[Bug] Navigation/Visual Element: When attempting clone a VM the selection of target node appears to fluctuate with auto-refresh activity

[KowMangler]

Hey there! 👋 Thanks for taking the time to report this bug — we appreciate it!

Please keep in mind: PegaProx is developed and maintained entirely by volunteers in our free time. We do our best, but we can't work magic 🪄 — please be patient, we'll get to your issue as soon as we can.

Want to help keep PegaProx alive?

• ⭐ Star the project — it's free and helps a lot!
• 💖 Become a Sponsor — helps us dedicate more time to development
• 🤝 Contributing code or docs is always welcome too!

---

Describe the bug

When trying to clone a VM the node selection drop down seems to reorder the node list automatically based on the cpu/mem refresh rates in the main window. (probably due to current cpu/mem constraints)

Steps to Reproduce

1. Have an active cluster where the least used node list would fluctuate rapidly
2. Right click a VM
3. Select clone
4. Expand the node drop-down

Expected behavior

The node drop-down list should remain stable

Environment

• PegaProx Version: PegaProx Beta 0.9.10
• Installation Method: script
• OS: Debian 13
• Browser: Google Chrome 148.0.7778.167
• Behind Reverse Proxy? NO
• UI? Corporate

Logs

Screenshots

This one is difficult to describe so I made an animated GIF.

https://gifyu.com/image/b7a5r

Checklist

• [ X] I have searched existing issues to make sure this is not a duplicate
• [ X] I am using the latest version of PegaProx

[mkellermann97]

Yeah this one's a React rendering anti-pattern — the clone modal is reading the live node list from the polling-driven cluster state, and the upstream list is sorted by current load (lowest first), so every poll tick the order shifts. Dropdown needs to snapshot the order on open or sort by a stable key (node name).

Smallish UI fix, queued. Doesn't block anything functional — the option you pick still lands correctly on the right node, just the visual jitter is annoying.

Donation-funded — opencollective.com/pegaprox if PegaProx earns its keep in your homelab.

— MK

[KowMangler]

@mkellermann97 I've only had it installed for 2 days and am just reporting things i come across. Hard to fix anything if you don't know about it haha. I like where this product appears to be headed but i have a need for scale. And I mean 5000-10000 proxmox nodes total across over 1000 clusters in scattered geo locations. Running through typical workflows i would have is how I'm finding all this. I haven't even truly started digging into this product and have only really scratched the surface of what i intend to test.

[MrMasterbay]

@KowMangler
Thanks for reporting!
Honest heads-up since you just started testing: we're 3 people doing this in our free time without any big budget and also in an open Beta, Proxmox changes their API a lot, and we're currently also patching dependency CVEs so response times vary. Bug reports are still very welcome though, and we'll get to this one.

Appreciate the patience in advance.

[mkellermann97]

Hey @KowMangler — picked this one up on Saturday-morning patrol.

Fix landed on Testing (commit 14e5da7): the CloneVmModal now snapshots the node list once at mount via a useMemo with empty deps. Whatever ordering the parent dashboard happened to pass in at the moment you clicked Clone gets re-sorted alphabetically by node name and frozen for the lifetime of the modal. Dashboard's cpu/mem polling keeps running for the main view, but the dropdown stays put while you're using it. No more "the entry I was about to click jumped two slots down."

No translation work needed (no new strings — just sort logic). Verified the new stableNodes symbol lands in the built HTML and is served live by the dev instance.

Donations at https://opencollective.com/pegaprox if Saturday-morning fix sessions on minor paper-cuts earn their keep — totally voluntary.

— Marcus

[mkellermann97]

Hey @KowMangler — v0.9.10.3 shipped: https://github.com/PegaProx/project-pegaprox/releases/tag/v0.9.10.3

The Clone VM dropdown stability fix is in. After upgrading, opening the Clone modal on a cluster with fluctuating per-node load will snapshot the target-node list at mount via useMemo with an empty deps array — alphabetic sort, frozen for the modal's lifetime. The dashboard's cpu/mem auto-refresh keeps running for the main view but the dropdown stays put while you're picking from it.

Closing as completed. If the dropdown still reshuffles in a way you wouldn't expect (e.g. on a re-mount triggered by something specific), drop a fresh ticket with the trigger and we'll trace that separately.

Donations at https://opencollective.com/pegaprox if frozen-snapshot-at-mount UX patterns earn their keep.

— Marcus