Thursday, 2024-06-13

stanCan I have both s3 and file backend enabled for glance with kolla ansible? Basically I want a file backend with netapp nfs, they seem to do some internal optimisation for copying the glance image to cinder volume when they are using same backend. basically I want a highly available glance, when using a shared file backend?04:06
stanis this possible, I understand having a file backend that is not shared and having glance highly available, would lead to problems, but having a shared filesystem like nfs mounted on all controllers would greatly benefit from a highly available glance?04:08
stanI would think adding an additional mount to the glace container, and changing the globals yaml with glance_api_hosts and data_dir param would be enough?04:24
opendevreviewPierre Riteau proposed openstack/kayobe master: Use TinyIPA images from stable/2023.2  https://review.opendev.org/c/openstack/kayobe/+/92170506:56
opendevreviewMartin Hiner proposed openstack/kolla-ansible master: Move Nova roles after OVS  https://review.opendev.org/c/openstack/kolla-ansible/+/92184607:39
opendevreviewRafal Lewandowski proposed openstack/kolla-ansible master: [WIP] Enable ML2/OVN and distributed FIP by default  https://review.opendev.org/c/openstack/kolla-ansible/+/90495907:58
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: Bump Ansible versions to 2.16 and 2.17  https://review.opendev.org/c/openstack/kolla-ansible/+/92174308:12
opendevreviewPierre Riteau proposed openstack/kayobe master: Revert "[release] Use OpenStack 2024.1 release"  https://review.opendev.org/c/openstack/kayobe/+/92192208:43
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: Bump Ansible versions to 2.16 and 2.17  https://review.opendev.org/c/openstack/kolla-ansible/+/92174308:51
opendevreviewRafal Lewandowski proposed openstack/kolla-ansible master: [WIP] Enable ML2/OVN and distributed FIP by default  https://review.opendev.org/c/openstack/kolla-ansible/+/90495909:08
opendevreviewGaël THEROND proposed openstack/kolla master: [WIP] - Add new hardware prometheus exporter:  https://review.opendev.org/c/openstack/kolla/+/92165809:20
Fl1ntHi everyone!09:22
Fl1ntquick question09:22
Fl1ntis there something existing on our image build process for sources where the name of the resulting folder once the archive is extracted is a bit clancky? 09:23
Fl1ntI'm working on this: https://review.opendev.org/c/openstack/kolla/+/92165809:23
Fl1ntbut right now, I'm kinda required to use a very dirty hardcoded release versioning on the Dockerfile template :'(09:24
Fl1ntI'm pretty sure you guys already thought about it, so I'm wondering ^^09:24
opendevreviewGaël THEROND proposed openstack/kolla master: Add new hardware prometheus exporter:  https://review.opendev.org/c/openstack/kolla/+/92165809:55
Fl1nt@mnasiadka, regarding the bmc_exporter name, I just renamed the binary as the original one was named idrac_exporter dispit supporting multiple bmc type (iLO/iDRAC/xClarity).09:58
mnasiadkait's either idrac exporter or bmc exporter, the idrac one looks bloody weird that all metrics are prefixed with idrac name09:59
Fl1ntyep, I'll switch to the one you suggested10:00
SvenKieskeFl1nt: are you referring to this? ./idrac_exporter-helm-idrac-exporter-${prometheus-bmc-exporter-version}/cmd/idrac_exporter10:03
SvenKieskenot sure what you are specifically asking for right now10:03
SvenKieskeif you are talking about https://review.opendev.org/c/openstack/kolla/+/921658/3/kolla/common/sources.py you should specify the github download location via the "releases" path on github imho, not via "achive/refs/tags", like the other packages do as well10:06
Fl1ntSvenKieske, yep10:06
SvenKieskemhm, do the releases only contain the helm charts (the name seems to inidcate that)? weird project structure..10:07
Fl1ntSvenKieske, the project structure is a bit weird, yes, the helm named archive is only containing the charts, the sources.tar.gz one does contain the code, hence why I can't use the release/ path10:08
SvenKieskewell if you gonna switch to another bmc exporter as mnasiadka suggested maybe the problem goes away ;) if not I would suggest opening a bug report at their github, maybe they are not aware that they are doing suboptimal releases?10:10
Fl1ntBut I'll switch to the one mnasiadka suggested as it's much better suited for our usage and structure on kolla10:10
SvenKieskenice10:10
Fl1ntI'll suggest them the change too as I already suggested them to publish the binaries as release.10:10
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: CI: Skip SLURP jobs on master  https://review.opendev.org/c/openstack/kolla-ansible/+/92193510:30
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: rabbitmq: Add support for using stream queues for fanout  https://review.opendev.org/c/openstack/kolla-ansible/+/91691110:32
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: [CI] Remove SLURP related ansible installation  https://review.opendev.org/c/openstack/kolla-ansible/+/92193710:38
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: [CI] Remove SLURP related ansible installation  https://review.opendev.org/c/openstack/kolla-ansible/+/92193710:39
opendevreviewRafal Lewandowski proposed openstack/kolla-ansible master: [WIP] Enable ML2/OVN and distributed FIP by default  https://review.opendev.org/c/openstack/kolla-ansible/+/90495911:28
Fl1ntmnasiadka, SvenKieske, any reason why we do not cleanup the symlink and temporary working directory on our exporter images? It duplicate data onto the image, sure it's not that of a big deal (few Mb) but yet seems weird enough.11:35
mnasiadkafeel free to raise a bug and fix it :)11:35
Fl1ntok, cool, not a feature tho?11:35
mnasiadkaI don't think so11:35
Fl1ntok, nice11:36
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: CI: Skip SLURP jobs on master  https://review.opendev.org/c/openstack/kolla-ansible/+/92193511:37
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: [CI] Remove SLURP related ansible installation  https://review.opendev.org/c/openstack/kolla-ansible/+/92193711:38
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: CI: Disable SLURP jobs for D cycle  https://review.opendev.org/c/openstack/kolla-ansible/+/92193511:39
opendevreviewBartosz Bezak proposed openstack/kolla-ansible master: [CI] Remove SLURP related ansible installation  https://review.opendev.org/c/openstack/kolla-ansible/+/92193711:40
Fl1ntarggh, the gebn exporter is using IPMI approach where the mrlhansen exporter do use the normalized redfish APIs to get much more informations, any thoughts on this?11:45
opendevreviewVerification of a change to openstack/kolla-ansible master failed: Move Nova roles after OVS  https://review.opendev.org/c/openstack/kolla-ansible/+/92184611:46
Fl1ntOn my own I'm fine with the IPMI approach for now tbh.11:52
SvenKieskeI hate to ask again, but what is the reasoning for the above barrage of CI changes? the commit messages are void of any reasoning, as usual.11:52
SvenKieskethis is not how we can foster more contributions to kolla, by making everything extra opaque and hard to understand :(11:53
opendevreviewGaël THEROND proposed openstack/kolla master: Add new hardware prometheus exporter:  https://review.opendev.org/c/openstack/kolla/+/92165811:54
opendevreviewMerged openstack/kolla-ansible master: CI: increase timeout for server resize confirmation  https://review.opendev.org/c/openstack/kolla-ansible/+/91639612:04
SvenKieskeokay I _think_ I know why these are removed, nvm, I guess I need to get some sleep after the next slew of meetings12:10
SvenKieskestill it would be nice if Commit Messages could adhere to https://docs.openstack.org/contributors/common/git.html#body12:11
SvenKieskenamely: " The body should contain all the important information related to the problem, without assuming that the reader understands the source of the problem or has access to external sites."12:12
SvenKieskeI honestly waste a lot of time on reviews chasing information not contained in commit messages which should be there. I don't know if I'm the only one.12:13
Fl1ntSvenKieske, yes, sometimes I do have long time digging for information too, but I don't have any insight on your example in here as I don't know what it is about exactly out of "CI related jobs".12:15
Fl1ntbut I do agree, meaningful commit body should be more important than stupid rules inherited from 1980 where screens couldn't handle more than 80 columns.12:16
SvenKieskewell to add the context for the CI Jobs: I think the reasoning is, that the next cycle upgrade will not be a slurp upgrade and we only need to reintroduce slurp upgrade testing after the next-next(?) release again on master branch? something like that I think is the cause for it, I really need some sleep.12:17
SvenKieskeand I agree with your perception of the rules in openstack land (to be fair they are not kolla specific afaik), but I wouldn't voice it this sharply. I can live with short lines, that's just annoying, not a big problem, there are bigger issues (like contributor count) :)12:19
SvenKieskemaybe I should finally start a blogpost on openstack contributing, title would be smth like "death by a thousand papercuts, or how to improve developer onboarding for large projects"12:50
SvenKieskeI must say I like the new additions in gerrit lately, e.g. the ability to have a "suggestion" button in the comment field is really nice, a small thing, but very nice.12:51
opendevreviewVerification of a change to openstack/kolla-ansible master failed: Move Nova roles after OVS  https://review.opendev.org/c/openstack/kolla-ansible/+/92184613:29
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: Bump Ansible versions to 2.16 and 2.17  https://review.opendev.org/c/openstack/kolla-ansible/+/92174314:28
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: CI: drop docker pin  https://review.opendev.org/c/openstack/kolla-ansible/+/92046014:29
opendevreviewMichal Nasiadka proposed openstack/kolla-ansible master: WIP: CI: Template out kolla-build.conf using role  https://review.opendev.org/c/openstack/kolla-ansible/+/90377114:30
opendevreviewMerged openstack/kolla-ansible master: Move Nova roles after OVS  https://review.opendev.org/c/openstack/kolla-ansible/+/92184616:10
Fl1ntSvenKieske, yep, I'm all in sync with your comments ^^18:11
opendevreviewGaël THEROND proposed openstack/kolla master: Add new hardware prometheus exporter:  https://review.opendev.org/c/openstack/kolla/+/92165820:01
opendevreviewGaël THEROND proposed openstack/kolla master: Add new hardware prometheus exporter:  https://review.opendev.org/c/openstack/kolla/+/92165820:03
opendevreviewFreerk-Ole Zakfeld proposed openstack/kolla-ansible master: Allow haproxy to listen to more than 1 address  https://review.opendev.org/c/openstack/kolla-ansible/+/91658620:34

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!