[03:25] <cc111>  Hi. Does anyone have an example of an autoinstall yaml that includes a disk path: "*sdb" to choose an installl disk by chance? The installer keeps selecting a different disk.
[04:55] <foo> Nginx question (I realize there is an nginx channel, too): Getting this error:  [error] 23619#23619: *136 client intended to send too large body: 761038926 bytes ... am I missing something, shouldn't client_max_body_size 1500M solve that?
[04:55] <foo> Might be that nginx is the best place for that, in which case, please disregard :) stumped on this.
[11:37] <Walex> "Context: http, server, location", you may have to specify it in some or all contexts, depending on your configuration.
[13:51] <cc111> Is there an example of the autoinstall that includes matching a disk like path: "*sdb" to choose an install disk? The installer keeps selecting a different disk.
[15:03] <RoyK> cc111: what makes it even more interesting, is that device naming isn't very stable in linux ;)
[15:57] <cc111> I am fine with matching the pcipath like in /dev/disk/by-path/
[16:00] <RoyK> cc111: that's probably better, but then, I don't know much about autoinstall, so please excuse me there - others may know more
[16:08] <cc111> is there any way  i could tell why it's ignoring the parameters? and if i could restart the installer with a new user-data file without rebooting?
[16:19] <RoyK> cc111: personally, I don't know, but again, others may know
[21:22] <blackboxsw> athos: for ubuntu server future support of ansible: are we looking at whether ansible-core (minimal package) is a viable install target as opposed? cloud-init may opt to install the minimal ansible-core instead of ansible, but I notice ansible-core deb can't install on kinetic. 
[21:23] <blackboxsw> filing a bug against ansible-core for reference
[21:24] <blackboxsw> maybe it's this bug https://bugs.launchpad.net/ubuntu/+source/ansible-core/+bug/1969917 digging in
[21:40] <blackboxsw> here's the full symptom athos https://bugs.launchpad.net/ubuntu/+source/ansible-core/+bug/1969917/comments/5 broken symlink due to deb package putting release.py in unversioned /usr/lib/python3 vs versioned /usr/lib/python3.10 and py3compile not tying the links together properly I think
[23:38] <blahdeblah> blackboxsw: Out of curiosity, how did we get into the situation where ansible & ansible-core are different versions?  I was just building myself a dev environment container for it and realised that the version numbering is a bit weird.
[23:39] <blackboxsw> wow blahdeblah that is strange and from the looks of ansible:debian/control  it doesn't generate the ansible-core module as I would have expected.
[23:40] <blackboxsw> I didn't realize they were generated from separate source packages until just now.
[23:42] <blahdeblah> Was there perhaps an upstream Debian package maintainer dispute?
[23:42] <blackboxsw> seems like we'd want to make sure we are syncing the right compatible versions. I see a couple breaks/replaces in control files that need sorting I think
[23:42] <blahdeblah> There's also the ansible PPA, which seems to more closely follow upstream ansible.
[23:42] <blackboxsw> ansible-core: Breaks: ansible (<< 4.6.0-1~), ansible-base                                     
[23:43] <blackboxsw> Replaces: ansible (<< 4.6.0-1~), ansible-base                                   
[23:43] <blackboxsw> so yeah a bit muddy to sort I think
[23:43]  * blackboxsw has to run for dinner. will check in later
[23:43] <blackboxsw> ttfn