INTERNAL MULTI UPS: First ultra-mega-beta (glitchy) React build is here

[DartSteven]

Ok folks: here’s the link + the docker-compose to try the first INTERNAL image I’ve ever shipped.
It’s ultra mega beta, it’s glitchy, and it’s not pretty yet, but I’m genuinely happy because the React port is finally starting to pay off.

PLEASE DON’T ROAST ME (YET)

Migrating a real project from “plain JavaScript stuff” to a proper React structure is… long. Very long.
But I’m convinced it’ll be worth it once the foundation is solid.

KNOWN ISSUES

Yes: I’m already aware there are graphic bugs, layout glitches, weird UI behaviors, and overall “beta”.
I’m tracking them as I spot them, but I’d love help from anyone testing this.

HOW TO REPORT BUGS (THIS MATTERS)

If you find anything (from the smallest detail to the most critical bug) please report it, but use a bit of criteria so we keep things manageable.

Please use this format:

• Title: DEV-Internal BUG: <short description>
• Include:
  • What you expected vs what happened
  • Steps to reproduce (even rough steps help)
  • Screenshot / short video if possible
  • Your environment (OS, browser, resolution)
  • If it happens every time or randomly

If you’re unsure it’s a bug, report it anyway — just label it clearly.
If anything fails, paste the logs in your issue and I’ll dig in.

WHAT I NEED MOST RIGHT NOW

• Layout / responsive issues
• Any console errors
• Anything that looks broken or inconsistent
• Performance hiccups (no need to benchmark like crazy yet)

EXTRA IDEAS (OPTIONAL BUT USEFUL)

• Add a short “What this build is / is not” note: INTERNAL dev preview, not stable, not feature-complete
• Add “Top priorities” (3–5 bullets) so know what to focus on
• Pick one reporting channel (Issues) to avoid splitting feedback

Thanks to everyone who tests and reports stuff :)

Looking for a bit of UI/CSS help (if anyone feels like it)

I’m working on the React port and I’m trying my best, but the UI/CSS part is not my strongest area.
If someone here enjoys polishing interfaces and has the skills, I’d genuinely appreciate a hand.

Even small help is welcome: quick suggestions, fixing a layout bug, improving spacing, making things more consistent, proposing a nicer style… anything.

What would help me a lot

• Better spacing and alignment (desktop + mobile)
• Cleaning up CSS / responsive behavior
• Making the UI look more coherent
• Optional: simple, lightweight animations/transitions

If you’d like to collaborate, please drop a message in a Discussion (or comment on the preview thread) and use a title like:
DEV-Internal UI: Collaboration

Thanks for even reading this ( and no pressure at all. )

[DartSteven]

WHAT’S INCLUDED

• docker-compose.yml:

services:
  nut:
    image: dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.02
    container_name: Nutify
    privileged: true
    cap_add:
      - SYS_ADMIN
      - SYS_RAWIO
      - MKNOD
    devices:
      - /dev/bus/usb:/dev/bus/usb:rwm
    device_cgroup_rules:
      - 'c 189:* rwm'
    volumes:
      - ./Nutify/logs:/app/nutify/logs
      - ./Nutify/instance:/app/nutify/instance
      - ./Nutify/ssl:/app/ssl
      - ./Nutify/etc/nut:/etc/nut
      - /dev:/dev:rw              
      - /run/udev:/run/udev:ro 
    environment:
      - SECRET_KEY=test1234567890 
      - UDEV=1                   
    ports:
      - 3493:3493
      - 5050:5050
      - 443:443
    restart: always
    user: root

HOW TO RUN (DOCKER)

docker compose up -d

⚠️ Please don’t run this on a “dirty” setup (seriously)

Not just recommended — mandatory (and yeah, I’m half joking… half).

• Run this instance in a clean environment
• Do NOT reuse the old database
• Ideally: clone/extract it into a brand-new, empty folder and run from there

Also: the database has been heavily reworked and it is not compatible with the previous one in any way.
If you try to plug in the old DB, you will get errors (or worse: subtle, cursed behavior).

[DartSteven]

Docker Hub Images

Updated image links and a small changelog for each build will live here.

• build link: 19/02/2026 dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.01
  Changelog : First commit in multi nut ups

• build link: 19/02/2026 dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.02
  Changelog : Fixed an issue in the database that was causing it to grow excessively, and resolved a glitch in the Wizard.

• build link: 22/02/2026 dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.03
  Changelog : - Fixed the layout of the log options buttons, Added Scope ID and Target support for multi-UPS in each notification section, Fixed reports, Rebuilt database assets, Added an upsmon script for single-UPS NUT “vanilla” mode: monitors the UPS and sends notifications to Nutify via HTTP

• build link: 26/02/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.05-amd64

• build link: 26/02/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.13-amd64

• build link: 04/03/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-amd64

• build link: 04/03/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-mac-arm64

• build link: 04/03/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi5-arm64

• build link: 04/03/2026 dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi4-armv7

[czech12]

Hello,

I just learned about Nutify yesterday. I wanted to deploy nut (which I'm pretty new to as well) but was looking for a nice UI to go along with it. I found your project on reddit. Thanks for your hard work, this is a great tool and extremely useful.

I deployed the latest version (0.1.7) and got it working with my first UPS. I wanted to add another UPS and that is when I found that this feature is in active development. A side note, but I was still able to get nut working with multiple UPSs on version 0.1.7, it's just that only the first one added shows up in the web gui for reporting. You can still have nut reference 2 UPSs and configure clients to use one or the other, it's just command line only.

My installation is a bit unique. Both of my UPSs are network connected, so I only needed the snmp-ups driver. I don't need any support for USB devices. I also prefer deploying containers in kubernetes, so this is how I'm using nutify. The web frontend is exposed using traefik reverse proxy and the nut port is exposed using a 'loadbalancer' ip.

I was able to take my deployment of 0.1.7 and clone it to a brand new deployment for testing, only changing the image to dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.02. To be clear, this wasn't a dirty upgrade, I started fresh again. Everything seems to be working the same as it was in the 0.1.7 version.

The wizard is mostly okay, but when configuring snmp-ups as the driver, the config is incomplete since it doesn't ask for the SNMP community string. This was not a problem for me though, I just manually modified the ups.conf file once I was able to access the settings. This was the same case in 0.1.7 though.

Once I got into the settings, I was able to update ups.conf and add both of my UPSs correctly. I also manually configured upsd.users upsmon.conf to my liking. After restarting services, nutify picked up both of my UPSs and the switch between them is working as well! Awesome!

I noticed it named my first UPS "Primary UPS". It took me a minute to figure out how to adjust this. I kept hitting the edit button but it didn't look like it was doing anything. I then realized I could edit the names above. One thing that is a little weird is the IPs listed in this section are 127.0.0.1, not the IP configured in ups.conf.

Here are a list of bugs and/or weirdness I saw when using the test version.

• When editing the nut files, the window is really small and I only have an option to make it longer, not wider. This was not a problem in 0.1.7.

• Setting the currency is a per UPS setting. I think it makes more sense for this to be a global option.

• Nominal Power is something I need to manually set. When I saved this setting, the container restarted without warning. It restarted with the setting saved though. It looks like this is a global setting, not sure it if makes more sense to be per UPS.

• Very minor, but the side bar seems to be vertically centered now. I think it looked better when it was locked to the top.

Everything else seems functional. Thanks again for a great application. I hope this feedback helps.

BONUS: Below is my k8s deployment:

---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: nutify-logs-pvc
  namespace: nutify
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: nutify-instance-pvc
  namespace: nutify
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 500Mi
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: nutify-ssl-pvc
  namespace: nutify
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 100Mi
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: nutify-nut-pvc
  namespace: nutify
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 100Mi
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nutify
  namespace: nutify
  labels:
    app: nutify
spec:
  replicas: 1
  selector:
    matchLabels:
      app: nutify
  template:
    metadata:
      labels:
        app: nutify
    spec:
      containers:
      - name: nutify
        image: dartsteven/nutify-internal-testing:amd64-0.2.0.Int.Testing.v.02
        securityContext:
          privileged: true
          capabilities:
            add: ["SYS_ADMIN"]
        env:
        - name: SECRET_KEY
          valueFrom:
            secretKeyRef:
              name: nutify-secret
              key: SECRET_KEY
        - name: UDEV
          value: "0"
        ports:
        - containerPort: 3493
          name: nut
        - containerPort: 5050
          name: web
        volumeMounts:
        - mountPath: /app/nutify/logs
          name: logs
        - mountPath: /app/nutify/instance
          name: instance
        - mountPath: /app/nutify/ssl
          name: ssl
        - mountPath: /etc/nut
          name: nut-config
        resources:
          requests:
            cpu: 200m
            memory: 256Mi
          limits:
            cpu: 1000m
            memory: 1024Mi
        stdin: true
        tty: true
      volumes:
      - name: logs
        persistentVolumeClaim:
          claimName: nutify-logs-pvc
      - name: instance
        persistentVolumeClaim:
          claimName: nutify-instance-pvc
      - name: ssl
        persistentVolumeClaim:
          claimName: nutify-ssl-pvc
      - name: nut-config
        persistentVolumeClaim:
          claimName: nutify-nut-pvc
---
apiVersion: v1
kind: Service
metadata:
  name: nutify-service  # Keep for Ingress
  namespace: nutify
spec:
  selector:
    app: nutify
  ports:
  - name: web
    port: 5050
    targetPort: 5050
  type: ClusterIP
---
apiVersion: v1
kind: Service
metadata:
  name: nutify-nut-lb
  namespace: nutify
  labels:
    app: nutify-nut
spec:
  selector:
    app: nutify
  ports:
  - name: nut
    port: 3493
    targetPort: 3493
  type: LoadBalancer

[czech12]

I just realized that I initially deployed the .02 test version. (as shown in my k8s manifest above).

I was able to deploy .03 test version by only switching the image. All configuration and persistent storage remained intact.

The issue I noted above with the editor and the small window size is resolved.

Also, hitting the edit button on the NUT Manager page makes it much more obvious since the settings fly out below.

I'll report back if I find anything else. Thanks again!

[DartSteven]

Hi, tonight or tomorrow morning I’ll make a new build of the project, which has now reached about 95% of the goals I had set.

If you can, I’d really appreciate it if you could test it thoroughly, because in a few days (no later than the end of next week) I’ll push the final progress update of the project to GitHub.

[Dewi4nt]

A bug i just noticed, the changes in "Power Flow" settings do not save. At least for me.

[czech12]

I just upgraded to v13 about an hour ago. I did need to start fresh again, but I had a much better experience with the wizard. I like how the timezone and currency were both selectable in there.

I just tried to change the power flow settings and I didn't have a problem. I changed the currency from USD to EUR and validated that it updated. I just changed it back to USD and that worked too. I also adjusted the Cost per kWh, it saved the change but it's difficult to tell if the cost adjusted. Are these the settings you are referring to?

[Dewi4nt]

Hmmmm... It just updated for me as well... after like 15 minutes of still showing EUR... Dunno, probably will set up net, clean container tomorrow.

Thanks for retesting that :)

[DartSteven]

docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-amd64

[czech12]

Do you have this latest testing version in arm64 architecture? I'd like to test this on the cluster it will eventually run on. Thanks.

[DartSteven]

Do you have this latest testing version in arm64 architecture? I'd like to test this on the cluster it will eventually run on. Thanks.

docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-mac-arm64
docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi5-arm64
docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi4-armv7

[Dewi4nt]

I just wonder, where the battery Health is coming from?

In the stable version of Nutify I can see 100% there, but in this beta there's only 54% so I wonder :D

[czech12]

Do you have this latest testing version in arm64 architecture? I'd like to test this on the cluster it will eventually run on. Thanks.

docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-mac-arm64 docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi5-arm64 docker pull dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi4-armv7

Thanks for posting these. I haven't had much time to test, but I did not have the same experience when using 0.2.0.Int.Testing.v.24-raspberrypi5-arm64. Everything appears to work application wise. I was able to get through the wizard, but both of my snmp-ups driver ups's now say NOCOMM. It's the first time I've seen this error in any of my testing.

When I used version 0.1.7, I chose the arm64 package and everything worked for me. I was a little confused which package to use when you posted 3 variants of the arm packages. I went with the arm64 variant because that's what I was using before. However, I am running raspberry pi 4 hardware.

I also just tried dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi4-armv7. When using this one, the container didn't start properly, so I think I'm using the correct image when using dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.24-raspberrypi5-arm64.

[DartSteven]

The Battery Health widget computes a score from four live inputs:

Battery charge
Runtime
Voltage
Temperature
Each input is normalized to a safe range before scoring:

Charge: 0–100
Runtime (minutes): 0–240
Voltage: 0–64
Intermediate subscores are then calculated:

runtimeScore = (runtime / 180) * 100 (bounded to 0–100)
voltageScore = (voltage / 32) * 100 (bounded to 0–100)
temperatureScore uses bands:
Missing temperature: 60
20–30°C: 100
15–35°C: 70
10–40°C: 40
Outside these ranges: 20
Final score (bounded to 0–100):

score = round(charge * 0.45 + runtimeScore * 0.25 + voltageScore * 0.15 + temperatureScore * 0.15)
Health levels:

= 85 → Excellent
= 65 → Good
= 45 → Watch
< 45 → Critical
Update Behavior
The widget updates from real-time UPS snapshots when real-time mode is active.
If real-time data is missing, it falls back to the latest available stats/metrics values.

The score is recalculated whenever any of the four inputs changes.
The panel shows:

Gauge
Health level
Summary text
First recommendation item
If there is not enough meaningful battery data, the health block may be hidden.

Important Note : This is the frontend widget logic for UI health visualization; Report generation on the backend uses a separate health-calculation path and can produce different results by design.

[Azbests]

Some feedback on new test builds: 0.2.0.Int.Testing.v.24-amd64. Tried scenario where I have two of same model APC Back-UPS added to VM via USB-passthrough for use with docker. I am not sure if auto config part handles them correctly 🤔Ended up changing ups.conf file and only then second ups started getting detected correctly.

First I think that UPS0 was added twice. the one above and two commented bellow. Over all it seems that for successful setup it needed serial= fields to differentiate UPS'es correctly. I just added some of other values so it would be more similar to one that was auto generated.

[root@nutify nut]# cat ups.conf
[ups0]
  driver = "usbhid-ups"
  port = "auto"
  vendorid = "051D"
  productid = "0002"
  desc = "ups0"
  user = nut
  group = nut
  product = "Back-UPS BGM2200B-GR FW:481600G5-481700G2"
  serial = "xxxxxxxxxxxxx0"
  vendor = "American Power Conversion"

[ups1]
  driver = "usbhid-ups"
  port = "auto"
  vendorid = "051D"
  productid = "0002"
  desc = "ups1"
  pollinterval = 1
  pollfreq = 1
  user = nut
  group = nut
  product = "Back-UPS BGM2200B-GR FW:481600G5-481700G2"
  serial = "xxxxxxxxxxxxx1"
  vendor = "American Power Conversion"

# bus = "002"
# device = "003"
# busport = "002"
# WARNING: The bus, device and busport parameters are commented out to prevent reconnection issues.
# When USB devices are disconnected and reconnected, these values often change,
# which can cause NUT to fail finding the UPS. Leave them commented for better reliability.
#[ups0]
#    driver = "usbhid-ups"
#    port = "auto"
#    desc = "Back-UPS BGM2200B-GR FW:481600G5-481700G2 American Power Conversion S/N:xxxxxxxxxxxxx0"
#    pollinterval = 1
#    pollfreq = 1
#    user = nut
#    group = nut
#
#[ups1]
#    driver = "usbhid-ups"
#    port = "auto"
#    pollinterval = 1
#    pollfreq = 1
#    user = nut
#    group = nut

Also nice upcoming updated with multi-ups support. Thanks for sharing this NUT enhancement project of yours!

Also I can re-test/test things, if there is any interest for this specific setup.

[DartSteven]

Please try this dartsteven/nutify-internal-testing:0.2.0.Int.Testing.v.31-amd64 only amd64 now. Fixes wizard config files :)

[Azbests]

hmm. I still get config generated like this. Copied while still in setup wizard's conf preview tabs

edit: adding some more printout

Mode: Standalone (Local UPS only)
Monitoring Profile: Multi Monitor
Fleet Topology: Local Targets Only
Derived Host Mode: standalone
UPS Targets: 2

[ups0]
driver = "usbhid-ups"
port = "auto"
vendorid = "051D"
productid = "0002"
product = "Back-UPS BGM2200B-GR FW:481600G5-481700G2"
serial = "xxxxxxxxxxxxx0"
vendor = "American Power Conversion"
# bus = "002"
# device = "003"
# busport = "002"
# WARNING: The bus, device and busport parameters are commented out to prevent reconnection issues.
# When USB devices are disconnected and reconnected, these values often change,
# which can cause NUT to fail finding the UPS. Leave them commented for better reliability.
[ups0]
    driver = "usbhid-ups"
    port = "auto"
    desc = "Back-UPS BGM2200B-GR FW:481600G5-481700G2 American Power Conversion S/N:xxxxxxxxxxxxx0"
    pollinterval = 1
    pollfreq = 1
    user = nut
    group = nut

[ups1]
    driver = "usbhid-ups"
    port = "auto"
    desc = "ups1"
    pollinterval = 1
    pollfreq = 1
    user = nut
    group = nut

[DartSteven]

In the wizard, when you choose between Manual and Auto, if you select Manual from the dropdown you can pick the correct UPS, add it using Test & Save, and then add a second one.
Does this work as expected? I’m asking because I only have one USB UPS connected to the server, while the other one is connected via LAN. And please always start with a clean environment, without old databases or any other leftover data.

[Azbests]

1. Yes as you had mentioned in one of previous replies. I am testing with data dir deleted each time when i do not need to manually edit that ups.conf conf file.

2. Hmm this approach produces again different results

And usb detection drop down found only single entry:

Conf file that was generated this time:

# Network UPS Tools: ups.conf
# UPS definitions for NUT

[ups0]
    driver = "usbhid-ups"
    port = "auto"
    desc = "Local UPS"
    pollinterval = 1
    pollfreq = 1
    user = nut
    group = nut 
[ups0]
    driver = "usbhid-ups"
    port = "auto"
    desc = "Local UPS"
    pollinterval = 1
    pollfreq = 1
    user = nut
    group = nut

[ups1]
    driver = "usbhid-ups"
    port = "auto"
    desc = "ups1"
    pollinterval = 1
    pollfreq = 1
    user = nut
    group = nut

Maybe some other logs you might be interested in?

[DartSteven]

No thanks, I think I’ve understood the problem. I’ll make a note of it and try to reproduce it on my side. Thanks!