Proxmox causing node to shut off - is Proxmox rebooting or is session expiring?

I

illmortal

New Member
Mar 20, 2025
7
0
1
Hello everyone,

I've decided to utilize Proxmox, currently PVE 8.3.5. I added one node (Ubuntu VM) thus far for testing purposes. After successfully getting the node up and running to my liking - a process that took me about 4 hours, I called it a day and put my computer to sleep - the computer I'm using to access Proxmox via Web GUI. The next morning I noticed that the node was offline (not pingable nor was Proxmox). I get back to my computer to find that Proxmox was logged out of the session which in turn shuts my node off.

My Proxmox setup is simply a headless server that I should be able to leave on at all times running 24/7, if it needs to do a reboot, I should make that decision and manually run the reboot command. Instead at this time, for the last 4 days, I keep encountering Proxmox going offline and taking my node offline with it.

So at this time, my assumption is that the session is expiring which is turning off my node, or there are updates/upgrades and reboots occurring without my interaction / intervention which is a big nono.

I need whatever is happening to come to a full stop and operate 24/7/365.

I've checked my BIOS, all is good there.

I've also configured /etc/systemd/logind.conf:

Code: 
HandleLidSwitch=ignore
HandleLidSwitchExternalPower=ignore
HandleSuspendKey=ignore
HandleHibernateKey=ignore


If there's a way to disable session timeout, that would be great. Or a way to make it that Proxmox does not rely on another device having a tab open actively viewing the session, that would be great too.



Please let me know if there's anything I need to provide.
 
Last edited:
Hello!

Just so I am clear, the hardware you installed PVE onto was a laptop? From my experience, uncommenting and setting the HandleLidSwitch=ignore will stop the node from hibernating when you close the lid. You must either reboot the node or run a systemctl restart systemd-logind to have the changes in /etc/systemd/logind.conf take affect.

If this is a repurposed PC, server tower, or rack mounted server you should never really need to set those settings. The PVE node should serve up the GUI 24/7/365 without issues, you should not need to have a tab open or anything like. There shouldn't be any any extra settings you need to change to facilitate this unless you are closing the lid on a laptop with PVE installed on it.

Let me know!
Jonathan
 
weehooey-jf said:
Hello!

Just so I am clear, the hardware you installed PVE onto was a laptop? From my experience, uncommenting and setting the HandleLidSwitch=ignore will stop the node from hibernating when you close the lid. You must either reboot the node or run a systemctl restart systemd-logind to have the changes in /etc/systemd/logind.conf take affect.

If this is a repurposed PC, server tower, or rack mounted server you should never really need to set those settings. The PVE node should serve up the GUI 24/7/365 without issues, you should not need to have a tab open or anything like. There shouldn't be any any extra settings you need to change to facilitate this unless you are closing the lid on a laptop with PVE installed on it.

Let me know!
Jonathan
Click to expand...
Edit: Sorry, I didn't directly answer your question.

I'm using a repurposed HP EliteDesk 800 G4 (mini desktop) to host Proxmox. It was originally a Seedbox and media server. I upgraded the RAM and NVMe for this purpose. Everything works except when the system is idle for a couple of hours; it logs me out or sometimes reboots, requiring a login refresh to access again.


The disconnect/reboot (whatever is stopping Proxmox from functioning consistently) has happened once again.

I can't find it in the log where this is happening other than there are updates that are taking place:


Mar 20 15:42:37 proxmox systemd[1]: Started pvestatd.service - PVE Status Daemon.
Mar 20 15:42:37 proxmox pvedaemon[1088]: starting server
Mar 20 15:42:37 proxmox pvedaemon[1088]: starting 3 worker(s)
Mar 20 15:42:37 proxmox pvedaemon[1088]: worker 1089 started
Mar 20 15:42:37 proxmox pvedaemon[1088]: worker 1090 started
Mar 20 15:42:37 proxmox pvedaemon[1088]: worker 1091 started
Mar 20 15:42:37 proxmox systemd[1]: Started pvedaemon.service - PVE API Daemon.
Mar 20 15:42:37 proxmox systemd[1]: Starting pve-ha-crm.service - PVE Cluster HA Resource Manager Daemon...
Mar 20 15:42:37 proxmox systemd[1]: Starting pveproxy.service - PVE API Proxy Server...
Mar 20 15:42:37 proxmox systemd[1]: systemd-rfkill.service: Deactivated successfully.
Mar 20 15:42:38 proxmox pve-ha-crm[1096]: starting server
Mar 20 15:42:38 proxmox pve-ha-crm[1096]: status change startup => wait_for_quorum
Mar 20 15:42:38 proxmox systemd[1]: Started pve-ha-crm.service - PVE Cluster HA Resource Manager Daemon.
Mar 20 15:42:38 proxmox kernel: e1000e 0000:00:1f.6 eno1: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Mar 20 15:42:38 proxmox kernel: vmbr0: port 1(eno1) entered blocking state
Mar 20 15:42:38 proxmox kernel: vmbr0: port 1(eno1) entered forwarding state
Mar 20 15:42:38 proxmox pveproxy[1097]: starting server
Mar 20 15:42:38 proxmox pveproxy[1097]: starting 3 worker(s)
Mar 20 15:42:38 proxmox pveproxy[1097]: worker 1098 started
Mar 20 15:42:38 proxmox pveproxy[1097]: worker 1099 started
Mar 20 15:42:38 proxmox pveproxy[1097]: worker 1100 started
Mar 20 15:42:38 proxmox systemd[1]: Started pveproxy.service - PVE API Proxy Server.
Mar 20 15:42:38 proxmox systemd[1]: Starting pve-ha-lrm.service - PVE Local HA Resource Manager Daemon...
Mar 20 15:42:38 proxmox systemd[1]: Starting spiceproxy.service - PVE SPICE Proxy Server...
Mar 20 15:42:39 proxmox spiceproxy[1103]: starting server
Mar 20 15:42:39 proxmox spiceproxy[1103]: starting 1 worker(s)
Mar 20 15:42:39 proxmox spiceproxy[1103]: worker 1104 started
Mar 20 15:42:39 proxmox systemd[1]: Started spiceproxy.service - PVE SPICE Proxy Server.
Mar 20 15:42:39 proxmox pve-ha-lrm[1105]: starting server
Mar 20 15:42:39 proxmox pve-ha-lrm[1105]: status change startup => wait_for_agent_lock
Mar 20 15:42:39 proxmox systemd[1]: Started pve-ha-lrm.service - PVE Local HA Resource Manager Daemon.
Mar 20 15:42:39 proxmox systemd[1]: Starting pve-guests.service - PVE guests...
Mar 20 15:42:40 proxmox kernel: fbcon: Taking over console
Mar 20 15:42:40 proxmox pve-guests[1107]: <root@pam> starting task UPID:proxmox:00000456:00000D29:67DC6FB0:startall::root@pam:
Mar 20 15:42:40 proxmox pve-guests[1107]: <root@pam> end task UPID:proxmox:00000456:00000D29:67DC6FB0:startall::root@pam: OK
Mar 20 15:42:40 proxmox systemd[1]: Finished pve-guests.service - PVE guests.
Mar 20 15:42:40 proxmox systemd[1]: Starting pvescheduler.service - Proxmox VE scheduler...
Mar 20 15:42:41 proxmox pvescheduler[1113]: starting server
Mar 20 15:42:41 proxmox systemd[1]: Started pvescheduler.service - Proxmox VE scheduler.
Mar 20 15:42:41 proxmox systemd[1]: Reached target multi-user.target - Multi-User System.
Mar 20 15:42:41 proxmox systemd[1]: Reached target graphical.target - Graphical Interface.


Mar 20 15:42:41 proxmox systemd[1]: Starting systemd-update-utmp-runlevel.service - Record Runlevel Change in UTMP...
Mar 20 15:42:41 proxmox systemd[1]: systemd-update-utmp-runlevel.service: Deactivated successfully.
Mar 20 15:42:41 proxmox systemd[1]: Finished systemd-update-utmp-runlevel.service - Record Runlevel Change in UTMP.
Mar 20 15:42:41 proxmox systemd[1]: Startup finished in 24.131s (kernel) + 10.354s (userspace) = 34.485s.
Mar 20 15:43:15 proxmox chronyd[860]: Selected source 50.205.57.38 (2.debian.pool.ntp.org)
Mar 20 15:43:15 proxmox chronyd[860]: System clock TAI offset set to 37 seconds



Mar 20 15:55:24 proxmox pvedaemon[1089]: <root@pam> successful auth for user 'root@pam'
Mar 20 15:57:13 proxmox systemd[1]: Starting systemd-tmpfiles-clean.service - Cleanup of Temporary Directories...
Mar 20 15:57:13 proxmox systemd[1]: systemd-tmpfiles-clean.service: Deactivated successfully.
Mar 20 15:57:13 proxmox systemd[1]: Finished systemd-tmpfiles-clean.service - Cleanup of Temporary Directories.
Mar 20 15:57:13 proxmox systemd[1]: run-credentials-systemd\x2dtmpfiles\x2dclean.service.mount: Deactivated successfully.
Mar 20 16:11:24 proxmox pvedaemon[1090]: <root@pam> successful auth for user 'root@pam'
Mar 20 16:12:34 proxmox smartd[690]: Device: /dev/sda [SAT], SMART Usage Attribute: 194 Temperature_Celsius changed from 154 to 151
Mar 20 16:16:34 proxmox smbd[6700]: pam_unix(samba:session): session opened for user all(uid=1000) by (uid=0)
Mar 20 16:16:59 proxmox pvedaemon[6782]: starting termproxy UPID:proxmox:00001A7E:00033185:67DC77BB:vncshell::root@pam:
Mar 20 16:16:59 proxmox pvedaemon[1091]: <root@pam> starting task UPID:proxmox:00001A7E:00033185:67DC77BB:vncshell::root@pam:
Mar 20 16:16:59 proxmox pvedaemon[1090]: <root@pam> successful auth for user 'root@pam'
Mar 20 16:16:59 proxmox login[6789]: pam_unix(login:session): session opened for user root(uid=0) by (uid=0)
Mar 20 16:16:59 proxmox systemd[1]: Created slice user-0.slice - User Slice of UID 0.
Mar 20 16:16:59 proxmox systemd[1]: Starting user-runtime-dir@0.service - User Runtime Directory /run/user/0...
Mar 20 16:16:59 proxmox systemd-logind[693]: New session 1 of user root.
Mar 20 16:16:59 proxmox systemd[1]: Finished user-runtime-dir@0.service - User Runtime Directory /run/user/0.
Mar 20 16:16:59 proxmox systemd[1]: Starting user@0.service - User Manager for UID 0...
Mar 20 16:16:59 proxmox (systemd)[6795]: pam_unix(systemd-user:session): session opened for user root(uid=0) by (uid=0)
Mar 20 16:16:59 proxmox systemd[6795]: Queued start job for default target default.target.
Mar 20 16:16:59 proxmox systemd[6795]: Created slice app.slice - User Application Slice.
Mar 20 16:16:59 proxmox systemd[6795]: Reached target paths.target - Paths.
Mar 20 16:16:59 proxmox systemd[6795]: Reached target timers.target - Timers.
Mar 20 16:16:59 proxmox systemd[6795]: Listening on dirmngr.socket - GnuPG network certificate management daemon.
Mar 20 16:17:00 proxmox systemd[6795]: Listening on gpg-agent-browser.socket - GnuPG cryptographic agent and passphrase cache (access for web browsers).
Mar 20 16:17:00 proxmox systemd[6795]: Listening on gpg-agent-extra.socket - GnuPG cryptographic agent and passphrase cache (restricted).
Mar 20 16:17:00 proxmox systemd[6795]: Listening on gpg-agent-ssh.socket - GnuPG cryptographic agent (ssh-agent emulation).
Mar 20 16:17:00 proxmox systemd[6795]: Listening on gpg-agent.socket - GnuPG cryptographic agent and passphrase cache.


4:16PM is when I approached my computer, took it out of sleep, and saw that Proxmox once again was logged off with my node offline.


Edit: Here's a pastebin of a portion of the log - I don't know what other keywords to look for other than reboot.

https://pastebin.com/1cRv7jvg
 
Last edited:
Bump!

Anyone have any ideas as to what's causing Proxmox to take my node offline? I'm at work at the moment, but before leaving my home, I tried pinging Proxmox and once again it's offline. I'll have to wait til I get home to touch it again - what should I immediately look for to show what triggered my Proxmox to go down/offline/reboot?

I'm willing to scrap the entire installation I have now and start from scratch. I just need a guide where during or after the installation process I need to configure Proxmox to stay online 24/7 and keep my nodes online 24/7. With this current installation, out of the box, Proxmox is not staying online for more than at least 2 hours.
 
Last edited:
MarkusKo said:
does the node power off or crash?

if it crashes:
https://forum.proxmox.com/threads/p...th-hardware-transcoding-in-plex.132187/page-4

https://www.reddit.com/r/homelab/comments/1baoy0y/fresh_proxmox_install_on_hp_elitedesk_800_g4_mini/
Click to expand...
That's the current challenge I'm facing. I have no issues pulling logs, but I can't find anywhere in the logs what's happening, for example the possibility of a crash happening. And because there's so many lines in the logs, I keep using the Find feature for the keyword, "reboot". Assuming a reboot is taking place.

This issue has yet to occur while I'm in front of my PC to witness it. The only thing I've witnessed is Proxmox session timing out / ending and when I noticed it, my node was offline. Before I logged back into Proxmox through the web GUI, I attempted to ping my node but it was offline. After logging back into Proxmox, I went straight to the system logs where I had discovered, "reboot" and that threw a red flag. So it's one of two things, I assume, Proxmox disconnects my node when the session expires or Proxmox is rebooting without my consent. I can't imagine a sessions expiring would also take down nodes but that's what I had witnessed and if that's the case, I need "session expiration" to be non-existent. If it's rebooting, I need to know why it's rebooting so that I can't put a stop to it.

Thanks for providing the two links, also. I'm going to review them.

Curious, when I get home, as soon as I log back into Proxmox, what log should I look into? And what keywords should I search/find for?

In a Linux environment, at least from my experience, when errors exist in the log, it's usually red text. My System Log, the last 4 times I've checked it, hasn't showed a visual indicator of errors, I would assume it would but maybe that feature doesn't exist - hence then using the Find feature.


Edit: Oh jees.. I remember I had to purchase a DisplayPort headless ghost display emulator (yeah I had to Google it to make sure I had it right) just to run my Ubuntu Server on this machine. So I'll have to re-add that adaptor and then test again. That might actually be the the issue at hand. Interestingly with Ubuntu Server, the side effect of not having a display attached to the machine (headless) was not being able to VNC into the machine. It didn't randomly reboot or anything like that, it just wasn't accessible via VNC, the connection would get rejected.
 
Last edited:
it is normal that the web ui session expires, thats a security feature :)

if the node is offline -> it hangs or crashes
if it would reboot you could ping it again after some time but thats not the case
 
MarkusKo said:
it is normal that the web ui session expires, thats a security feature :)

if the node is offline -> it hangs or crashes
if it would reboot you could ping it again after some time but thats not the case
Click to expand...
Correct, whatever happens to take Proxmox offline, it does not allow me to ping Proxmox until I go to the Web GUI (sometimes refresh the page) and then log in. After logging in, I have to go to the node and manually start it. The machine that Proxmox is installed onto, while Proxmox is in this offline status, is on and appears to be operating just fine, I can even see there's network activity (blinking green light).

I edited my last response, I now have a feeling that it's the headless issue. Now that I'm thinking about it, while I was setting up Proxmox I had my portable 10" monitor attached to it, but I was also very active on that machine/Proxmox for 4 or so hours as I was setting it up and getting my first node up and running. I didn't experience any random crashes/reboots only a couple times that the session had timed out while I was away.

I have a DP emulator that I had used previously on this machine when I was running Ubuntu Server, I'll just re-add it to the machine and play the waiting game again.
 
Update! Looks like that was the issue all along. This specific computer the HP Mini series appears to require a monitor signal? It will not work in headless mode without adding an adaptor that mimics a monitor connection.

Thanks for providing the links - hope this helps someone else later on down the road.
 
You must log in or register to reply here.
Top Bottom