On Sat, Jul 10, 2021, at 11:30 AM, Stuart Henderson wrote: > On 2021-07-10, Peter Nicolai Mathias Hansteen <pe...@bsdly.net> wrote: > > For whatever reason your pf.conf did not parse to a valid config, so rc’s > > own default rules were kept in place. > > Yep. dmesg -s might give a clue.
Thank you both, I suspected it might be default startup rules but I didn't know they were in /etc/rc. Since I've rebooted again since the crash I don't think dmesg -s is helpful now, but there are other indicators in /var/log/messages that something wasn't right: ... Jul 9 17:13:58 ********** ntpd[24315]: constraints configured but none available Jul 9 17:14:14 ********** ntpd[24954]: no reply received in time, skipping initial time setting Jul 9 17:14:14 ********** savecore: no core dump Jul 9 17:14:24 ********** reorder_kernel: kernel relinking done Jul 9 17:28:58 ********** ntpd[24315]: constraints configured but none available Jul 9 17:43:58 ********** ntpd[24315]: constraints configured but none available Jul 9 17:58:58 ********** ntpd[24315]: constraints configured but none available ... For now I'll chalk it up to an issue with the KVM host (I have a ticket open with the hosting provider asking about the event, since both my VMs on that node crashed at the same time). If it happens again I will try dmesg -s. Allan