Skip to content

Latest commit

 

History

History
300 lines (212 loc) · 6.97 KB

README.md

File metadata and controls

300 lines (212 loc) · 6.97 KB
Raw

Integrating Salicru SPS SOHO+ UPS on Proxmox VE and Home Assistant Core

Table of Contents

Scope and objective

This procedure depicts all the steps made on an attempt to fully integrate these cheap UPS in Proxmox VE and Home Assistant Core running on it.

The model tested was a Salicru SPS SOHO+ 500 VA, as such:

salicru_ups

Requirements

  • Proxmox VE 6.x
  • Home Assistant Core 0.112.4
  • USB port free on your hardware

Proxmox VE integration

Connecting the UPS to the Proxmox hardware

Use the UPS USB Type B port (on the back!) and the provided USB cable to connect it straight into another free USB port, after powering up the UPS on the front power button.

Detecting the new USB device from the UPS

Run in the Proxmox shell, through its web GUI or SSH console:

# lsusb

You should see a similar line within the response:

Bus 001 Device 028: ID 06da:ffff Phoenixtec Power Co., Ltd

If not, check the USB connection or use a different USB port on your hardware.

Installing the NUT server and client

Run:

# apt update && apt install nut -y

Configuring both NUT services

If does not exist, create the following file and its contents:

# nano /etc/nut/ups.conf

maxretry = 3
[salicru]
driver = usbhid-ups
port = auto
desc = "Salicru"
# 15 seconds between each polling
pollinterval = 15

Create the following rule that allows the nut user to access the UPS driver later on:

# nano /etc/udev/rules.d/90-nut-ups.rules

# Rule for a Salicru UPS
ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="06da", ATTR{idProduct}=="ffff", MODE="0660", GROUP="nut"

Restart the udev to apply it:

# service udev restart

Next, replug the USB cable into the hardware USB port, waiting a couple of seconds between.

Then, paste the following contents inside:

# nano /etc/nut/nut.conf

MODE=standalone

Edit the file and adjust accordingly:

# nano /etc/nut/upsd.conf

# LISTEN <address> [<port>]
LISTEN 0.0.0.0 3493
# This will cause the NUT server service to be listening all remote IP addresses, among the Proxmox localhost. Use it wisely.

Set the monitor username (upsmonitor) and its password (YOUR_STRONG_PASSWORD). Use a strong one!

# nano /etc/nut/upsd.users

[upsmonitor]
password = YOUR_STRONG_PASSWORD
upsmon master

As the last file to edit:

# nano /etc/nut/upsmon.conf

# Commands for shutdown on power loss
MONITOR salicru@THE_PROXMOX_LAN_IP_OR_LOCALHOST 1 upsmonitor YOUR_STRONG_PASSWORD_HERE master
POWERDOWNFLAG /etc/killpower
SHUTDOWNCMD "/sbin/shutdown -h now"

Set permissions correctly:

# chown root:nut /etc/nut/*

and

# chmod 640 /etc/nut/*

Enabling and starting the NUT services

Set them as enabled, both server and client services:

# systemctl enable nut-server.service

and:

# systemctl enable nut-client.service

Finally, start both:

# service nut-server start

and

# service nut-client start

If you are confronted with errors, check them through:

# journalctl -xe

Home Assistant Core integration

Configuring the NUT client integration

Go to your Home Assistant instance and open the following path:

Configuration -> Integrations -> + sign (at the buttom-right side) -> search for NUT

Fill the correct parameters:

home_assistant_nut_client_config_options

After all the steps, you should find a proper configured integration like this:

home_assistant_nut_client_configured

Setting up a lovelace card for all the entities

Finally, create a manual lovelace card with this proposal:

entities:
  - sensor.salicru_battery_runtime
  - sensor.salicru_battery_charge 
  - sensor.salicru_battery_voltage
  - sensor.salicru_load
  - sensor.salicru_status
  - sensor.salicru_status_data
  - sensor.salicru_battery_chemistry
type: entities
title: Salicru SPS SOHO+ 500 VA

Should be similar to this:

lovelace_card

Automating the shutdown process

In case of doubt, you can setup a safe guard, using Home Assistant, to monitor the UPS in discharge and trigger a Proxmox VE shutdown, on low battery level.

Automations.yaml

automation:
- id: 'upsdischarged'
  alias: 'UPS discharged - Start Shutdown'
  trigger:
    - platform: numeric_state
      entity_id: sensor.salicru_battery_charge
      below: 20
      for:
        seconds: 5
    - platform: numeric_state
      entity_id: sensor.salicru_battery_runtime
      below: 240
  condition:
    condition: and
    conditions:
    - condition: template
      value_template: >
        {{ not(is_state('sensor.salicru_status_data','OL CHRG')) }}
  action:
    - service: shell_command.proxmox_shutdown_host_pve

Configuration.yaml

shell_command:
 proxmox_shutdown_host_pve: !secret proxmox_shutdown_host_pve

Secrets.yaml

proxmox_shutdown_host_pve: bash /config/shell_scripts/proxmox_shutdown_host_pve.bash IP_PROXMOX USER_WITH_SYS.POWERMGMT PASSWORD NODE_NAME

/config/shell_scripts/proxmox_shutdown_host_pve.bash

#!/bin/bash 

APINODE=$1
USER=$2
PASSWORD=$3
NODE=$4

curl --silent --insecure --data "username=${USER}@pve&password=${PASSWORD}" https://$APINODE:8006/api2/json/access/ticket | jq --raw-output '.data.ticket' | sed 's/^/PVEAuthCookie=/' > cookie
curl --silent --insecure --data "username=${USER}@pve&password=${PASSWORD}" https://$APINODE:8006/api2/json/access/ticket | jq --raw-output '.data.CSRFPreventionToken' | sed 's/^/CSRFPreventionToken:/' > csrftoken
curl --silent --insecure  --cookie "$(<cookie)" --header "$(<csrftoken)" -X POST https://$APINODE:8006/api2/extjs/nodes/$NODE/status --data 'command=shutdown'
rm cookie csrftoken

Reboot your Home Assistant instance.

Have fun! :-)

References