[Solved] Local expression in Rule does not evaluate as they used to do

[wmarcolin]

@toggledbits

Hi Patrick

Well, I got all the installations done, and NUT is connected to the MSR.

Looking at the Reactor log, no error:

[latest-22149]2022-06-07T04:07:18.147Z <Structure:INFO> Structure#1 loading controller interface nut (NUTController)
[latest-22149]2022-06-07T04:07:18.158Z <default:null> Module NUTController v22152

[latest-22149]2022-06-07T04:07:18.272Z <Structure:INFO> Starting controller NUTController#nut
[latest-22149]2022-06-07T04:07:18.276Z <NUTController:NOTICE> NUTController#nut connecting to NUT server at "192.168.33.25":3493
[latest-22149]2022-06-07T04:07:18.278Z <NUTController:INFO> NUTController#nut started NUT client connection; waiting for ready

[latest-22149]2022-06-07T04:07:19.258Z <NUTController:INFO> NUTController#nut connected and ready!
[latest-22149]2022-06-07T04:07:19.260Z <NUTController:NOTICE> Controller NUTController#nut is now online.

In the VM using the command: upsc smart1500lcd@192.168.33.25 shows all the UPS data.

So, my difficulty as can be seen on the first screen, that I don't get the detailed UPS information in the MSR.

[toggledbits]

You are not searching correctly in the Entities list. Your UPS name does not have "nut" in it. Rather than seaching for entities containing the word "nut", clear all fields and select your NUTController instance in the controller filter field.

[wmarcolin]

@toggledbits

Perfect, and thanks again!!!

[wmarcolin]

Hi!

A query, my NUT service stops working recurrently, and when I make the query upsc myups@localhost I get the message Error: Data stale.

To get it working again I have to run upsdrvctl start to restart.

Has anyone experienced this difficulty?

[wmarcolin]

@toggledbits

Okay, job done, UPS > NUT > MSR > HE

The only thing that bothers me is that the UPS drive stops working without explanation, and I had to put a routine in crontab for it to restart every 10 minutes, I find this horrible but it was the only solution from everything I read, and I am not the only one to do it.

[toggledbits]

Does the NUTController's system entity show that the connection is down when that happens? If so, and NUT is running on the same host as MSR, it's probably pretty easy to just write a rule that restarts the NUT service when the controller has been marked down for X minutes (sustained for).

[wmarcolin]

@toggledbits

The NUT and MSR are in the same HW, but NUT is in a Linux VM, and MSR I run in Windows.

I was able to build a rule in MSR that analyzes 3 attributes of NUT, that if they stay without changing the content in 5 minutes gives me an alert of failure.

Strange, when I look at the service, it is all active, what crashes is the UPS driver, so I have to run

upsdrvctl stop
upsdrvctl start

Now, is it possible via Windows MSR to trigger a command in the Linux VM to do this restart? It would avoid having to crontab, only reboot in real failure conditions, and not every 10 minutes.

Thanks.

[Andr]

I rescued an Eaton Ellipse ups from work that come back to life with fresh batteries.

Thought it would be easy to get some kind of system status or notification (from Windows) that somehow could be picked up in Reactor or at least in HA. Oh my... what a horror

After a long time following this guide I got NUT up and running on Windows @wmarcolin (Another VM is not a resonable option here)
Monitor a UPS attached to a Windows machine

Then I added the HA integration just to se if it worked and it actually did

So now I want over the finish line, and install NUTController in Reactor.
It looked easy, but still, no luck.
This got me thinking on another possible solution. Reactor already monitors "System Memory" "Uptime" and so on, so maybe it also could monitor Powersource, Battery Level, Charging State? Then let Reactor react with its magic @toggledbits ?
For a Windows user that seems a lot easier and efficient as Windows just handle my UPS as the battery it actually is so it looked like "a laptop".

Anyway, here is were I got stuck. No clue what more to try since the Nutserver is up and running and speaks with HA.
reactor.log gives me two errors:

[latest-23196]2023-07-28T19:34:28.862Z <Structure:CRIT> TypeError: self.nut.setUsername is not a function [-]
TypeError: self.nut.setUsername is not a function
    at NUTController.start_client (c:\reactor\ext\NUTController\NUTController.js:221:22)
    at NUTController.start (c:\reactor\ext\NUTController\NUTController.js:44:14)
    at Structure.start (c:\reactor\server\lib\Structure.js:362:56)
[latest-23196]2023-07-28T19:34:28.863Z <NUTController:CRIT> TypeError: Cannot read properties of undefined (reading 'end') [-]
TypeError: Cannot read properties of undefined (reading 'end')
    at Nut.close (c:\reactor\ext\NUTController\node_modules\node-nut\node-nut.js:65:15)
    at NUTController.stop (c:\reactor\ext\NUTController\NUTController.js:53:26)
    at c:\reactor\server\lib\Structure.js:366:48
[latest-23196]2023-07-28T19:34:28.864Z <Controller:NOTICE> NUTController#nut stopping

NUTController.js:221:22 looks like this

NUTController.js:44:14

Structure.js:362:56

node-nut.js:65:15

NUTController.js:53:26

Structure.js:366:48

[toggledbits]

Go into the extension directory (ext/NutController) and run npm list and post the output here.

Also you need to show your NutController config.

[Andr]

@toggledbits
So, back from vacation so time to take another shot at this

npm list:

C:\reactor\ext\NUTController>npm list
npm WARN config global `--global`, `--local` are deprecated. Use `--location=global` instead.
reactor-nutcontroller@0.1.22305 C:\reactor\ext\NUTController
`-- node-nut@1.0.3

NutController config

  - id: nut
    implementation: NUTController
    enabled: true
    #name: Eddie UPS NUTController
    config:
      server: 22.33.44.11        # modify the IP address as needed
      port: 3493              # optional, default shown
      username: "***"      # optional, no user auth if not set
      password: "***"  # optional, must be specified if username is used

[toggledbits]

Try commenting out the username from config. Authentication isn't needed for the queries being used.

[Andr]

@toggledbits
Yep, that did it
Thanks once again @toggledbits

Any thougts about getting more System Info into MSR by the way?
This got me thinking on another possible solution. Reactor already monitors "System Memory" "Uptime" and so on, so maybe it also could monitor Powersource, Battery Level, Charging State? Then let Reactor react with its magic

[toggledbits]

Hmmm... are you (mis)understanding the way NUTController publishes its data? It publishes UPS info on an entity for each UPS that NUT reports. Here, for example, is the data for a test "dummy" UPS that one can configure in NUT:

battery_maintenance.charging=false
battery_maintenance.rechargeable=true
battery_maintenance.replace=true
battery_power.level=0.2
battery_power.since=1692217178132
power_source.source="battery"
x_nut_ups.battery_charge="20"
x_nut_ups.battery_charger.status="discharging"
x_nut_ups.battery_date.maintenance="05/01/2022"
x_nut_ups.device_mfr="Dummy Manufacturer"
x_nut_ups.device_model="Dummy UPS"
x_nut_ups.device_type="ups"
x_nut_ups.driver_name="dummy-ups"
x_nut_ups.driver_parameter.mode="dummy"
x_nut_ups.driver_parameter.pollinterval="2"
x_nut_ups.driver_parameter.port="/etc/nut/dummy.seq"
x_nut_ups.driver_parameter.synchronous="no"
x_nut_ups.driver_version="2.7.4"
x_nut_ups.driver_version.internal="0.14"
x_nut_ups.input_voltage=1580
x_nut_ups.ups_mfr="Dummy Manufacturer"
x_nut_ups.ups_model="Dummy UPS"
x_nut_ups.ups_status="OB LB"

So you can see it's reporting battery state, power level, current power source, etc.

Or are you asking for something else and I'm not getting it?

[Andr]

@toggledbits
No, I get how NutController works. So far at least...

But when I got stuck several times trying to set up NUT I almost gave up. At the same time I remembered seeing some System Info in MSR and if could find data about battery level, charging and so on I would not even need the NUT Installation.
And that would be a big bonus for me as I realize that when my NUT installation fails one day it would take me long time to figure it out and also one less process/config to handle.

And from System Info I assume we could get info about temperatures, storage levels, fan speeds and so on?
This could be used to send warning notifications from MSR.

So yes, I was asking something else that come up in my head when I tried to find another way than using NUT.