I'm not going to be too much use on this one...
My main machine has gone through too many upgrades and even the latest wasn't clean; an exim4 config problem seemed to leave the upgrade mid-session due to an old customisation:
# MC : remove log warnings 'Warning: purging the environment'
# 'Suggested action: use keep_environment'
# see https://forum.bytemark.co.uk/t/exim-change-logwatch-and-keep-environment-cve-2016-1531/2372/8
2018-08-07 23:17:23 Exim configuration error in line 333 of /etc/exim4/exim4.conf.tmp:
"keep_environment" option set for the second time
Makefile:34: recipe for target 'test' failed
make: *** [test] Error 1
dpkg: error processing package symbiosis-email (--configure):
subprocess installed post-installation script returned error exit status 2
dpkg: dependency problems prevent configuration of symbiosis-xmpp:
symbiosis-xmpp depends on symbiosis-email; however:
Package symbiosis-email is not configured yet.
dpkg: error processing package symbiosis-xmpp (--configure):
dependency problems - leaving unconfigured
While I was looking at what had happened and attempting to install the roundcube php5 dependency the machine rebooted (I think unattended-updates has changed and reboots [with logged in users]) but this seemed to help. After that, all seemed to be well except for a couple of things. I was getting two
motd messages at login - one for jessie at time of upgrade (picked up from
/run, I think) and one for the current stretch session. Secondly,
unattended-upgrades showed no action overnight. It was fine run manually but that's not how it's supposed to work.
Meanwhile, I gave php7 a go: only one site required knocking into shape so I was pleased to retire php5 an hour or so after upgrade. Performance looks much improved. After a day or so I thought I'd see if another reboot would kick everything into shape. It did; I can't see any issues now.
So, drifting back on topic, I've just added a new site and
sudo symbiosis-httpd-configure --verbose just does its stuff. However,
public/logs hasn't been created. I've only tested with an unregistered domain so I don't know if it'll be created on demand (probably not going by the previously quoted jessie low memory httpd logger thread). So, it may or may not be a bug (I told you I wouldn't be much use ).
If you can't sort it out and don't mind a workaround, then adding
/etc/symbiosis/skel.d/public/logs may be worth a shot. (I'm considering adding a lazy site-wide
Looks like I was wrong about the
unattended-upgrades reboot -- it still defaults to
( but this is a new default...