[CentOS] shutdown or poweroff?

Sun Feb 4 17:38:15 UTC 2018
Joseph L. Casale <jcasale at activenetwerx.com>

-----Original Message-----
From: CentOS [mailto:centos-bounces at centos.org] On Behalf Of Walter H.
Sent: Sunday, February 4, 2018 9:03 AM
To: centos at centos.org
Subject: [CentOS] shutdown or poweroff?

> just a simple question, my router has CentOS 6 with the apcupsd running,
> in the log of
> apcupsd I see this:
> 
> 2018-02-01 19:05:54 +0100  apcupsd 3.14.12 (29 March 2014) redhat
> startup succeeded
> 2018-02-04 15:52:43 +0100  Power failure.
> 2018-02-04 15:52:49 +0100  Running on UPS batteries.
> 2018-02-04 15:53:00 +0100  Reached remaining time percentage limit on
> batteries.
> 2018-02-04 15:53:00 +0100  Initiating system shutdown!
> 2018-02-04 15:53:00 +0100  User logins prohibited
> 2018-02-04 15:53:37 +0100  apcupsd exiting, signal 15
> 2018-02-04 15:53:37 +0100  apcupsd shutdown succeeded
> 
> does this mean the shutdown was successfull?
> is there other log where I can verify this: because shutting down squid
> takes almost a minute or so ...

Things changed a bit from the sysv implementation (see http://www.apcupsd.org/manual/manual.html#system-shutdown-test)
when systemd came into the picture as many naïve or hack approaches
were no longer necessary. It looks like the current packages drop the
needed scripts in /lib/systemd/system-shutdown (see https://www.freedesktop.org/software/systemd/man/systemd-halt.service.html)
So now with the state file and that script, your UPS (if setup and possible)
will be signaled to interrupt power at a safe point once the daemon
initiates shutdown.

Your next startup log will indicate if a dirty mount was discovered.

You should also check /etc/apcupsd/apcupsd.conf for the behavior
and time values, there aren't many options and the file is short.