| *** mjturek has quit IRC | 00:32 | |
| *** chhagarw has joined #openstack-dib | 03:24 | |
| *** chhagarw has quit IRC | 03:28 | |
| *** chhagarw has joined #openstack-dib | 04:09 | |
| *** hwoarang has quit IRC | 04:16 | |
| *** hwoarang has joined #openstack-dib | 04:18 | |
| *** chhagarw has quit IRC | 06:22 | |
| *** chhagarw has joined #openstack-dib | 06:22 | |
| *** yolanda_ has joined #openstack-dib | 07:07 | |
| *** dtantsur has joined #openstack-dib | 09:03 | |
| dtantsur | hi folks! does it ring any bells? http://logs.openstack.org/47/653347/1/check/bifrost-integration-dibipa-debian-centos-7/5bcdc61/job-output.txt.gz#_2019-04-17_08_48_38_252694 | 09:04 |
|---|---|---|
| dtantsur | jessie seems also broken because jessie-backports is no longer a thing | 09:04 |
| dtantsur | what's the best supported distribution right now? ubuntu? | 09:05 |
| dtantsur | another job passed, so maybe the stretch issue is transient.. | 09:14 |
| *** strigazi has quit IRC | 10:09 | |
| *** strigazi has joined #openstack-dib | 11:25 | |
| *** strigazi has quit IRC | 11:50 | |
| *** strigazi has joined #openstack-dib | 11:51 | |
| *** strigazi has quit IRC | 11:52 | |
| *** strigazi has joined #openstack-dib | 11:52 | |
| *** strigazi has quit IRC | 11:55 | |
| *** strigazi has joined #openstack-dib | 11:55 | |
| *** dtantsur is now known as dtantsur|brb | 12:07 | |
| *** dtantsur|brb is now known as dtantsur | 13:07 | |
| *** mjturek has joined #openstack-dib | 13:22 | |
| *** Vorrtex has joined #openstack-dib | 13:27 | |
| *** mjturek has quit IRC | 14:25 | |
| *** mjturek has joined #openstack-dib | 14:29 | |
| clarkb | dtantsur: the infra team uses dib to build ubuntu centos fedora suse debian | 15:01 |
| clarkb | I would expect any of those to work if using an up to date release version | 15:01 |
| dtantsur | okay, thanks | 15:03 |
| clarkb | note we use the -minimal elements for the distros that exists for | 15:04 |
| *** dtantsur is now known as dtantsur|afk | 16:24 | |
| *** hwoarang has quit IRC | 16:54 | |
| *** hwoarang has joined #openstack-dib | 16:56 | |
| *** jangutter has quit IRC | 17:39 | |
| *** chhagarw has quit IRC | 17:43 | |
| *** chhagarw has joined #openstack-dib | 17:48 | |
| *** hwoarang has quit IRC | 18:08 | |
| *** hwoarang has joined #openstack-dib | 18:14 | |
| *** chhagarw has quit IRC | 18:23 | |
| *** chhagarw has joined #openstack-dib | 18:23 | |
| Vorrtex | Hey clarkb quick question... Can you think of an environment variable that contains what architecture is being built for in DIB? Or would I need to create that? | 18:50 |
| clarkb | Vorrtex: $ARCH set in diskimage_builder/lib/common-defaults | 18:51 |
| Vorrtex | clarkb awesome... off the top of your head is that set from the argument "-a" to disk-image-create ? | 18:52 |
| clarkb | no that seems set from uname -m? | 18:53 |
| clarkb | so it is the builders arch not necessarily the target arch (though I'm not sure how well dib supports cross builds) | 18:53 |
| Vorrtex | clarkb so you're not sure if there is a target architecture environment variable? | 18:54 |
| clarkb | I'm not. We do our arm64 builds on an arm64 host | 18:54 |
| clarkb | and x86 on x86 etc | 18:54 |
| clarkb | ianw may know when he starts his day | 18:54 |
| *** chhagarw has quit IRC | 18:55 | |
| Vorrtex | Right, that's what I would have expected in the first place, but thanks. I'll look for info in the mean time. Much appreciated. | 18:55 |
| *** mjturek has quit IRC | 20:20 | |
| ianw | cross building has been achieved ... but i don't think with master; it was a custom effort | 20:54 |
| Vorrtex | So IIUC, ianw you're saying that using the $ARCH environment variable will provide "current expected results" where the built image will be the host's architecture, is that correct? | 20:55 |
| ianw | Vorrtex: built image == host's architecture ; yes ... in other words not cross building. $ARCH is used to do things like pick different bootloader packages etc | 20:57 |
| ianw | but not to specify a full "give me an arm image even though i'm on x86_64" | 20:58 |
| Vorrtex | ianw all good. I have a review trying to add ppc64le to octavia's amphora building, and in one of the elements I'm using an "if/else" that includes "$(uname -m)" and it was suggested I use an environment variable from DIB instead. | 20:59 |
| ianw | ok, yeah, that would be a way we'd use it | 20:59 |
| Vorrtex | Yep, cool cool. | 20:59 |
| ianw | cross building *has* been achieved using basically qemu emulation, but never been "productised" if you will into the main repo | 21:00 |
| johnsom | ianw This looks correct? https://review.openstack.org/#/c/653488/2/elements/haproxy-octavia/pre-install.d/01-backports | 21:03 |
| johnsom | It seemed odd to use uname -m on the build system to select a yum report URL that is installed in the image. | 21:03 |
| johnsom | s/report/repo/g | 21:03 |
| Vorrtex | If "http://mirror.centos.org/altarch/7/paas" existed, I wouldn't need a different url... | 21:05 |
| ianw | yeah i would suggest $ARCH, to avoid this being the type of thing that needs patching for cross-building should we go that way :) | 21:05 |
| Vorrtex | Sorry, that sounded snappy, just wanted to point out why its a different url | 21:05 |
| Vorrtex | ianw right, will update to using $ARCH. | 21:06 |
| johnsom | Yeah, ok, that was my thought. | 21:06 |
| Vorrtex | ianw is $ARCH *not* set by uname -m in the first place? Or are you saying that *if* cross-support is added, then that variable would potentially be repurposed to accommodate a targeted arch instead of the host's arch? | 21:08 |
| ianw | Vortex: yep, the second bit | 21:10 |
| Vorrtex | Gotcha gotcha, just curious. Thanks! | 21:11 |
| *** logan- has quit IRC | 21:34 | |
| *** logan- has joined #openstack-dib | 21:37 | |
| *** Vorrtex has quit IRC | 21:38 | |
| *** fdegir has quit IRC | 21:55 | |
| *** zigo has quit IRC | 22:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!