NUT package (2.8.0 and below)

uk_dave

Hi everyone, sorry my first post here is an ask for help but.. well I need some help!
I have a slightly older APC Back-UPS 1500 RS UPS connect via a USB cable and have managed to get the basics of NUT working as a server (on pfSense) to shutdown itself as well as a remote FreeNAS and Ubuntu machine.

My problem is that when power is restored, as soon as the host that talks to the UPS starts to boot up (i.e. immediately after going through POST) it triggers the UPS to go on battery (despite being plugged into the wall with AC power on) which it then stays on until the battery dies. Clearly this is not great for any machine attached that started at the same time as the firewall came up so trying to resolve this for a clean startup. If I leave the system to do it's thing it will sit there cycling; power up, trigger UPS to go on battery, continue booting, battery dies, abrupt shutdown, power up, rinse and repeat.

I've just moved the UPS from my rack to my study to do some more testing and I'm experiencing the same behaviour running the NUT master on a ubuntu test machine.. seems unlikely now it's anything specific to the pfSense iteration.

I'm hoping someone with experience configuring NUT and UPS's here might know what the issue since it seems like either a driver config option I've not set or something to do with startup / shutdown / init scrips etc that is a whole other kettle of fish to get my head around! In desperation I installed the windows utility (on a windows machine) to see if there were any behaviour options I could set on the UPS but there was nothing.

Again, apologies this isn't really a pfSense specific issue (other than ultimately I want the UPS managed by it). Alternatively if anyone could just give me the keywords to search on google for the behaviour I'm trying to stop that would be a great start!

Thanks,
Dave

Edit: Just tried apcupsd which I think works as I would expect so I guess I have a partial workaround but will need to see if I can get that to notify NUT on FreeNAS.. but that's definitely outside the scope of this forum!

dennypage

@uk_dave A few questions:

Are you using running the current version of pfSense (2.4.5)?
Are you using the current version of the nut package (2.7.4_7)?
Do you have anything in Advanced Settings? In particular, any of these?
- ignorelb
- override.battery.charge.warning
- override.battery.charge.low
- override.battery.runtime.low

uk_dave

@dennypage Thanks for replying,

Running 2.4.5-RELEASE-p1 (amd64) of pfSense and 2.7.4_7 of the NUT package.

I'm currently doing this testing on a fresh install of pfSense, all I had added to the config was:

USB Type: Local USB
UPS Name: ups
Driver: usbhid
ups.users: [remoteuser]
password = upsmon
upsmon slave

Here's my UPS: BR1500GI

I have disabled the master / slave behaviour of the UPS so all 5x sockets on the battery backup side should be in the 'Battery backup outlets with surge protection only' mode. I was assuming that NUT tells the UPS to shutdown and therefore having 5x sockets I can use rather than 3x was preferable.

Thanks,
Dave

dennypage

@uk_dave said in NUT package:

BR1500GI

Can you post output of upsrw and upsc for the device please?

uk_dave

@dennypage Output below:

[2.4.5-RELEASE][admin@pfSense.localdomain]/root: upsrw ups
[battery.charge.low]
Remaining battery level when UPS switches to LB (percent)
Type: STRING
Maximum length: 10
Value: 10

[battery.runtime.low]
Remaining battery runtime when UPS switches to LB (seconds)
Type: STRING
Maximum length: 10
Value: 120

[input.sensitivity]
Input power sensitivity
Type: STRING
Maximum length: 10
Value: medium

[input.transfer.high]
High voltage transfer point (V)
Type: STRING
Maximum length: 10
Value: 294

[input.transfer.low]
Low voltage transfer point (V)
Type: STRING
Maximum length: 10
Value: 176

[ups.delay.shutdown]
Interval to wait after shutdown with delay command (seconds)
Type: STRING
Maximum length: 10
Value: 20

[2.4.5-RELEASE][admin@pfSense.localdomain]/root: upsc ups
battery.charge: 100
battery.charge.low: 10
battery.charge.warning: 50
battery.date: 2001/09/25
battery.mfr.date: 2015/03/19
battery.runtime: 771
battery.runtime.low: 120
battery.type: PbAc
battery.voltage: 27.3
battery.voltage.nominal: 24.0
device.mfr: American Power Conversion
device.model: Back-UPS RS 1500G
device.serial: 3B1512X15061
device.type: ups
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 2
driver.parameter.port: auto
driver.parameter.synchronous: no
driver.version: 2.7.4
driver.version.data: APC HID 0.96
driver.version.internal: 0.41
input.sensitivity: medium
input.transfer.high: 294
input.transfer.low: 176
input.voltage: 245.0
input.voltage.nominal: 230
ups.beeper.status: disabled
ups.delay.shutdown: 20
ups.firmware: 878.L4 .I
ups.firmware.aux: L4
ups.load: 29
ups.mfr: American Power Conversion
ups.mfr.date: 2015/03/19
ups.model: Back-UPS RS 1500G
ups.productid: 0002
ups.realpower.nominal: 865
ups.serial: 3B1512X15061
ups.status: OL
ups.test.result: No test initiated
ups.timer.reboot: 0
ups.timer.shutdown: -1
ups.vendorid: 051d
[2.4.5-RELEASE][admin@pfSense.localdomain]/root:

Thanks again :)

dennypage

I don't see anything troubling in your upsrw or upsc output.

@uk_dave said in NUT package:

My problem is that when power is restored, as soon as the host that talks to the UPS starts to boot up (i.e. immediately after going through POST) it triggers the UPS to go on battery (despite being plugged into the wall with AC power on) which it then stays on until the battery dies.

More questions:

Are you testing by pulling mains from the UPS? Assuming yes...
When the pfSense shutdown occurs, is the UPS turning off the output?
How long are you waiting before restoring mains to the UPS?

uk_dave

@dennypage

yes, I turn the mains socket off
I don't believe it is, I've got a halogen lamp on a dimmer that can pull up to 435W so I've varied the load to full to bring the estimated run time down to 1 min remaining to force the shutdown process then ease up so it can last a few minutes and the UPS doesn't shut down until the battery dies... definitely longer than 20s
I usually wait about a minute before restoring power.

dennypage

@uk_dave said in NUT package:

I don't believe it is

Okay, this definitely is a problem. The last thing nut does is instruct the UPS to remove power after the configurable delay (20s by default).

What I would do next is to move the pfSense box to not be powered by the UPS, but still have the USB connected. Then, test the power kill command of the UPS using "upsdrvctl shutdown" or the -k option to usbhid-ups.

4o4rh

@dennypage can you help with the correct config pls.
I have a APC UPS master on TrueNAS box

I have a dumb UPS connected to pfsense and need to shutdown pfsense based on a timer triggered by the master due to the non-shared UPS.

I have looked at the NUT documentation and examples below but i can't figure out if i need to use the heartbeat config or not.

http://rogerprice.org/NUT/ConfigExamples.A5.pdf
https://wiki.ipfire.org/addons/nut/detailed

i simply want to shutdown the slave 20min after power loss, but obviously cancel the timer if the power comes back.

is the a way to simulate a power failure on the master to test the slave without actually cutting power to the master?

uk_dave

@dennypage
Hi, I have pfSense plugged into the wall and connected to the UPS via USB.

running 'upsdrvctl shutdown' does indeed turn off the outputs, 20s later shutdowns the UPS which then turns back on since AC is still active to the UPS. Watching the console of pfSense it goes through:

Initiating UPS shutdown
UPS ups on battery
UPS ups on line power

I wondered if somehow it had automagically fixed itself so after booting up again (successfully) I pulled the power to the UPS. pfSense shutdown correctly but the UPS outputs are still active and being powered. Left it for about 5 minutes but the UPS did not shutdown.

This behaviour is consistent using NUT over 3x different machines and 2x different USB cables (bought another for testing). The machines I've been testing on (each as a standalone master) are:

HP T730 thin client (currently using as a test pfSense environment)
Asus Chromebox running Ubuntu
Supermicro pfSense machine based around a X11SCL-IF

Two common denominators... the UPS and me, one of them is doing something wrong I think :)
Cheers

dennypage

@uk_dave said in NUT package:

I wondered if somehow it had automagically fixed itself so after booting up again (successfully) I pulled the power to the UPS. pfSense shutdown correctly but the UPS outputs are still active and being powered. Left it for about 5 minutes but the UPS did not shutdown.
This behaviour is consistent using NUT over 3x different machines and 2x different USB cables (bought another for testing). The machines I've been testing on (each as a standalone master) are:

HP T730 thin client (currently using as a test pfSense environment)
Asus Chromebox running Ubuntu
Supermicro pfSense machine based around a X11SCL-IF

Very weird. It seems like the the shutdown command is being ignored if the UPS is on battery. It sounds like a firmware bug, but you said that apcupsd worked correctly. I'm afraid I'm out of ideas. You might approach folk on the NUT mailing list and see what they might have to offer.

uk_dave

@dennypage

Thanks for all your help anyway. I've sent an email to the NUT mailing list so see if anyone has any ideas.
Cheers

4o4rh

@gwaitsi said in NUT package:

@dennypage can you help with the correct config pls.
I have a APC UPS master on TrueNAS box

I have a dumb UPS connected to pfsense and need to shutdown pfsense based on a timer triggered by the master due to the non-shared UPS.

I have looked at the NUT documentation and examples below but i can't figure out if i need to use the heartbeat config or not.

http://rogerprice.org/NUT/ConfigExamples.A5.pdf
https://wiki.ipfire.org/addons/nut/detailed

i simply want to shutdown the slave 20min after power loss, but obviously cancel the timer if the power comes back.

is the a way to simulate a power failure on the master to test the slave without actually cutting power to the master?

any of you guys can help me with the correct config for upssched and upsmon? Do i need the heartbeat for what i want to do?

dennypage

@gwaitsi said in NUT package:

is the a way to simulate a power failure on the master to test the slave without actually cutting power to the master?

Look for FSD in the NUT documentation.

dennypage

@gwaitsi said in NUT package:

I have a dumb UPS connected to pfsense and need to shutdown pfsense based on a timer triggered by the master due to the non-shared UPS.
...
i simply want to shutdown the slave 20min after power loss, but obviously cancel the timer if the power comes back.

A dumb UPS isn't something supported by the NUT pfSense package. In general dumb USPs are not a good idea because there is nothing that will trigger the restart of the system when power returns.

4o4rh

@dennypage an ideological argument over why it is not a good idea doesn't help solve the problem. My NAS box is 3 floors above my pfsense box (which is in the basement where there is no room for a full UPS). The dumb UPS i have, can simply replace the existing power adapter and on the PTZ i tested with, works for above 24hrs. I expect it will provide about 2-3hrs with the pfsense box, so I want to shutdown the box after about 50% capacity.

The last time i had a power failure a few weeks ago, i lost my system drive on pfsense and spent the entire weekend trying to recover and get the system back to its previous state. So while it may not be ideal to go and have to manually restart in a worse case scenario, it is better than having to rebuild the system. This year we've had about 4-5 utility outages lasting up to an hour. So in all cases this year, the device would have restore power before needing to shutdown. (although, having said that. the APC UPS only lasts about an hour before shutdown - so there may have been 1 or 2 shutdowns due to the master shutting down).

dennypage

@gwaitsi said in NUT package:

@dennypage an ideological argument over why it is not a good idea doesn't help solve the problem.

No, it won't. But then again, there isn't anything in the pfSense NUT package that will help you solve the problem as defined.

Operating within the constraints you have defined, I would just use an ssh script run from your NAS box to perform the shutdown.

Over the weekend I restarted my old QNAP NAS. That left me to solve the problem of one Synology NAS and one QNAP NAS that I wanted to have run as UPS slaves to NUT on pfsense.

Spent a minute looking at changing the default UPS name on both and quickly decided that wasn't going to be worth the effort.

Set up a "dummy-ups" in NUT thinking that would work. That didn't work the way I thought it would, adding the lines to 'additional configuration line to ups.conf'. For some reason that prepends the additional lines not append. Bad.

So, I then added the lines to 'extra arguments to driver' which appended the lines to ups.conf and that resulted in the ups.conf I wanted.

You want this in the right place, after your real UPS:

[qnapups]
driver = dummy-ups
port = ups@127.0.0.1
desc = "dummy-ups in repeater mode"

I now have a real UPS named ups and a replicate UPS named qnapups. I DO have additional lines to override low battery that come BEFORE the repeater configuration. If you want to do this in the GUI best to have a shell open and see what is going on in the config files in /usr/local/etc/nut.

My ups.conf looks like this:

[ups]
driver=usbhid-ups
port=auto
ignorelb
override.battery.charge.low = 30
[qnapups]
driver = dummy-ups
port = ups@127.0.0.1
desc = "dummy-ups in repeater mode"

As best I understand (please correct me if I'm wrong) the NUT package is all but unmaintained at this point. I wonder if the best way to get things working is to install the package, stay out of the GUI and configure things by hand in the various config files.

Gertjan

@jwj said in NUT package:

That left me to solve the problem of one Synology NAS

Like :

To have the Syno NAS connect to the NUT on pfSEnse as a salve :
Set the IP to the IP of pfSense in which the NUT package is running.
Define on pfSense the IP it has to expose it's NUT interface on :

and of course tell it what user / slave could 'login".
This user and password are somewhat hard coded into the NUT core files of the Syno (if I recall that correctly).

@Gertjan Yes. Required to get things working with a Synology NAS. I had a quick look at the Marketing blurb concerning DSM 7 beta. Supposedly more control over UPS configuration. We'll see. All this hard coded stuff is a pain in the butt!

Good post!