Which UPS to use with FreeNAS?

[Harsesis]

Hi,

so I'm currently buring-in my new server. I'm now looking into UPS and I'm not quite sure which will fit my needs best. In another topic I found that APC and Cyberpower are common choices. Depending on the alternatives I would like to buy a UPS to protect my freenas system only (E3-1230 with planned 8+ drives) or both of my servers, so addionally my I5-2500k esxi machine.

APC offers Back-UPS and Smart-UPS products and I found out that Smart-UPS seems to be the higher-quality product? Are there any problems using the USB interface of the APC devices with freenas?

What are the important points to look for?

 

[Ericloewe]

Smart-UPS is better, with pure sine wave output. Back-UPS is crummy. Back-UPS Pro is okay. At least I hope it's a reasonable approximation of a sine wave, I haven't actually seen the waveform (add that to the "When I finally get an oscilloscope" list).

CyberPower might have pure sine wave models around the price of APC's Back-UPS Pro, so look at those, too.

In any case, check out NUT's compatibility list before buying.

 

[Darren Myers]

I have 4 smart UPS's and theyre great, all mine are managed via a management card, now when i say that dont be scared of the 200$ price tag APC wants for new. i just picked up 2 more APC cards for 20$ each off ebay. I really like the SUA1500 , heres the rack mount mode and the pedestal one , and the management card that will work in either one

 

[Harsesis]

Thank you for your answers. Do you mean this list? If I interpret it correctly all APC Smart-UPS modesl are supported (at least using UBS)? I'm not sure about the size of the UPS, I think the 980Watt version is quite a bit overkill, at least for one server? I would have looked more into the 500 Watt range for the freenas server only?

I think I have to look into the management cards, at least when I want to hook up both servers to one UPS? (than for sure a bigger model)

 

[mattbbpl]

Thank you for your answers. Do you mean this list?

Yes, that's the list. I can't speak to your other questions yet, but that is the list Eric referred to.

 

[Fraoch]

Not to hijack this thread, but my preferred retailer has started stocking Eaton UPSes. Eaton is a very large commercial electrical supplier, they ought to know what they're doing. Also I seem to recall they cooperate very well with the NUT developers (the same can't be said for APC).

Anyone have any experience with Eaton UPSes?

It could be just the range of models my retailer stocks but they seem suspiciously cheap for their power output.

My main workstation's UPS might be on its way out so I'm starting to look around.

BTW two more knocks against APC: it reports virtually no data to NUT, just enough to report that it's online or on battery and not much else. There's also anecdotal evidence on another forum that APC UPSes are very hard on batteries - overcharging them, overheating them and generally shortening their life.

 

[cuban8]

I'm running an Eaton 5P850IR for ca. 3/4 of a year now with my 8-bay Freenas. Connection via USB and setup was simple, I didn't have to edit config files manually. Haven't really needed it so far though, predicted holdover time is about 1 hour with the NAS idling.

upsc output:

battery.capacity: 7.20
battery.charge: 100
battery.charge.low: 20
battery.charge.restart: 0
battery.energysave: no
battery.protection: yes
battery.runtime: 3600
battery.type: PbAc
battery.voltage: 25.8
battery.voltage.nominal: 24
device.mfr: EATON
device.model: 5P 850
device.serial: XXXXXXXXXX
device.type: ups
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 2
driver.parameter.port: /dev/ugen1.3
driver.version: 2.7.2
driver.version.data: MGE HID 1.33
driver.version.internal: 0.38
input.current: 0.00
input.frequency: 50.0
input.frequency.extended: no
input.frequency.nominal: 50
input.sensitivity: normal
input.transfer.boost.low: 184
input.transfer.high: 294
input.transfer.low: 160
input.transfer.trim.high: 265
input.voltage: 229.5
input.voltage.extended: no
input.voltage.nominal: 230
outlet.1.autoswitch.charge.low: 0
outlet.1.delay.shutdown: 65535
outlet.1.delay.start: 3
outlet.1.desc: PowerShare Outlet 1
outlet.1.id: 1
outlet.1.status: on
outlet.1.switchable: yes
outlet.2.autoswitch.charge.low: 0
outlet.2.delay.shutdown: 65535
outlet.2.delay.start: 6
outlet.2.desc: PowerShare Outlet 2
outlet.2.id: 2
outlet.2.status: on
outlet.2.switchable: yes
outlet.desc: Main Outlet
outlet.id: 0
outlet.switchable: no
output.current: 0.30
output.frequency: 50.0
output.frequency.nominal: 50
output.powerfactor: 0.86
output.voltage: 229.5
output.voltage.nominal: 230
ups.beeper.status: enabled
ups.delay.shutdown: 20
ups.delay.start: 30
ups.efficiency: 68
ups.firmware: 02.06.0014
ups.load: 11
ups.load.high: 105
ups.mfr: EATON
ups.model: 5P 850
ups.power: 69
ups.power.nominal: 850
ups.productid: ffff
ups.realpower: 60
ups.realpower.nominal: 600
ups.serial: XXXXXXXXXX
ups.shutdown: enabled
ups.start.auto: yes
ups.start.battery: yes
ups.start.reboot: yes
ups.status: OL CHRG
ups.test.interval: 2592000
ups.test.result: Done and passed
ups.timer.shutdown: 0
ups.timer.start: 0
ups.type: offline / line interactive
ups.vendorid: 0463

 

[rogerh]

I would check that the UPS you choose has a programmable delay before shutting down after a shutdown command. And that this is programmable from NUT. You can find out which NUT driver is applicable and read the man page for that driver. This is because the FreeNAS NUT implementation sends a shutdown command to the UPS before shutting down the operating system, and it is best to set a delay of two or three minutes to make sure FreeNAS shuts down cleanly. For some reason, UPS makers, including APC, make it incredibly difficult to establish whether this should happen, and even if it should it doesn't always work in practice. You can just not shut down the UPS, but this is only sensible if you will always be around when the power goes, otherwise it tends to be bad for UPS batteries: and the system won't restart on power return if the UPS never switched off.

 

[Fraoch]

I'm running an Eaton 5P850IR for ca. 3/4 of a year now with my 8-bay Freenas. Connection via USB and setup was simple, I didn't have to edit config files manually. Haven't really needed it so far though, predicted holdover time is about 1 hour with the NAS idling.

upsc output:

battery.capacity: 7.20
battery.charge: 100
battery.charge.low: 20
battery.charge.restart: 0
battery.energysave: no
battery.protection: yes
battery.runtime: 3600
battery.type: PbAc
battery.voltage: 25.8
battery.voltage.nominal: 24
device.mfr: EATON
device.model: 5P 850
device.serial: XXXXXXXXXX
device.type: ups
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 2
driver.parameter.port: /dev/ugen1.3
driver.version: 2.7.2
driver.version.data: MGE HID 1.33
driver.version.internal: 0.38
input.current: 0.00
input.frequency: 50.0
input.frequency.extended: no
input.frequency.nominal: 50
input.sensitivity: normal
input.transfer.boost.low: 184
input.transfer.high: 294
input.transfer.low: 160
input.transfer.trim.high: 265
input.voltage: 229.5
input.voltage.extended: no
input.voltage.nominal: 230
outlet.1.autoswitch.charge.low: 0
outlet.1.delay.shutdown: 65535
outlet.1.delay.start: 3
outlet.1.desc: PowerShare Outlet 1
outlet.1.id: 1
outlet.1.status: on
outlet.1.switchable: yes
outlet.2.autoswitch.charge.low: 0
outlet.2.delay.shutdown: 65535
outlet.2.delay.start: 6
outlet.2.desc: PowerShare Outlet 2
outlet.2.id: 2
outlet.2.status: on
outlet.2.switchable: yes
outlet.desc: Main Outlet
outlet.id: 0
outlet.switchable: no
output.current: 0.30
output.frequency: 50.0
output.frequency.nominal: 50
output.powerfactor: 0.86
output.voltage: 229.5
output.voltage.nominal: 230
ups.beeper.status: enabled
ups.delay.shutdown: 20
ups.delay.start: 30
ups.efficiency: 68
ups.firmware: 02.06.0014
ups.load: 11
ups.load.high: 105
ups.mfr: EATON
ups.model: 5P 850
ups.power: 69
ups.power.nominal: 850
ups.productid: ffff
ups.realpower: 60
ups.realpower.nominal: 600
ups.serial: XXXXXXXXXX
ups.shutdown: enabled
ups.start.auto: yes
ups.start.battery: yes
ups.start.reboot: yes
ups.status: OL CHRG
ups.test.interval: 2592000
ups.test.result: Done and passed
ups.timer.shutdown: 0
ups.timer.start: 0
ups.type: offline / line interactive
ups.vendorid: 0463

Nice. My APC Smart-UPS C 1500 has more data than I recalled, but nothing quite that extensive.

Code:

# upsc APC_Smart-UPS_C
battery.charge: 100
battery.charge.low: 10
battery.charge.warning: 50
battery.runtime: 10380
battery.runtime.low: 120
battery.type: PbAc
battery.voltage: 27.3
battery.voltage.nominal: 24.0
device.mfr: American Power Conversion
device.model: Smart-UPS C 1500
device.serial: XXXXXXXXXX
device.type: ups
driver.flag.pollonly: enabled
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 45
driver.parameter.port: /dev/ugen0.3
driver.version: 2.7.2
driver.version.data: APC HID 0.95
driver.version.internal: 0.38
ups.beeper.status: disabled
ups.delay.shutdown: 20
ups.firmware: UPS 09.8 / ID=1005
ups.mfr: American Power Conversion
ups.mfr.date: 2014/07/01
ups.model: Smart-UPS C 1500
ups.productid: 0003
ups.serial: XXXXXXXXXX
ups.status: OL
ups.timer.reboot: -1
ups.timer.shutdown: -1
ups.vendorid: 051d

And that's it.

Just recently, there was a power failure long enough to have FreeNAS shut itself down cleanly, but it's not set up to shut the UPS off as well - I don't think it's capable of that.

I recall I had problems getting it set up at first.

 

[rogerh]

Just recently, there was a power failure long enough to have FreeNAS shut itself down cleanly, but it's not set up to shut the UPS off as well - I don't think it's capable of that.

I recall I had problems getting it set up at first.

The capability to shutdown the UPS was added to a fairly recent 9.3 update; and the capability to not shut it down when you don't want to was enabled in the latest update. So if you're using a recent 9.3 update I would check whether it is going to surprise you by shutting down your UPS before FreeNAS shuts down. I notice the (default?) shutdown delay on your UPS is 20 seconds, not long enough if e.g. jails have to shut down.

 

[Fraoch]

The capability to shutdown the UPS was added to a fairly recent 9.3 update; and the capability to not shut it down when you don't want to was enabled in the latest update. So if you're using a recent 9.3 update I would check whether it is going to surprise you by shutting down your UPS before FreeNAS shuts down. I notice the (default?) shutdown delay on your UPS is 20 seconds, not long enough if e.g. jails have to shut down.

Whoa, I totally missed that - thanks!

 

[cyberjock]

Not to hijack this thread, but my preferred retailer has started stocking Eaton UPSes. Eaton is a very large commercial electrical supplier, they ought to know what they're doing. Also I seem to recall they cooperate very well with the NUT developers (the same can't be said for APC).

As someone that worked in electrical generation, I've learned that the larger commercial electrical companies are just as screwed up, money hungry, and quick to earn a buck at the behest of their customers as anyone else. I'm not saying Eaton is bad or anything, just that I don't think it's a good idea to judge them by size. ;)

I have no idea if they worked with NUT developers or not. I'd assume that if they did then they'd have a vast majority of the UPS market and this would be something that is well known, not because of the work they did, but because of the wide net they cast with perspective buyers looking for NUT-compatible UPSes.

BTW two more knocks against APC: it reports virtually no data to NUT, just enough to report that it's online or on battery and not much else. There's also anecdotal evidence on another forum that APC UPSes are very hard on batteries - overcharging them, overheating them and generally shortening their life.

Well, some APC UPSes do report virtually no data. Some report a crapload of data that has not been reverse engineered. It does depend on the model. I learned a lot about APC UPSes when I was setting up nut manually on a linux box 3 years ago and I had 3 different UPSes that I could use.

I'm not sure I buy the "very hard on batteries" story. Everyone fears for that kind of problem. I've never had a battery last less than 5 years, I've never had a battery overheat, and I've never replaced a battery that was swollen, misshapen, or had signs of any kind of overcharging. I could make the same anecdotal evidence that Seagate hard drives fail more frequently because they've tuned their firmware for performance instead of reliability. Proving Seagate does that is just as difficult as proving that APC is hard on the batteries. Unless someone is going to hook up a bunch of oscilloscopes, log UPS usage over time, along with time current, etc then it's just someone's opinion. I've owned about 90% APCs (currently own 5 of them) and they've all worked great except for one. One had a bad post on the battery so the UPS would touch the battery terminal and it would show as good, but if you did a battery test or put it under load (anything that required a good connection from the battery to the UPS itself for current flow) then the test would fail or the battery would lose power.

So yeah, calling it fake. But, that's just my opinion versus everyone else's. And we know opinions are like a$$holes...

 

[Fraoch]

I have no idea if they worked with NUT developers or not.

Huh, I had some old information. From http://www.networkupstools.org/acknowledgements.html:

Eaton, has been the main NUT supporter in the past, between 2007 and 2011, continuing MGE UPS SYSTEMS efforts. As such, Eaton has been:

providing extensive technical documents (Eaton protocols library),
providing units to developers of NUT and related projects,
hosting the networkupstools.org webserver (from 2007 to August 2012),
providing artwork,
promoting NUT in general,
supporting its customers using NUT.

Warning

The situation has evolved, and since 2011 Eaton does not support NUT anymore.
This may still evolve in the future.
But for now, please do not consider anymore that buying Eaton products will provide you with official support from Eaton, or a better level of device support in NUT.

Well, some APC UPSes do report virtually no data. Some report a crapload of data that has not been reverse engineered. It does depend on the model. I learned a lot about APC UPSes when I was setting up nut manually on a linux box 3 years ago and I had 3 different UPSes that I could use.

I'm not sure I buy the "very hard on batteries" story. Everyone fears for that kind of problem. I've never had a battery last less than 5 years, I've never had a battery overheat, and I've never replaced a battery that was swollen, misshapen, or had signs of any kind of overcharging.

So yeah, calling it fake. But, that's just my opinion versus everyone else's.

Fair enough. As I said, it was anecdotal, just a few posts on another forum. My APC UPS did come through for me when it was needed about a month ago. The batteries are so expensive around here though that if the batteries need to be replaced I might as well buy a new UPS - they cost about 2/3 what the UPS did new. So it's either look for a new one or find generic batteries.