[Bug] VMs list of ESXI empties

[Nekima]

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

Approximately a few hours after adding ESXi hosts, the VM list empties and the connection test fails, simply re-entering the root password resolves the issue for a few hours before it malfunctions again.

Steps to Reproduce

1. Add an ESXI Host
2. Wait ~2h
3. The VMs list empties

Expected behavior

Re-entering the root password of the ESXI

Environment

• PegaProx Version: Latest
• Installation Method: LXC
• OS: Debian 13
• Browser: Brave latest
• Behind Reverse Proxy? No
• UI? Modern

Logs

Screenshots

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]

Hey @Nekima — thanks for the report. Your description (works, then empties out after ~2h, fixed by re-entering the password) is textbook pyvmomi session expiry: the vSphere web session ticket has a default idle timeout of about 30 min on most setups, some go up to 2h. Once the ticket goes stale, every follow-up query returns an empty list instead of an auth error, which is why the UI looks like the cluster "forgot" its VMs.

We do refresh the connection on a timer, but there may be a path where the background refresh loses the ticket and we keep using a dead session instead of re-authenticating. To confirm that's what's happening, could you:

1. Open the PegaProx server logs (in the container: docker logs <container> or journalctl -u pegaprox on a host install) around the time the list empties. Look for lines containing ESXi, vSphere, pyvmomi, or session.
2. On the ESXi side, check the hostd log (/var/log/hostd.log) at the same timestamp — a "session not authenticated" or "403" entry would confirm the ticket died.

If we see the ticket-expiry pattern it's a straightforward fix on our side — bump the session refresh to proactively re-auth before the ESXi-side timeout. Self-assigning.

[Nekima]

Hello, thanks for the quick reply,

In the journalctl, i don't have logs with lines containing ESXI, etc

In the ESXI side, i have this when i trying to refresh in Pegaprox:
2026-04-23T07:08:59.891Z Db(167) Hostd[6826052]: [Originator@6876 sub=Solo.Vmomi opID=2cebf9d9 sid=52313cad] Arg container:
2026-04-23T07:08:59.891Z Db(167) Hostd[6826031]: --> 'vim.Folder:ha-folder-root'
2026-04-23T07:08:59.891Z Db(167) Hostd[6826052]: [Originator@6876 sub=Solo.Vmomi opID=2cebf9d9 sid=52313cad] Arg type:
2026-04-23T07:08:59.891Z Db(167) Hostd[6826031]: --> (vmodl.TypeName) [
2026-04-23T07:08:59.891Z Db(167) Hostd[6826031]: --> "vim.VirtualMachine"
2026-04-23T07:08:59.891Z Db(167) Hostd[6826031]: --> ]
2026-04-23T07:08:59.891Z Db(167) Hostd[6826052]: [Originator@6876 sub=Solo.Vmomi opID=2cebf9d9 sid=52313cad] Arg recursive:
2026-04-23T07:08:59.891Z Db(167) Hostd[6826031]: --> true
2026-04-23T07:08:59.891Z In(166) Hostd[6826052]: [Originator@6876 sub=Solo.Vmomi opID=2cebf9d9 sid=52313cad] Throw vim.fault.NotAuthenticated
2026-04-23T07:08:59.891Z In(166) Hostd[6826052]: [Originator@6876 sub=Solo.Vmomi opID=2cebf9d9 sid=52313cad] Result:
2026-04-23T07:08:59.891Z In(166) Hostd[6826031]: --> (vim.fault.NotAuthenticated) {
2026-04-23T07:08:59.891Z In(166) Hostd[6826031]: --> object = 'vim.Folder:ha-folder-root',
2026-04-23T07:08:59.891Z In(166) Hostd[6826031]: --> privilegeId = "System.View",
2026-04-23T07:08:59.891Z In(166) Hostd[6826031]: --> msg = "",
2026-04-23T07:08:59.891Z In(166) Hostd[6826031]: -->

[Nekima]

I just noticed that when I restart the Pegaprox LXC container, all the ESXi hosts appear without having to re-enter the password

[Nekima]

The bug has been fixed by updating to the latest version, thanks!