[03:16] <blahdeblah> Repeating earlier question now that my IRC bouncer isn't bouncing up & down:
[03:16] <blahdeblah> Anyone able to comment on what's current best/easiest practice when setting up a VM host with VMs bridged onto multiple VLANs?
[03:16] <blahdeblah> I know the Linux kernel bridging code grew VLAN filtering in semi-recent kernels - is there an easy configuration method for it exposed in userland?
[08:41] <crohmann> coreycb: jamespage: May I kindly nag you about triggering a package update for UCA for neutron-vpnaas? - see https://bugs.launchpad.net/cloud-archive/+bug/1995861
[08:41] -ubottu:#ubuntu-server- Launchpad bug 1995861 in Ubuntu Cloud Archive "[SRU] neutron-vpnaas packages need updates as they are missing two critical patches" [Undecided, New]
[14:19] <sdeziel> blahdeblah: apparently systemd-networkd supports it but netplan doesn't yet (it's be asked for via https://bugs.launchpad.net/netplan/+bug/1793128 though)
[14:19] -ubottu:#ubuntu-server- Launchpad bug 1793128 in netplan "support vlan-aware bridges (vlan-filtering)" [Wishlist, Triaged]
[17:19] <foo> teward: when you say "i would NOT have things live in your home directory" (and sdeziel agrees) - you're referring to web code? Or socket only? 
[17:19] <foo> sdeziel: thanks, didn't think about strace
[17:29] <tds> blahdeblah: ifupdown2 has good support for it
[17:34] <tds> ah, though looks like ifupdown2 is missing from recent ubuntu releases :(
[18:43] <Kehet> hi, is it possible to dump crashlog from ubuntu-server installer to usb stick I use for booting? 
[18:43] <Kehet> or some other way to transfer logs 
[18:48] <teward> foo: technically speaking *both* because if there's a leak in your code it can traverse into your home directory and steal information.  but your socket should probably sit in a runtime directory space like was suggested by me and sdeziel like we said yesterday/overnight
[18:48] <sarnold> Kehet: when I was troubleshooting some booting problems a few years ago, I did a fair amount of copying of those crash dumps onto a different USB stick
[18:49] <sarnold> Kehet: depending upon how far you made it into the installer, you could configure the networking up, and use nc to move things around that way
[18:50] <Kehet> ohh I can drop to shell ofc, I'm dumb
[19:04] <blahdeblah> sdeziel, tds: thanks
[19:34] <foo> teward: thank you, that makes sense. I didn't consider this.
[19:34] <foo> teward: appreciate you sharing, thank you
[19:35] <teward> foo: you're welcome, but at the very least consider a runtimedirectory via systemd in /run/ for the socket at the least so www-data is guaranteed to access it proper
[19:36] <teward> i never trust www-data to work with sub/alt groups proper in daemons for the reasons you ran into - sometimes being in a secondary group doesnt work ;)
[19:38] <adac> effendy[m], I tried to remove the mount in the kubernetes env. That alone didn"t resolve the slow htop/iotop. However I also upgraded the packages on my machine and then docker got restarted. After that htop and iotop run fast again.
[19:39] <adac> So i suspect it you probably were still right. It might about this mount that stuff was slow on the machine. For whatever reason
[19:45] <effendy[m]> adac: interesting, yeah. Maybe there was something hanging there, even on the remote share
[20:03] <foo> teward: I like that.