| stephenfin | sean-k-mooney: idk if you got that resolved (there's a lot of scrollback) but https://docs.openstack.org/nova/2026.1/install/ loads just fine for me? | 11:51 |
|---|---|---|
| sean-k-mooney | clarkb: fungi i have as this before in a differnt vain but would there be an objection to creating a lable for debian-13 that could fun on any provider excpt the raxspace xen cloud? it may not be needed but the kernel module im creating for emulating pci devices was unable to modprone on the xen vms because of a lack of mmio space in the vm. xen does not really emulate pci | 11:52 |
| sean-k-mooney | the saem as kvm or realisty. i think i can actully mkae the kernel model work on xen by making the mmio allcoation more robust but my pior quesiton is basiclly plan b. if i cant make it work relyibly on the xen nodes could we consider a lable so i can create a nodeset tha will avoid them | 11:52 |
| sean-k-mooney | stephenfin: install worked it was specificly the contibuetor docs that dont | 11:52 |
| stephenfin | isn't that unversioned? | 11:52 |
| stephenfin | like release notes | 11:52 |
| sean-k-mooney | https://docs.openstack.org/nova/2026.1/contributor/index.html | 11:52 |
| sean-k-mooney | no | 11:53 |
| sean-k-mooney | at least not that im aware of. i defeinly have links to older ones in my history | 11:53 |
| stephenfin | in any case, the only recent docs changes that I'm aware of is using tox to build release notes instead of calling sphinx-build directly | 11:53 |
| sean-k-mooney | i can see the argument for makign it that way but i dont thinkthat is how it has generally worked | 11:53 |
| stephenfin | but that's clearly irrelevnat | 11:53 |
| stephenfin | *irrelevant | 11:54 |
| sean-k-mooney | the tl;dr of why i notice this is i was try8ing to link to how runways used to work in nova for someone | 11:55 |
| sean-k-mooney | i just used the github link instead | 11:55 |
| sean-k-mooney | so its ntore really a big problem fom my presepctive but its just not what i was expecting to see | 11:55 |
| sean-k-mooney | and i dont knwo how or where this redirect has been configured | 11:56 |
| sean-k-mooney | so i was trying to figure out if this was intentional or a bug, and if its not a bug how is it working | 11:56 |
| sean-k-mooney | there is nothing in https://github.com/openstack/nova/blob/master/doc/source/conf.py to confiure this so it obvioulsy done as part fo the job, some external configuration that is non obvious form just looking at the repo | 11:59 |
| *** jgilaber_ is now known as jgilaber | 12:02 | |
| fungi | sean-k-mooney: consider using something like the debian-trixie-nested-virt-8GB label? xen doesn't have nested kvm acceleration, so we don't apply it in the xen regions | 13:25 |
| fungi | i expect some time in the near future rackspace will want us to wind down our use of their classic regions in favor of flex, at which point it won't matter | 13:26 |
| sean-k-mooney | fungi: that would also work yes | 13:27 |
| sean-k-mooney | i condier that breifly but nested virt is not stircly requried | 13:27 |
| sean-k-mooney | fungi: im goint to see if makeing the mmio code more robust is an option if not ill try the nested virt lable | 13:27 |
| fungi | sean-k-mooney: as for the docs.openstack.org redirects, they're a complicated ruleset managed in https://opendev.org/openstack/openstack-manuals/src/branch/master/www/.htaccess | 13:29 |
| sean-k-mooney | currntly the kernel dirver as for a whole 1MB!!! belvow the 4G moduary in teh 32bit mmio space. im going ot update it to supprot 64bit pci adressign which sould allwo other mmio windows to be used as well | 13:29 |
| sean-k-mooney | ah so https://opendev.org/openstack/openstack-manuals/src/branch/master/www/.htaccess#L208-L209 | 13:30 |
| sean-k-mooney | https://opendev.org/openstack/openstack-manuals/commit/4aed59d263f47458bd9e8d49f84b69f1fb611adf | 13:30 |
| sean-k-mooney | ok its been this way for 5 years | 13:30 |
| sean-k-mooney | ok i gues mystory solved | 13:31 |
| sean-k-mooney | fungi: i get the logic for why we woudl redirect it but i just didnt expect it so all good | 13:31 |
| sean-k-mooney | there are some part of the conti9butor docs that are branch specific | 13:32 |
| sean-k-mooney | liek the evenetlet/threading model but in practic when we care we normally are loong at the git repo anyway | 13:33 |
| sean-k-mooney | so for the redndered vesion thsi is effectivly a non issue | 13:33 |
| fungi | sean-k-mooney: there are also a lot of one-off carve outs in the rules there if a project needs something special for their case, so if there's a particular problem the blanket redirects are causing for parts of the nova docs it wouldn't be unusual to add another special case handler | 13:35 |
| sean-k-mooney | oh https://opendev.org/openstack/openstack-manuals is also where all the top level conte for th docs pages are | 13:35 |
| fungi | yes | 13:38 |
| fungi | that's the main repository for the docs site, essentially | 13:38 |
| fungi | and all the project-specific docs are basically grafted in by publishing at sub-paths under there | 13:39 |
| clarkb | yes I would just use a nested virt label for consistency with what we already have | 14:47 |
| clarkb | then it can be useful beyond a single experiment | 14:47 |
| opendevreview | Merged openstack/project-config master: Import drizzle revival and add a bindep-rs support project https://review.opendev.org/c/openstack/project-config/+/989417 | 18:06 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!