[07:47] <cpaelzer> athos: I've had a look and found your fix in the salsa repo (after I've created the same, gladly it is trivial) :-)
[07:48] <cpaelzer> athos: I've also looked on the changes landed upstream since 1.2.0 but they change a lot (listed attributes, rework of the callout handling, ...), I can't just pick those without upstream declaring it "the thing" by tagging 1.3.0
[07:49] <cpaelzer> athos: while dev seems to be done on fedora (some hints in the commits indicate that) this also got no new version yes
[07:49] <cpaelzer> athos: so my upload will for now be just for your dependency fix
[07:53] <cpaelzer> athos: this time I really pushed the tag and branch :-)
[09:54] <uzee> Hi, is there a way in ubuntu 22 server autoinstall to get hostname from dhcp server? there doesn't seem to be user-data or identity section options to do that. the only references I found were like early or late commands. Is that the right approach?
[11:55] <Mmike> Where do I ask about do-release-upgrade (going from 20.04 to 22.04) removing libvirt-clients and some other needed packages? Is this a bug, or am I missing something?
[11:57] <cpaelzer> Mmike: it sounds wrong indeed
[11:58] <cpaelzer> Mmike: does it tell you due to which conflict it would want to remove it?
[11:59] <Mmike> cpaelzer, I just tested it via lxd, started xenial image, installed only libvirt-bin, all went fine up to 20.04 (on bionic it installed libvirt-clients, on 20.04 it removed libvirt-bin), but then on 22.04 upgrade libvirt-clients and bunch of others got removed
[11:59] <Mmike> it did not tell me why it was removing it
[11:59] <cpaelzer> I did the same test starting at focal and did not hit it
[11:59] <cpaelzer> redoing the same starting from xenial
[11:59] <Mmike> and I didn't really go trough logs - trying to repeat the same in KVM now, not sure how lxd images are close to vanilla ubuntu-server installs I have. 
[11:59] <Mmike> cpaelzer, start from bionic
[12:00] <cpaelzer> what except libvirt-clients did you install on xenial manually before an upgrade?
[12:00] <cpaelzer> s/on xenial/on bionic/
[12:00] <Mmike> just libvirt-bin 
[12:00] <cpaelzer> ah, that could be it
[12:01] <cpaelzer> libvirt-bin was a virtual package for ages and ages
[12:01] <cpaelzer> and being gone might move libvirt-clients to auto-remove or such
[12:01] <Mmike> yup, but was not on xenial :) 
[12:02] <Mmike> so folks having their VMs on xenial had libvirt-bin installed (as virsh was part of it, and so on)... and once they (well, me) upgraded to 22.04, quite-a-few libvirt-* packages got yanked
[12:02] <cpaelzer> hrm, that is wrong indeed
[12:02] <cpaelzer> maybe I forgot a provides or anything like it along the way
[12:02] <cpaelzer> so TL;DR a bug indeed
[12:02] <cpaelzer> would you file it or should I?
[12:03] <Mmike> cpaelzer, I'd appreciate if you could do it, since I left canonical I never gotten to re-activating my ubuntu-one account (can't find the recovery codes :/), so I don't have access to launchpad atm.
[12:03] <Mmike> I can bribe you with some Lagavulin in Madrid, if you're going to be there? :D 
[12:04] <cpaelzer> I'll be there, but not doing it for the bribing except you bribe by re-activating that account :-)
[12:05] <Mmike> oh! didn't think of that at all!!! :D Thank you! :D
[12:05] <Mmike> So, this is my container, upgraded from xenial to focal: https://pastebin.com/4JFf3a0h
[12:06] <athos> thanks, cpaelzer (mdevctl)! 
[12:11] <athos> cpaelzer: btw, the ppc64el build failed (flaky?). Could we retry that one?
[12:16] <Mmike> cpaelzer, so, there packages are removed during final stage of do-release-upgrade from 20.04 to 22.04: libvirt-clients libvirt-daemon libvirt-daemon-config-network libvirt-daemon-config-nwfilter libvirt-daemon-driver-qemu libvirt-daemon-system libvirt-daemon-system-systemd libvirt0
[12:16] <Mmike> They're listed under `Remove (was auto installed)` 'section'.
[12:17] <Mmike> (There is around 200 of them, but I only noticed issues on machines where I have libvirt installed while those were fresh xenial boxes)
[15:29] <teward> bryceh: looks like Robie confirmed a flaw in your written config at https://discourse.ubuntu.com/t/how-to-configure-nginx/39680 which I have independently looked at and concur is an issue.  Do you want to edit your documentation or do you mind me leveraging the moderator button I have on Discourse that I almost never use to edit the documentation?
[17:06] <teward> rbasak: not sure if bryceh is just timezone-shifted out of my timezone and such for us to handle, but if nobody else argues i'm going to handle that discourse triage you made and Just Edit It to remove the discrepancy.  A long while ago while admins were figuring out Discourse perms they elevated me (and a few others) to mod for us to have edit access in Community Council.  Discourse perms are weird.
[17:08] <rbasak> teward: if you're confident in your amendment then that'd be great - please just make it!
[17:09] <teward> rbasak: 100% confident, straight from the NGINX config handbook :P
[17:09] <rbasak> Maybe it's worth a comment to say it's been addressed - I'm not sure we have an established process for wrapping up once done.
[17:09] <teward> rbasak: ye i made a note on the thread that i already handed it to proper people
[17:09] <teward> so when i make the revision i'll make a note
[17:09] <teward> just want a second "ACK" on the JFDI call before JFDI :P
[17:11] <rbasak> Thank you!
[17:24] <teward> rbasak: bryceh: revised the post to have an http-to-https redirect block as well as the SSL.
[17:25] <teward> so the Discourse post triage is handled now.
[17:25] <teward> ... time for lunch
[18:23] <bryceh> thanks teward