an important point dib is a good place to consume these suggsted replacements rather than wholesale replacement | 16:51 |
| clarkb | red hat has a thing they keep trying to get dib replaced with too but makes more sense having that tool bootstrap the image for dib imo | 16:51 |
| clarkb | because then you can have reconsumable image modifications over the top that aren't tool opinionated | 16:52 |
| pabelanger | I hear people like virt-customize :) | 16:52 |
| clarkb | the cache is the other big dib feature that is lacking in many tools though mkosi seems to at least grok that for packages | 16:54 |
| *** electrofelix has quit IRC | 16:56 | |
| clarkb | odyssey4me: fwiw DIB has been advocated reasonably well and is used in non openstack places. I think image building is very much an opinionated problem space. Distros don't want their users making undistro like choices, outside the distro space you have image providers (say rax) that have opinions on what images should look like (nova agent anyone). Its a particularly bda problem space for NIH for reasons | 17:00 |
| odyssey4me | the more I explore dib, the more I see how flexible and useful it is | 17:02 |
| odyssey4me | it's a little daunting at first though | 17:02 |
| odyssey4me | nowhere near as daunting as openstack ;) | 17:03 |
| clarkb | juts off the top of my head I can name a half dozen image build tools. coreos image builder, docker, packer, dib, virt-customize, and the thing ubuntu uses that I forget its name | 17:03 |
| clarkb | odyssey4me: ya I think the thing that makes dib unique in the space is it tried to be less opinionated and allow you tobuild images the way you want. Which has made it more complicated | 17:03 |
| clarkb | its more of a framework for executing image builds. Provides the run levels and run parts execution and the idea of a cache | 17:04 |
| clarkb | but beyond that you can build in chroots, in containers, in VMs etc if you want | 17:04 |
| clarkb | https://github.com/CanonicalLtd/ubuntu-image is the ubuntu one | 17:05 |
| jlk | I wrote an image builder at one point | 17:08 |
| jlk | https://docs.pagure.org/pungi/about.html | 17:09 |
| jlk | and there's OS-Tree | 17:10 |
| jlk | which is the basis of a number of image tools https://ostree.readthedocs.io/en/latest/ | 17:10 |
| *** hashar is now known as hasharDinner | 17:19 | |
| *** harlowja has joined #zuul | 17:20 | |
| *** bhavik1 has quit IRC | 17:20 | |
| jlk | mordred: anything I can do to help in the lead up? | 17:47 |
| clarkb | jlk: https://etherpad.openstack.org/p/zuulv3-cutover is tracking the work for toda I think | 17:48 |
| clarkb | there are changes that cna be reviewed | 17:48 |
| * jlk rolls up his eyeball sleeves | 17:48 | |
| jeblair | github bug for the backlog: http://paste.openstack.org/show/621862/ | 18:08 |
| jlk | oh fun! | 18:09 |
| jlk | okay | 18:09 |
| pabelanger | statuses (reads as a snake) :) | 18:29 |
| clarkb | what is "filesystem layouts on executors"? | 18:30 |
| clarkb | is that hte /opt update? | 18:30 |
| clarkb | er wrong window | 18:32 |
| *** hasharDinner is now known as hashar | 19:49 | |
| *** dkranz has quit IRC | 19:54 | |
| *** hashar has quit IRC | 19:59 | |
| *** hashar has joined #zuul | 19:59 | |
| *** contra-bot has joined #zuul | 21:02 | |
| *** contra-bot has quit IRC | 21:03 | |
| *** contra-bot has joined #zuul | 21:03 | |
| *** contra-bot has joined #zuul | 21:04 | |
| *** contra-bot has quit IRC | 21:06 | |
| *** hashar has quit IRC | 21:07 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: It's branches, not branch https://review.openstack.org/507285 | 21:40 |
| fungi | given the ensuing excitement in #openstack-infra, should we consider skipping the zuul meeting in 15 minutes? | 21:45 |
| Shrews | for sure | 21:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!