CentOS 5.6 and Auto Shutdown when UPS Power is On

[amithad]

I followed all the steps in this documentation with me local parameters and when I try upscmd command with the following parameters is gives the error

[root@proxy /]# upscmd -u root -p xxxxxx MGE test.battery.start.quick
Unexpected response from upsd: ERR ACCESS-DENIED

my NUT setup supports only the following UPS commands

[root@proxy /]# upscmd -l MGE@localhost
Instant commands supported on UPS [MGE]:

beeper.disable - Disable the UPS beeper
beeper.enable - Enable the UPS beeper
beeper.mute - Temporarily mute the UPS beeper
beeper.off - Obsolete (use beeper.disable or beeper.mute)
beeper.on - Obsolete (use beeper.enable)
load.off - Turn off the load immediately
load.off.delay - Turn off the load with a delay (seconds)
load.on - Turn on the load immediately
load.on.delay - Turn on the load with a delay (seconds)
outlet.1.load.off - Turn off the load on outlet 1 immediately
outlet.1.load.on - Turn on the load on outlet 1 immediately
outlet.2.load.off - Turn off the load on outlet 2 immediately
outlet.2.load.on - Turn on the load on outlet 2 immediately
shutdown.return - Turn off the load and return when power is back
shutdown.stayoff - Turn off the load and remain off
shutdown.stop - Stop a shutdown in progress
test.battery.start.deep - Start a deep battery test
test.battery.start.quick - Start a quick battery test
test.battery.stop - Stop the battery test

and the command /etc/init.d/ups start has the following result

[root@proxy /]# /etc/init.d/ups start
Starting UPS driver controller:                            [  OK  ]
Starting upsd:                                             [  OK  ]
Starting UPS monitor (master):                             [  OK  ]

[atreyu]

I followed all the steps in this documentation with me local parameters and when I try upscmd command with the following parameters is gives the error

[root@proxy /]# upscmd -u root -p xxxxxx MGE test.battery.start.quick
Unexpected response from upsd: ERR ACCESS-DENIED

Unfortunately, I don't have a UPS w/which to test, so I can't follow along with you. Looking at the man page, I can only suggest taking a look at your upsd.users file setup:

DIAGNOSTICS
upscmd won't work unless you provide a valid username and password. If you get "access denied" errors, make sure that your upsd.users(5) has an entry for you, and that the username you are using has permissions to SET variables.

[amithad]

Thanks Atreyu,

I edited the upsd.users file so that it looks like below

[root]
        password = xxxxxx
        upsmon = master
        instcmds = test.battery.start.quick
        instcmds = test.batery.stop

and after that command worked

[root@proxy ~]# upscmd -u root -p xxxxxx 3000 test.battery.start.quick
OK

Could I know how long I have to wait to shutdown the PC with these default settings when it runs from UPS battery?

Thanks

[atreyu]

Do you mean how long does your UPS battery last? I don't know that, it should be in your UPS specs, but 30 minutes is common.

If you mean how soon should you initiate a shutdown on your server when the UPS software detects a power outage, i'd say immediately. That is: immediately start to shut things (daemons, processes, warn users about the server going down w/a wall message, etc.) in a graceful fashion. You want your filesystems to have no I/O happening, if the battery kicks it and you are not done bringing down the system.

I'm not really sure how the ups software works w/the system to shut it down, but i imagine you define it somewhere. I would just make sure that things like Oracle and Samba are shut down first, and then a normal "shutdown -h now" (or similar, lots of options to shutdown) can be done.

[amithad]

Dear Atreyu

Yes the second option ie the server shutdown process, not the UPS battery dead time. If I can configure NUT to start the process after 5 minutes; when the server runs from the UPS battery that would be fine because some times the main power is available within fine minute time during frequent power interruptions. I suppose it should be configured in

MONITOR MGE@localhost 1 root xxxx master
MINSUPPLIES 1
SHUTDOWNCMD "/sbin/shutdown -h +0"
POLLFREQ 5
POLLFREQALERT 5
HOSTSYNC 15
DEADTIME 15
POWERDOWNFLAG /etc/killpower
RBWARNTIME 43200
NOCOMMWARNTIME 300
FINALDELAY 5

somewhere in the upsmon.conf file when executing vi /etc/ups/upsmon.conf command

[amithad]

Hi I tested the UPS by Disconnecting the main A/C current supply to the UPS

when that was being done I got the following message at the Linuxbox terminal

UPS 3000@localhost on battery

Broadcast message from nut@proxy.server (Wed Oct 24 17:21:15 2012):

UPS 3000@localhost on line power

But I waited for three minutes my CentOS server didn't shutdown

[atreyu]

But I waited for three minutes my CentOS server didn't shutdown

I thought you configured it for 5 minutes? Anyway, is there anything in /var/log/messages? If not, does the ups daemon have its own log that you can check?

[amithad]

Yes Atreyu,

those are...

Oct 24 17:14:54 proxy usbhid-ups[2741]: Startup successful
Oct 24 17:14:54 proxy upsd[2743]: listening on 127.0.0.1 port 3493
Oct 24 17:14:54 proxy upsd[2743]: Connected to UPS [3000]: usbhid-ups-3000
Oct 24 17:14:54 proxy upsd[2743]: Unknown upsmon type =
Oct 24 17:14:54 proxy upsd[2744]: Startup successful
Oct 24 17:14:54 proxy upsmon[2747]: Startup successful
Oct 24 17:14:54 proxy upsd[2744]: User root@127.0.0.1 logged into UPS [3000]
Oct 24 17:18:34 proxy upsmon[2748]: UPS 3000@localhost on battery
Oct 24 17:18:34 proxy kernel: type=1400 audit(1351079314.997:4): avc:  denied  { search } for  pid=2757 comm="sh" name="home" dev=sda3 ino=9175041 scontext=unconfined_u:system_r:nut_upsmon_t:s0 tcontext=system_u:object_r:home_root_t:s0 tclass=dir
Oct 24 17:18:35 proxy wall[2757]: wall: user nut broadcasted 1 lines (31 chars)
Oct 24 17:21:15 proxy upsmon[2748]: UPS 3000@localhost on line power
Oct 24 17:21:15 proxy kernel: type=1400 audit(1351079475.002:5): avc:  denied  { search } for  pid=2767 comm="sh" name="home" dev=sda3 ino=9175041 scontext=unconfined_u:system_r:nut_upsmon_t:s0 tcontext=system_u:object_r:home_root_t:s0 tclass=dir
Oct 24 17:21:15 proxy wall[2767]: wall: user nut broadcasted 1 lines (34 chars)
Oct 24 17:23:41 proxy dbus: [system] Rejected send message, 1 matched rules; type="method_call", sender=":1.64" (uid=0 pid=2804 comm="/usr/bin/python) interface="org.freedesktop.NetworkManager" member="GetDevices" error name="(unset)" requested_reply=0 destination=":1.46" (uid=500 pid=2433 comm="nm-applet))
Oct 24 17:24:05 proxy kernel: ip_tables: (C) 2000-2006 Netfilter Core Team
Oct 24 17:24:05 proxy kernel: nf_conntrack version 0.5.0 (16384 buckets, 65536 max)
Oct 24 17:24:05 proxy kernel: ip6_tables: (C) 2000-2006 Netfilter Core Team
Oct 24 17:24:22 proxy init: tty (/dev/tty2) main process (2147) killed by TERM signal
Oct 24 17:24:22 proxy init: tty (/dev/tty3) main process (2149) killed by TERM signal
Oct 24 17:24:22 proxy init: tty (/dev/tty4) main process (2151) killed by TERM signal
Oct 24 17:24:22 proxy init: tty (/dev/tty5) main process (2153) killed by TERM signal
Oct 24 17:24:22 proxy init: tty (/dev/tty6) main process (2155) killed by TERM signal
Oct 24 17:24:22 proxy avahi-daemon[1757]: Got SIGTERM, quitting.
Oct 24 17:24:22 proxy avahi-daemon[1757]: Leaving mDNS multicast group on interface eth0.IPv4 with address 192.168.101.31.
Oct 24 17:24:23 proxy abrtd: Got signal 15, exiting
Oct 24 17:24:26 proxy acpid: exiting
Oct 24 17:24:27 proxy upsmon[2748]: Signal 15: exiting
Oct 24 17:24:27 proxy upsd[2744]: User root@127.0.0.1 logged out from UPS [3000]
Oct 24 17:24:27 proxy upsmon[2747]: upsmon parent: read
Oct 24 17:24:27 proxy upsd[2744]: mainloop: Interrupted system call
Oct 24 17:24:27 proxy upsd[2744]: Signal 15: exiting
Oct 24 17:24:27 proxy NetworkManager[1745]: <info> caught signal 15, shutting down normally.
Oct 24 17:24:27 proxy NetworkManager[1745]: <info> exiting (success)

[atreyu]

it looks like things are starting to shut down...is there anything in dmesg?

it looks like SELinux is enabled. Could that be interfering? You can check if enabled with:

getenforce

and temporarily disable it with:

setenforce 0

[amithad]

Yes SELinux is in enforcing mode I temporarily disabled it. Will give you the resuls of the setting.

Thanks