Rysz

I can't really help more there, unfortunately, it's most probably a driver issue. The default refresh is 5 seconds for the status, so I'm guessing the connection is going stale. You could open an issue on NUT GitHub about it: https://github.com/networkupstools/nut/issues

I'd still check if the cables are somehow loose on the board or the HDDs when the parity check is done. Just to be on the safe side... connection issues can cause a lot of problems down the road even if it works now.

Honestly I'd go the Docker route... Running all that on the OS itself and maintaining the dependencies is going to be hell. I only see disadvantages here as opposed to a Docker image which is detached from the OS.

Also habe diese Probleme selbst nie gehabt und diese Überlegungen zum Glück auch nie anstellen müssen. Mein System im Node 804 läuft bei mir seit 4 Jahren mit diesen Temperaturen auch bei 40 Grad im Hochsommer... 🙂 Mit zwei Lüftern vorne und zwei Lüftern hinten: 27°C im Leerlauf und 35°C unter Last (HDDs) Auch bei der CPU bin ich bei 40°C im Leerlauf und max. 60°C unter Last - mittlere Lüfterstufe (2/3). Intel-Stock-Kühler + 1x mitgelieferter Gehäuselüfter + 3x be quiet! Shadow Wings 2" (120 mm) Gehäuselüfter.

Überhaupt kein Problem von der Lautstärke, das Lauteste an dem System sind meine Exos Festplatten. CPU habe ich sogar den Intel-Stock-Lüfter und die restlichen Lüfter sind "be quiet! Shadow Wings 2" (120 mm). Habe den Server auch bei mir im Büro und draußen (also vor der Tür) hört man überhaupt gar nichts. Sitze jeden Tag am Schreibtisch direkt neben dem Server und fühle mich auch direkt daneben nicht gestört. Einzig die Exos Festplatten nerven manchmal bei vielen Lese- und Schreibvorgängen, das wäre meine einzige Kritik. Und neben dem Server schlafen könnte ich wahrscheinlich nicht, aber hinter geschlossener Tür hört man überhaupt nichts. Edit: Glaube ein Lüfter war bereits im Gehäuse, den habe ich auch belassen, der Rest ist aber von "be quiet!".

Also ich habe ein Node 804 mit maximaler Anzahl an Platten und kein Airflow-Problem... 🧐 Mit zwei Lüftern vorne und zwei Lüftern hinten: 27°C im Leerlauf und 35°C unter Last (HDDs) Auch bei der CPU bin ich bei 40°C im Leerlauf und max. 60°C unter Last - mittlere Lüfterstufe (2/3).

It shouldn't really cause very noticeable CPU utilization but depends on the UPS driver and your system too (CPU/RAM). Then again NUT does have 3 different services running (driver, upsd and upsmon) so it's not extremely lightweight. Could also be the USB bus or CyberPower subdriver needing to do more heavy lifting to speak the UPS required protocol and communication language. Bottom line it's hard to say, if you're on an old system with a weaker processor or few RAM reserves it might be more noticeable. But there's not really much we can do, if it persists you could attempt to restart the server at some point. A restart of the system will always freshly install NUT onto the RAM disk, so you'll start with a fresh package just to rule out any brokenness causing this. Another USB port is also worth attempting. Also check out the "pollfreq" option, may or may not be worth playing around with: https://networkupstools.org/docs/man/usbhid-ups.html

You can't really edit those files as they are only present in RAM and any changes will be lost after a reboot. The problem with allowing additional user-scripting within NUT's shutdown routine is users could in theory torpedo Unraid's graceful shutdown process if they're not careful about what they are doing and solving this with inbuilt timeouts and whatnot would be a bit out of NUT's scope. In general the NUT server should be the machine that shuts down last, in your case it's a bit of an exotic situation compared to most user's setups but I understand what you're trying to achieve. So in your case I'd advise to stick with the User Scripts solution because that way you'll still get the graceful shutdown if your script should fail for any reason because the Unraid shutdown time-out (configurable under "Disk Settings") will kick in and proceed with the graceful shutdown regardless of a hanging User Script. Worst case it won't be able to shutdown the ESXI hypervisor but your Unraid VM will be safely shutdown anyhow. Sorry I can't be of more help there!

Sure - set everything up in GUI first - then switch "Enable Manual Configuration:" to "Yes". Afterwards you can modify upsmon.conf inside the GUI editor: NOTIFYCMD "/usr/sbin/upssched" NOTIFYFLAG as needed (see manual) Rest of the configuration is done in upssched.conf also inside GUI editor. Manual mode ensures the changes won't get overwritten; manuals for configuration are here: https://networkupstools.org/docs/man/upssched.html https://networkupstools.org/docs/man/upssched.conf.html Make sure to keep "Enable Manual Configuration:" on "Yes" otherwise the changes are overwritten. Note that using upssched you'll only have the default NUT functionality (which is initiating shutdown on UPS low battery event) and no additional triggers as with the GUI (such as battery level, runtime remaining, time on battery) so you'll have to set those up yourself within upssched if you need them. Manual configuration mode only uses what's set up inside the configuration files, hence most GUI settings are not shown in this mode.

Can users with old licenses still use pre-6.12 versions, will the key remain valid for all those older versions as well? Meaning will users on older versions be forced to update at some point or will the old license always be valid for all versions no matter how old or new. What about the new licenses under the new model, can a new license be used to activate pre-6.12 releases for those who would like to run such older versions e.g. for hardware compatibility or whatever reasons? I think it would be fair for it to be valid on all old versions and one year into the future if it's not extended.

I'm not getting involved with the guesswork about a possible future licensing model. Should the financial need for a subscription model arise, I'll be happy to consider it for the system I've grown to love. But it really hurts me to see that Unraid as a brand is taking what seems like a lot of unnecessary PR damage right now. Unraid has always been very community reliant and driven and I'd say that was one the most liked points besides the OS. All of the hate and doomsday scenarios being spun online right now could've been avoided by better communication. I really hope a lesson is learnt from this and we'll see more up-front communication, rather than picking up the pieces later.

No, shows just fine here - perhaps you were looking for an older and deprecated version? The only currently and actively maintained one is by me (Rysz).

Can you try changing the NUT backend to "release (2.8.0 stable)" and see if that works? In NUT Settings using the "Network UPS Tools Backend Switch" (need to reboot after switching). If that doesn't work you can also try the oldest backend "legacy (2.7.4. stable)". Please let me know if any of those suggestions worked for you!

That's what I meant, you can create a seperate .page file in your plugin for this, specifically by using the "Buttons" menu your hidden code will be available on all GUI pages. Then you can use PHP to make sure it is included only on the "Main" page where it is needed. Using JS you could then communicate with (for example) a PHP script of your plugin that returns the parity check status, afterwards parsing that result and then using jQuery to enable/disable the respective buttons on the fly. That checking function could be run periodically using a timer, say every 5 seconds or whatever time you set. There's no real need to cause a full page refresh, you could just update the relevant parts using jQuery. Here's an example .page file that would achieve such a hidden code insertion: Menu="Buttons:198" Link="nav-user" --- <?if(stripos($path, "main") !== false):?> <script type="text/javascript"> // this section gets added to all GUI pages where path has 'main' in the URL // could check for parity check status here, e.g. communicating with PHP script // or checking for file existence. could parse result and enable/disable parity // buttons using jQuery afterwards. a timer could run check function periodically. </script> <?endif;?>

Why not do it via .page and JS, periodically checking the status of the parity operation and enabling/disabling buttons accordingly? There'd be no need to do a full page refresh with JS checking the parity operation's status and then enabling/disabling buttons via jQuery.

Then it's probably just reported wrongly by the driver, if the beeper is otherwise working that's the most important thing.

Can you try beeper.on instead of beeper.enable?

Did you enable ModBus on the UPS itself? It needs to be specifically enabled on the hardware and particularly if the regular USB driver works for you that seems to me a sign that ModBus is not enabled on your UPS.

It seems APCUPSD development has been stale for a couple of years now, so wouldn't expect much there. In the meantime it seems the NUT ModBus implementation seems to work (see above), so might be worth testing that.

Since your UPS doesn't provide the actual power draw it's calculated from the load and nominal power (wrongly 388W). There's a setting in the GUI: "UPS Power and Load Display Settings" where you can override the wrongly reported nominal power values ("Manual"). You should put your actual values 1500VA/900W there and it'll then show the correct power draw in the GUI. Please let me know if that worked for you!

Of course, just set "Start Network UPS Tools Service" to "Yes" and click on "Apply". If it is currently set to "Yes": Set it to "No" -> Click "Apply" -> Set it to "Yes" -> Click "Apply". Please let me know if that worked for you 🙂

Looking good, make sure to connect the UPS to your UNRAID (e.g. via USB) using the inbuilt "UPS Settings" (apcupsd) or the NUT plugin. This'll ensure you'll have graceful shutdowns when the power goes out. Some people also had their HBA cards run extremely hot and strapped a small fan to them. This of course depends on the airflow situation in your case, just something to watch out for in case of problems. 🙂

Cheers, I think this is what I'll do (more of the manual type) - thanks again!

Just to be clear, this would require me to update to 6.12.x and then a factory reset using that, right? 🙂

Hello experts! I have a really old UNRAID backup server running 6.8.3. which has never given me any trouble so far. As that version is so outdated I'd like to finally migrate to 6.12.x BUT preferably with a new installation of UNRAID rather doing the large step updating from 6.8.3 to 6.12.x Is there a way to start over with a fresh installation of UNRAID while preserving both parity and data on my current drives (meaning no rebuilds or parity-sync)? Essentially I would like to tell the new installation here's 2 parity and 6 data drives, parity and data is intact so please use that and don't do anything to my drives. 😄 Any pointers in the right direction would be much appreciated! 😉