more of an alert to something rather than an issue to resolve.
I have probably spent most of today fixing this.
This morning did a number of package updates (17) that popped up overnight one of these was an nginx-core (I have no idea if this did it or just brought it to my attention) anyway a system reboot was required.
After that nginx was down (well we have seen that before) but it refused to restart.
Digging around (old faithful nginx -t and there are problems with all the .conf files.
every sites_enabled was broken everyone has the line fascgi_read_timeout 60;
appended at the end.
That command repeated chokes nginx so prevents the start.
Resolved by removing the extra lines in all but one conf
I have no idea where these configurations come from or what is responsible for editing.
Whatever the cause the webserver started and all websites on this box are back online.
I have no idea. I am not even sure why it is there in the first place. It is not something I would add to a conf file. All I know is that all the domains on this box were working (nginx was up and running) yesterday and I just did the package updates + reboot. Perhaps it is also the cause of the old problem. Something is putting these lines in the .conf files (sites-enabled)
I mainly use Synaptic on Debian for updates. I know the updates run a script and makes you acknowledge certain things like any config file changes. I don’t know if Webmin/Virtualmin does the same. Hmm…
Just checked by doing an update on my desktop. apt-listchange is the text that appears in the box. It closes automatically if there is nothing that needs acknowledged. Sometimes it just alerts to changes that it think you need to know before moving forward. Sometimes the option to keep your config or go with the package maintainers as I recall. But, it doesn’t move forward without a prompt or reading through the list where appropriate.