Nut upsuser - Aug 2020 - Synology NAS is shutting down Ubuntu servers after very brief power outage (fwd)

> Since it's NUT in the NAS which is deciding to order the shutdown, it
would be
> useful to see upsmon.conf, upssched.conf and upssched-cmd (or whatever
Synology
> use if anything) from the NAS to see what criteria they use.
upsmon.conf (on Synology):
RUN_AS_USER root
MONITOR ups at localhost 1 monuser secret master
MINSUPPLIES 1
SHUTDOWNCMD ""
NOTIFYCMD /usr/sbin/upssched
POLLFREQ 5
POLLFREQALERT 5
HOSTSYNC 15
DEADTIME 15
POWERDOWNFLAG /etc/killpower
NOTIFYFLAG ONLINE EXEC
NOTIFYFLAG ONBATT EXEC
NOTIFYFLAG LOWBATT EXEC
NOTIFYFLAG NOCOMM EXEC
NOTIFYFLAG COMMBAD IGNORE
NOTIFYFLAG COMMOK IGNORE
NOTIFYFLAG SHUTDOWN IGNORE
NOTIFYFLAG FSD EXEC
NOTIFYFLAG NOPARENT SYSLOG
RBWARNTIME 43200
NOCOMMWARNTIME 300
FINALDELAY 5

upssched.conf (on Synology):
CMDSCRIPT /usr/syno/bin/synoups
PIPEFN /var/run/upssched.pipe
LOCKFN /var/run/upssched.lock
AT ONLINE * EXECUTE online
AT ONLINE * CANCEL-TIMER fsd
AT LOWBATT * EXECUTE lowbatt
AT NOCOMM * EXECUTE nocomm
AT FSD * EXECUTE fsd
AT ONBATT * EXECUTE onbatt

upssched-cmd (on Synology):
I don’t see this file.  Here are all the files in the ups folder:
nutscan-usb.h  ups.conf  upsd.conf  upsd.users  upsmon.conf  upssched.conf
> What is "Safe mode"?  Is it complete power down?, or some sort of
hibernation?
> If it's not a complete power down, how is the hibernation powered?
	"When the Synology NAS enters Safe Mode, it stops all services and
unmounts volumes in order to prevent data loss and shut down (halt for EDS14)
safely when the UPS device runs out of power. By default, the system enters Safe
Mode when the UPS device starts running low on power. You can also specify the
amount of time before the Synology NAS enters Safe Mode when power failure
occurs. However, if the UPS device reaches low battery before the specified
time, the system enters Safe Mode immediately.

	In situations where the Synology NAS shuts down during Safe Mode, it will
automatically turn on when power is restored if you have enabled the Restart
automatically after a power failure option (located at Control Panel >
Hardware & Power > General)."

I gather this puts it in a state where it is ready for imminent power loss and
won’t damage or lose any data.  I have enabled “Shutdown UPS when the system
enter Safe Mode” and I have also enabled “Restart automatically after a power
failure” as noted above.  I have always had these two options enabled.

Yesterday I recently disabled HDD Advanced Hibernation option as I thought that
may interfere with NUT on my servers being able to communicate with the NUT
server on the NAS, hence the immediate shutdown.  Now they will spin-up on their
own, but it takes 5-10 seconds.  My thought was if they can’t communicate
initially, they assume the server is dead and shut down.  Would that make sense
at all?

HDD Hibernation essentially spins down the hard drives when there has been no
activity.  I had it set at one hour; I have since disabled that setting.  I have
not ever set the Synology itself to go to sleep or to auto-shutdown.

From the Synology support site:
	You can specify the period of time hard disks remain idle before entering HDD
hibernation mode. During HDD hibernation, hard disks stop spinning, reducing
power consumption and extending the lifespan of hard disks. You can specify
different periods of time for internal hard disks and external eSATA/USB hard
disks.

	Phase 1: HDDs power down after a period of inactivity (idle time). You may go
to DSM > Control Panel > Hardware & Power > HDD Hibernation to
configure the length of inactivity for the HDDs to enter hibernation.

	Phase 2: Advanced Hibernation/Deep Sleep will be implemented for the HDDs to
further reduce power consumption. You may go to DSM > Control Panel >
Hardware & Power > HDD Hibernation to enable advanced HDD hibernation.
> You should remove line 1 : RUN_AS_USER nut
Ok, I will comment that back out.  That won’t interfere with anything?  What
user would it run as?  Will it have access to upsmon.conf?


Thanks,

Todd

--
Todd Benivegna // todd at benivegna.com
On Aug 9, 2020, 3:07 AM -0400, Roger Price <roger at rogerprice.org>,
wrote:
> On Fri, 7 Aug 2020, Todd Benivegna wrote:
>
> Hello Todd, sorry for the delay replying, I was away for a while. It's
ok to
> post configuration files in this list if blank lines and comments are
removed.
>
> > On the Synology (I didn’t edit any of these files):
> > ups.conf:  https://hastebin.com/dedereqizi.shell
>
> pollinterval = 5
> [ups]
> driver = usbhid-ups
> port = auto
>
> > upsd.conf:  https://hastebin.com/pupeseweda.css
>
> LISTEN 192.168.1.70
> LISTEN 2601:cf:8200:43d0:211:32ff:fe63:60d7
> LISTEN fe80::211:32ff:fe63:60d7
> LISTEN 127.0.0.1
> LISTEN ::1
>
> > upsd.users:  https://hastebin.com/ocenamecex.cs
>
> [monuser]
> password = secret
> upsmon master
>
> Since it's NUT in the NAS which is deciding to order the shutdown, it
would be
> useful to see upsmon.conf, upssched.conf and upssched-cmd (or whatever
Synology
> use if anything) from the NAS to see what criteria they use.
>
> Roger
> _______________________________________________
> Nut-upsuser mailing list
> Nut-upsuser at alioth-lists.debian.net
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/nut-upsuser
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://alioth-lists.debian.net/pipermail/nut-upsuser/attachments/20200809/07ce9cfc/attachment.html>

On Aug 9, 2020, at 3:18 PM, Todd Benivegna wrote:
> 
> Now they will spin-up on their own, but it takes 5-10 seconds.  My thought
was if they can’t communicate initially, they assume the server is dead and shut
down.  Would that make sense at all?
Sorry to jump in the middle here, but this can occur in certain cases, when
upsmon sees a server go dead *after seeing it go on battery*. See
"DEADTIME" description:
https://networkupstools.org/docs/man/upsmon.conf.html (defaults to 15 seconds)

On Sun, 9 Aug 2020, Todd Benivegna wrote:
> upssched.conf (on Synology):
> CMDSCRIPT /usr/syno/bin/synoups
> 
> upssched-cmd (on Synology):
> I don’t see this file.  
The CMDSCRIPT declaration says that in a Synology box the file that NUT calls 
upssched-cmd is called /usr/syno/bin/synoups.  Could we see this file?  Thanks.
>       What is "Safe mode"?  Is it complete power down?, or some
sort of hibernation? 
>       If it's not a complete power down, how is the hibernation
powered?
>
>       "When the Synology NAS enters Safe Mode, it stops all services
and unmounts volumes in order to prevent data loss and shut down (halt for
>       EDS14) safely when the UPS device runs out of power. By default, the
system enters Safe Mode when the UPS device starts running low on power.
>       You can also specify the amount of time before the Synology NAS
enters Safe Mode when power failure occurs. However, if the UPS device reaches
>       low battery before the specified time, the system enters Safe Mode
immediately.
>
>       In situations where the Synology NAS shuts down during Safe Mode, it
will automatically turn on when power is restored if you have enabled the
>       Restart automatically after a power failure option (located at
Control Panel > Hardware & Power > General)."
This reads like typical sales literature.  I admit to being none the wiser even 
after reading it several times.
> I gather this puts it in a state where it is ready for imminent power loss
and
> won’t damage or lose any data.  
How long can this last?  Who supplies the power to the NAS during "Safe
mode"?
> Ok, I will comment that back out.  That won’t interfere with
anything?  What
> user would it run as?  Will it have access to upsmon.conf?
Ubuntu very probably has user nut set up to run NUT and has the permissions set 
accordingly.  If it's possible to execute "command ps -elf | grep
nut" in a
running NAS you will see user nut in action.

Roger

Hi Roger,

Here is the /usr/syno/bin/synoups file

https://hastebin.com/sibopejuyu.bash

Thanks,

Todd

--
Todd Benivegna // todd at benivegna.com
On Aug 9, 2020, 4:49 PM -0400, Roger Price <roger at rogerprice.org>,
wrote:
> On Sun, 9 Aug 2020, Todd Benivegna wrote:
>
> > upssched.conf (on Synology):
> > CMDSCRIPT /usr/syno/bin/synoups
> >
> > upssched-cmd (on Synology):
> > I don’t see this file.
>
> The CMDSCRIPT declaration says that in a Synology box the file that NUT
calls
> upssched-cmd is called /usr/syno/bin/synoups. Could we see this file?
Thanks.
>
> > What is "Safe mode"?  Is it complete power down?, or some
sort of hibernation?
> > If it's not a complete power down, how is the hibernation powered?
> >
> > "When the Synology NAS enters Safe Mode, it stops all services
and unmounts volumes in order to prevent data loss and shut down (halt for
> > EDS14) safely when the UPS device runs out of power. By default, the
system enters Safe Mode when the UPS device starts running low on power.
> > You can also specify the amount of time before the Synology NAS enters
Safe Mode when power failure occurs. However, if the UPS device reaches
> > low battery before the specified time, the system enters Safe Mode
immediately.
> >
> > In situations where the Synology NAS shuts down during Safe Mode, it
will automatically turn on when power is restored if you have enabled the
> > Restart automatically after a power failure option (located at Control
Panel > Hardware & Power > General)."
>
> This reads like typical sales literature. I admit to being none the wiser
even
> after reading it several times.
>
> > I gather this puts it in a state where it is ready for imminent power
loss and
> > won’t damage or lose any data.
>
> How long can this last? Who supplies the power to the NAS during "Safe
mode"?
>
> > Ok, I will comment that back out.  That won’t interfere with
anything?  What
> > user would it run as?  Will it have access to upsmon.conf?
>
> Ubuntu very probably has user nut set up to run NUT and has the permissions
set
> accordingly. If it's possible to execute "command ps -elf | grep
nut" in a
> running NAS you will see user nut in action.
>
> Roger
> _______________________________________________
> Nut-upsuser mailing list
> Nut-upsuser at alioth-lists.debian.net
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/nut-upsuser
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://alioth-lists.debian.net/pipermail/nut-upsuser/attachments/20200810/30216ed3/attachment.html>