Repository navigation
pve-ups #2190
Description
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
Activity
CrazyWolf13 commentedon Aug 16, 2026
selfhst/icons#964
github-actions commentedon Aug 16, 2026
The pve-ups script is ready for testing:
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 commentedon Aug 17, 2026
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 commentedon Aug 19, 2026
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 commentedon Aug 19, 2026
Smooth installation through community-scripts - using with cyberpower ups. Everything's showing up as expected!
CrazyWolf13 commentedon Aug 19, 2026
@ffind-dev have you hit any roadblocks, or can we move over the script to our main repo?
ffind-dev commentedon Aug 19, 2026
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 commentedon Aug 19, 2026
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 commentedon Aug 19, 2026
@Schello805 you're using the webhooks for that? That's not meant to be human readable :)
Schello805 commentedon Aug 19, 2026
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 commentedon Aug 19, 2026
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.
Schello805 commentedon Aug 19, 2026
Thanks for the reply. Of course, you can monitor the container itself using Uptime Kuma or similar tools, but currently, only my smart plug detects a power outage. From that moment on, you have X minutes or hours to react. Those without a smart plug could receive a notification from the UPS. However, if that isn't possible (as is the case for me), it would be great if pve-ups could send a notification so that action can still be taken.
For me, even an unreadable message would suffice as an emergency alert. Though, naturally, a readable message would be nicer :-)
ffind-dev commentedon Aug 19, 2026
@Schello805 will be available with version 3.4.0:
ffind-dev commentedon Aug 19, 2026
@CrazyWolf13 i just tested "update" through community-script command as well. Looking good. So we have this tested at least once.
Schello805 commentedon Aug 19, 2026
Really cool. Thanks 😊🙏
ffind-dev commentedon Aug 21, 2026
@CrazyWolf13 i've tested updates on some more appliances. No roadblocks - what's needed to move on? Or how much feedback do you usually wait for?
github-actions commentedon Aug 21, 2026
A PR has been created for pve-ups: community-scripts/ProxmoxVE#16670
michelroegl-brunner commentedon Aug 22, 2026
Merged with #16670 in ProxmoxVE
github-actions commentedon Aug 22, 2026
Files deleted with PR #2199