[12:16] <ahasenack> hm, I have another jammy laptop, with an older installation (right after jammy came out), same problem
[12:20] <ahasenack> sarnold: https://github.com/lxc/lxd/issues/10537
[12:20] <ubottu> Issue 10537 in lxc/lxd "chgrp fails in lvm storage containers on Ubuntu 22.04" [Closed]
[12:21] <ahasenack> sdeziel: ravage: ^^^
[12:22] <lotuspsychje> think ravage had an lvm/zfs bug on jammy reported too
[12:23] <lotuspsychje> bug #1970066
[12:23] <ubottu> Bug 1970066 in snapd (Ubuntu Jammy) "(Encrypted) ZFS breaks 22.04 installation" [Critical, Confirmed] https://launchpad.net/bugs/1970066
[12:23] <lotuspsychje> (if that can help ahasenack )
[12:25] <ravage> im not sure if that is related. i dont have an older install anymore
[12:31] <ahasenack> simplest workaround at the moment is to run lxd with LXD_IDMAPPED_MOUNTS_DISABLE=1
[12:31] <ahasenack> (the daemon)
[13:31] <ahasenack> I noticed that frr does not restart rsyslog after frr is installed. It creates a /etc/rsyslog.d/45-frr.conf file that will only be picked up if rsyslog is restarted
[13:31] <ahasenack> I wonder if there shouldn't be a trigger for this, in all packages that install rsyslog config snippets
[13:35] <ahasenack> in haproxy's README.Debian: "you must restart rsyslog after installing this package"
[13:45] <sdeziel> ahasenack: I'm using ZFS which lacks idmap so that's why I couldn't reproduce your bug, thanks!
[13:46] <ahasenack> curious thing is that I'm using btrfs for the lxd storage pool, and that is on top of an LVM LV
[13:46] <ahasenack> I guess the FS doesn't matter
[13:46] <ahasenack> but before I had zfs on top of an LV too
[13:47] <ahasenack> sdeziel: what's your zfs storage, a file?
[13:47] <ahasenack> or partitions
[13:47] <sdeziel> ahasenack: partitions
[13:47] <ahasenack> not logical volumes?
[13:47] <sdeziel> no
[13:47] <ahasenack> ok
[14:01] <ahasenack> re rsyslog trigger, there is one already, nothing to see here
[17:54] <sarnold> ahasenack: aha! nice find. even nicer that a fix is in progress :)
[17:54] <ahasenack> yeah, but in the kernel, that will take a while
[17:56] <ahasenack> at least I wasn't crazy :)
[18:22] <sarnold> yeah, bummer about the fix being in the kernel ;( that adds so much latency. apparently the panics that have been taking out boxes left and right had a patch publically available back in april, and we'll probably get it in another two-ish weeks..
=== scortal is now known as scortal1
[18:42] <scortal1> hi
[19:28] <scortal1> hi
=== scortal is now known as scortal11
[20:20] <scortal11> hi
[20:20] <sarnold> scortal11: cmon, you've been asked to not do this..
[20:21] <sarnold> once is polite, odd, but whatever. three times is just weird
[20:21] <scortal11> 'offtopic