[00:00] no, so it doesn't use /dev/sdc etc [00:00] those names aren't persistant [00:00] they can change across reboots [00:04] that must be its default, correct? [00:06] there is no default -- zfs uses whatever names you used in the zfs create command [00:07] if you used the short names, that's what it'll use [00:07] and if they change, you'll have a Bad Time [00:08] ya that must be whats going on because i added drives to the system. i just figured it would find its zfs drives and do its thing 🤷‍♂️ or at least find the drives and use the new device names, if that is the situation. [00:12] because systems may have several thousand drives, it's not the default behaviour to go looking for drives, that'd take forever [00:12] that's the reason for the cache file.. [00:12] but if it's busted, it'd be nice if it tried to help out a bit, heh [04:48] sarnold: you still around? [04:52] i exported the pool and did zpool import -d /dev/disk/by-id -a. seems to have worked. thanks again! [05:14] Intelo: even i915 has working vGPU support (less powerful as a GPU, but more compatible and easy to use actually), I tried it on my Laptop once [12:08] jamespage: could you take a look at https://code.launchpad.net/~chris.macnaughton/ubuntu/+source/openvswitch/+git/openvswitch/+merge/387852 ? [13:03] auxin: #zfsonlinux may be a more useful venue for you to get help with recovering your data. === cpaelzer_ is now known as cpaelzer === kedar is now known as kedar_apte [18:03] kinghat: is everything all sorted out with your pool? :) [18:04] sarnold: it seems to be but i havent pulled the power and all that to really test it. i will in an hour or so. [18:07] kinghat: heh, yeah, I know the reluctance to actually *test* those kinds of things.. [18:49] <|\n> hello, since 20.04 i need some `ebtables -t broute` equivalent, is there any? [18:58] |\n: what happens when you use it? it's still documented in the ebtables manpage on my 20.04 laptop anyway [19:01] <|\n> sarnold, in my case it drops certain frames from certain interface [19:03] <|\n> well it was doing so at 18.04 [19:05] <|\n> it was pretty sweet and cozy to look into if_ether.h and drop something heh [19:28] is there a way to prevent a package upgrade from running the .postinst script? [19:30] <|\n> exclude it from package / repackage maybe [19:30] sdeziel: unpackage it by hand? [19:30] <|\n> though doesn't sound like a good idea [19:30] sarnold: it's what https://askubuntu.com/questions/482928/ignore-apt-get-postinstall-scripts-automatically suggests too :( [19:31] not what I was looking for but thanks anyway ;) [19:32] sdeziel: maybe try the touch postinst thing, then set immutable? [19:32] sdeziel: probaby dpkg will blow up and leave you with a big mess. but maybe it will ignore errors? [19:32] sarnold: that's what I tried at first but I'm in a container where the underlying FS doesn't let me [19:33] sdeziel: if the package in question shld do different things in a container, that's probably worth a bug report [19:33] in lxd containers backed by btrfs or zfs, I get a permission denied on the chattr +i [19:34] yeah, that part makes sense [19:34] yeah, I'll probably end up reporting this as a bug but it's not because it's in a container [19:35] it's too hard to run MySQL read-only replicas on Ubuntu IMHO [19:36] ugh yes, that's gonna be annoying [19:36] that script deserves a MYSQL_DO_NOTHING [19:37] I'd like the postinst to not explode when in read_only and/or super_read_only mode [19:40] sweet, the bug affects MySQL 5.7 only :) [19:59] closing the loop: LP: #1889472 [19:59] Launchpad bug 1889472 in mysql-5.7 (Ubuntu) "mysql-server-5.7 postinst fails when in read-only mode" [Undecided,New] https://launchpad.net/bugs/1889472 [20:03] sdeziel: beautiful [21:32] sarnold: looks like it survived the test [21:32] kinghat: woot! [21:32] thanks for checking back 🙏 === jancoow1 is now known as jancoow [23:26] Howdy. I am currently attempting to install Ubuntu 20.04 on a USB Stick in a HP server, it works until it gets to 'Installing Kernel ' and just sits forever at 'Unpacking linux-firmware'. Any ideas on what may be going wrong?