pve-ups

[CrazyWolf13]

Name of the Script

pve-ups

Script Type

CT (LXC Container)

Does this script support arm64?

arm64 not tested

📋 Script Details

a small appliance that shuts Proxmox hosts down when the UPS runs low, configured entirely through a web wizard

[CrazyWolf13]

selfhst/icons#964

[github-actions]

The pve-ups script is ready for testing:

bash -c "$(curl -fsSL https://github.com/community-scripts/ProxmoxVED/raw/main/ct/pve-ups.sh)"

Set a password and configure your Proxmox API token (requires Sys.PowerMgmt privilege) through the first-access setup wizard at http://CT-IP:8080.
Requires a UPS with an SNMP network card (RFC 1628 or a vendor MIB such as APC PowerNet) or a reachable NUT server on TCP port 3493.

Note: This is not in the official repo yet—it's just a dev version! After merging into ProxmoxVE, it will need to be recreated.

Discord testing thread: https://discord.com/channels/1302816934508630047/1538650247243497573

[ffind-dev]

First test-setup with default and advanced settings successful - verbose output without errors.
I'll give it some more test-installations, but so far: good job! :)

[Schello805]

Installation without issues and connection to 2 hosts very easy - testshutdown successful. I use a USV V7. So for me it works fine. Good work :-)

[fnbc]

Smooth installation through community-scripts - using with cyberpower ups. Everything's showing up as expected!

[CrazyWolf13]

@ffind-dev have you hit any roadblocks, or can we move over the script to our main repo?

[ffind-dev]

I've tested on a few Proxmox Hosts and within my dev-simulator and had no issues with the community-script wrapper so far. That's all I can test at the moment.

[Schello805]

By the way, I use a local ntfy installation for all my homelab notifications and the Notification content is... improvable :-)

This is, what came up during the test shutdown.

{"subject":"[PVE-UPS] Test shutdown executed","body":"Test (dry-run): order pve [USV V7, AND], pvemini [USV V7, AND]. NOTHING was shut down.","status":{"appliance":{"version":"3.3.0","uptime_s":99832,"engine_state":"ONLINE","dry_run":true,"config_valid":true,"alarm":false,"last_selftest_at":"2026-08-18T07:00:14.281397+00:00","last_selftest_ok":false,"next_selftest_at":"2026-08-19T09:00:00"},"ups":[{"id":"ups1","name":"USV V7","type":"snmp","mib":"rfc1628","reachable":true,"manufacturer":"\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000","model":"LIHVX1K5\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000","last_poll":"2026-08-18T11:52:54.110445+00:00","poll_interval_s":30,"power_source":"mains","battery_status":"normal","runtime_remaining_min":1,"battery_charge_pct":100,"seconds_on_battery":0,"triggered":false,"trigger_reason":null,"countdown_remaining_s":null,"comm_loss_remaining_s":null,"alarm":false,"error":null}],"shutdown":{"triggered":false,"reason":null,"triggered_at":null,"countdown_remaining_s":null,"comm_loss_remaining_s":null},"hosts":[{"name":"pve","this_host":false,"order":0,"ups_ids":["ups1"],"ups_policy":"all","feeds":[{"id":"ups1","name":"USV V7","triggered":false}],"eligible":false,"pending_reason":null,"reachable":null,"shutdown_state":"idle","last_action_at":null,"last_error":null},{"name":"pvemini","this_host":false,"order":0,"ups_ids":["ups1"],"ups_policy":"all","feeds":[{"id":"ups1","name":"USV V7","triggered":false}],"eligible":false,"pending_reason":null,"reachable":null,"shutdown_state":"idle","last_action_at":null,"last_error":null}]}}

[ffind-dev]

@Schello805 you're using the webhooks for that? That's not meant to be human readable :)

[Schello805]

Thanks, but How Else can I Let me in inform in case of losing power supply?
What is the Normal use Case you Expect?

[ffind-dev]

Currently you should actively query the status api or process webhooks through a specific service to parse the content.
But there are plans to make the webhooks content customizable e.g. for MS Teams, and Plaintext is an idea as well.

I personally think, actively querying the status API with an RMM Tool is the best option as you would notice if it fails. If pve-ups stopped working you would never notice if you're not monitoring it. And it might not be able to notify you by itself when it's unavailable. That's the case for nearly every tool.