Cyberpower UPS CP685AVR - can't start ups service

[ethereal]

i tried to configure the ups in freenas but i couldn't see which port to use - i thought if i start the ups service before i did the configuration then the port might show up. when i tried to start the ups service i got the following errors.

Dec 21 16:10:04 freenas notifier: Can't start /usr/local/libexec/nut/(null): No such file or directory
Dec 21 16:10:04 freenas notifier: Network UPS Tools - UPS driver controller 2.7.3
Dec 21 16:10:04 freenas root: /usr/local/etc/rc.d/nut: WARNING: failed precmd routine for nut
Dec 21 16:10:04 freenas notifier: /usr/local/etc/rc.d/nut: WARNING: failed precmd routine for nut
Dec 21 16:10:04 freenas notifier: Starting nut_upsmon.
Dec 21 16:10:04 freenas notifier: fopen /var/db/nut/upsmon.pid: No such file or directory
Dec 21 16:10:04 freenas notifier: Unable to use old-style MONITOR line without a username
Dec 21 16:10:04 freenas notifier: Convert it and add a username to upsd.users - see the documentation
Dec 21 16:10:04 freenas notifier: Fatal error: unusable configuration
Dec 21 16:10:04 freenas notifier: Network UPS Tools upsmon 2.7.3
Dec 21 16:10:04 freenas root: /usr/local/etc/rc.d/nut_upsmon: WARNING: failed to start nut_upsmon
Dec 21 16:10:04 freenas notifier: /usr/local/etc/rc.d/nut_upsmon: WARNING: failed to start nut_upsmon
Dec 21 16:10:04 freenas notifier: Starting nut_upslog.
Dec 21 16:10:04 freenas notifier: Warning: initial connect failed: Connection failure: Connection refused
Dec 21 16:10:09 freenas notifier: Stopping nut_upslog.
Dec 21 16:10:09 freenas notifier: Waiting for PIDS: 6219.
Dec 21 16:10:09 freenas notifier: nut_upsmon not running? (check /var/db/nut/upsmon.pid).
Dec 21 16:10:09 freenas notifier: nut not running? (check /var/db/nut/upsd.pid).

i'd appreciate a little help because i have never used an ups before - never mind configuring freenas

a Christmas kiss to anybody who can help

 

[ethereal]

i got it working - i am no longer a ups virgin.

i configured the ups service first - i noticed that there was another option in the port dropdown that wasn't there before to i selected that one. then when i started the ups service it seemed to go well.

i unplugged the ups - i got an email saying ONBATT and after 30 seconds freenas shutdown and i got another email saying SHUTDOWN.

i am a man now - christmas kisses for me alone

 

[rogerh]

I think that, as you seem to have found, some files don't get written unless you configure NUT first. And I don't think you can configure it until you have entered the right driver and it finds the ups. You also have to fill in all the other fields appropriately. I would leave the 'upsmon' user as the default and just give it a simple password for now.

Edit: glad you got it working while I was prevaricating: and I don't think there is any simple correlation between the allocation of kisses and gender nowadays.

 

[DrKK]

The CP685AVR, by the way, will power a small FreeNAS box for an hour. It certainly powers mine for at least an hour. You should feel free to switch the shutdown timer to "shut down when battery is low".

 

[ethereal]

i was a little worried about it because i have (10 hdd - 6 case fans - hba - m/b - cpu and memory) - i worked this out to need 387 watss at spin-up. this ups only gave out 390 watts. the only thing i connected to the battery, initially was freenas and when i measured it, freenas was only using 114w idle and 141 w scrubbing both pools. as i am a gambler, a rogerer and a puker i added my router (wrt1900ac) to the battery. Router and an idle freenas still only uses 121 w.

i thank you for your advice - i think i'll give it a go. i'm glad i purchased the cheap crap one. i was going to buy a true sine wave at about twice the price.

 

[DrKK]

This is a lot bigger than my system. I'm drawing more like 35W at idle, 50W under load, on my FreeNAS.

i was going to buy a true sine wave at about twice the price.

Nah. This is one of those nerd jihad issues where, in my view, I think the nerds are wrong in practice even when they're right in principle. About 100 of the guys right now are reading this, and are chomping at the bit to foam at the mouth about how you want pure sine wave UPS's. The reality, however, is that they are way, WAY more expensive, and the typical guy with a FreeNAS is usually running equipment---and at draws---that tolerate choppy step wave cheap UPS's very well. Remember, there's a lot of filtering, capacitors, inductors, transistors, between your devices (hard drives, mobo) and the dirty step wave. The worst thing I have ever seen happen on my equipment is to hear a little bit of noise suggesting that the PSU wasn't "happy" about the non-sinusoidal wave, but I have had all of my computer equipment, including my FreeNAS, log many, many hours on cheaper UPS's (APC Back-ups bottom-of-the-line models, Cyberpower various bottom-of-the-line models...) with no discernible short- or long-term damage. I mean, if the price difference was $30, then, of course, by all means, I would be recommending the pure sine wave things. But the price difference is very substantial, and to justify it, I'll need more than nerd anus-clenching philosophical arguments; on the ground floor with real world equipment, it seems to tolerate shitty UPS's quite well. By the time you're buying a pure-sine-wave UPS, for most home users, you are spending more already than the cost it would take to replace most of the equipment you're protecting with the UPS in the spectacularly unlikely event that the non-pure-sine-wave would damage it.

Just doesn't make sense to me.

Two biggest money-down-the-drain things I think a lot of home small-deploy FreeNAS users do is the over-buy CPU's (literally, almost no one, NO ONE, running a home FreeNAS needs a Xeon CPU. 92% of the people need just a G3220-like CPU, and 92% of what remains need an i3), and then over-buy UPS's as a close second.

 

[rogerh]

I am tempted by the inline UPSs that have the inverter constantly supplying power, rather than starting when line power is lost. However, I probably shan't buy one until I see a credible report of a reputable PSU failing to maintain a computer running when the UPS switched from line to battery power.

 

[Bidule0hm]

The ATX specs says an ATX PSU should keep the output stable during at least 16 ms (or was it 20?) without input power, any decent PSU should follow the spec and any decent UPS will take less than half that to switch from mains to battery.

 

[Ericloewe]

I am tempted by the inline UPSs that have the inverter constantly supplying power, rather than starting when line power is lost. However, I probably shan't buy one until I see a credible report of a reputable PSU failing to maintain a computer running when the UPS switched from line to battery power.

El-cheapos will certainly not always manage, but @Bidule0hm is right.

The ATX specs says an ATX PSU should keep the output stable during at least 16 ms (or was it 20?) without input power, any decent PSU should follow the spec and any decent UPS will take less than half that to switch from mains to battery.

 

[rogerh]

The ATX specs says an ATX PSU should keep the output stable during at least 16 ms (or was it 20?) without input power, any decent PSU should follow the spec and any decent UPS will take less than half that to switch from mains to battery.

Specs are good; but practical operation is even better!

 

[Ericloewe]

Specs are good; but practical operation is even better!

Seasonic tends to massively overengineer their high-end PSUs to beat the spec by some 50% or more, for instance. G-Series is slightly above spec.

Low-end stuff (think Corsair CX/CS) tends to not really meet the spec. Still, I've never seen it be a problem.

 

[Bidule0hm]

Yep, exactly that. That's one of the reasons you want a good quality PSU in your server ;)

 

[tvsjr]

I can say, from first-hand experience, some of the newer power-factor-correcting supplies do NOT play nicely with the "modified sine wave" outputs of cheap UPSes. They'll just turn off. Cyberpower got around this in their PFC series of UPS by using a triangle wave output... it's not as hard/expensive to produce as pure-sine, but it doesn't wig the PFC power supplies out. I suspect more recent PSUs have worked around this problem; however.

Now, my FreeNAS (and a few other servers) are plugged into a 3KVa online pure-sine UPS with extra battery packs (for about an hour of runtime)... but that might be a bit overkill :D

 

[Bidule0hm]

Excepted I have a pseudo sine wave UPS (more or less a square wave in fact...) and the PSU (SeaSonic X-650) didn't have any problem in all of my tests so I guess it's again not really PFC vs no PFC but good vs crappy quality PSUs :)

But of course a true sine wave UPS is far better ;)

 

[tvsjr]

My bad experience was in the early days of the PFC power supplies (which came about, primarily, due to the 80Plus certifications). There are so many non-sine output UPSs out there that the manufacturers likely had to find away to deal with the problem. Mandating that everyone go buy a pure-sine UPS wouldn't work.

The reality is, PSUs are pretty incredible these days - especially good ones like Seasonic. Feed them anything that sorta looks like AC, at most any frequency, at any reasonable voltage... and they'll make tons of nice clean DC.