No Access every few days

sw-omit said:
That looks like the still-in-boot screen to me. Once it is done booting, it should have a login-prompt as well as some text saying "connecting to x.x.x.x:8006 to configure" or something among those lines.

EDIT: So it should look something like this once it's done booting:
https://nguvu.org/images/160701-proxmox-login.png (Just the first thing I found)
Click to expand...
Yeah I thought so! I'm going to back-up everything to my NAS and do a reinstall....
 
rvgmofficial said:
check your network my problem was my my router's default gateway
Click to expand...
huh oke? everything is oke otherwise other devices should've had also a problem. I think the problem is with the device. I don't know if you saw my latest screenshot, but it doesn't suppose to look like that. Futhermore, nothing has change on the device or anything else. But know I'm unable to get an HDMT output....I tried everything but won't give me an output of the device.
 
If it doesn't provide any HDMI-output (and you don't have any other screen-outputs on the device, I think you might indeed by out of luck and the device might have cooked itself (literally), given your previous statements.
Only suggestion would be to try a smaller screen. Larger screen = more pixels = more processing to display things (although usually not noticeable much, with a system on its last legs, it might help)
Other then that, best you could try is see if you can attach the drive to some other (linux) system to rescue the data (if you haven't / hadn't made backups yet)
 
sw-omit said:
If it doesn't provide any HDMI-output (and you don't have any other screen-outputs on the device, I think you might indeed by out of luck and the device might have cooked itself (literally), given your previous statements.
Only suggestion would be to try a smaller screen. Larger screen = more pixels = more processing to display things (although usually not noticeable much, with a system on its last legs, it might help)
Other then that, best you could try is see if you can attach the drive to some other (linux) system to rescue the data (if you haven't / hadn't made backups yet)
Click to expand...
I created backups of all the machines and reinstall proxmox.
 
sw-omit said:
If it doesn't provide any HDMI-output (and you don't have any other screen-outputs on the device, I think you might indeed by out of luck and the device might have cooked itself (literally), given your previous statements.
Only suggestion would be to try a smaller screen. Larger screen = more pixels = more processing to display things (although usually not noticeable much, with a system on its last legs, it might help)
Other then that, best you could try is see if you can attach the drive to some other (linux) system to rescue the data (if you haven't / hadn't made backups yet)
Click to expand...
I'm back! After reinstalling proxmox, I still had the issue after 8 days! Then the day after that 2 times in one day!
Yesterday it happend again, so i connected an HDMI cable to my TV again. For now it didn't happen yet, but I saw the following message:

The e1000e caught my eye, because that's my NIC, I don't know what this means....
1723913130802.jpeg

The NIC error was today, didn't effect anything. But I also had a ProxMox kernel: tpm tpm0: A TPM error (256) occurred attempting get random error yesterdag....the whole day. I disabled TPM yesterday in the BIOS
 

Attachments

Last edited:
Hey!
Did you finally figure out the problem?
Just facing the same which is really really annoying.
System gets unresponsive and all VMs get unaccessible. No SSH or Web-GUI to proxmox or VMs possible but proxmox ping still available. It is happening very randomly. Sometimes it is running for few days with no issue, sometimes it occurs multiple times a day…
 
Last edited:
d3sp3rados13 said:
Hey!
Did you finally figure out the problem?
Just facing the same which is really really annoying.
System gets unresponsive and all VMs get unaccessible. No SSH to proxmox or VMs possible but proxmox ping still available. It is happening very randomly. Sometimes it is running for few days with no issue, sometimes it occurs multiple times a day…
Click to expand...
Nope I did not, I made it work with a workaround. I bought a Ethernet to USB connector and never had any issue ever again. I still running without every disconnected.
 
  • Like
Reactions: d3sp3rados13
LazeX7 said:
Nope I did not, I made it work with a workaround. I bought a Ethernet to USB connector and never had any issue ever again. I still running without every disconnected.
Click to expand...
Thanks for that quick reply!
Sad to hear, that the problem still isn't solved yet and this really drives me crazy. Already tested all the other hardware and that looked fine.I had installed ESXi before on that same machine and it did run flawlessly with no issues for years...

But I will also try the thing with the USB-Ethernet connector and hope this will do the trick...
 
Exact same issue here with Proxmox 8.4 in an ASUS NUC 15 Pro and Unify Dream Router 7. I have it on a smart plug and notice power increasing from 10W to 30W and then staying getting stuck at 30W indefinitely. This is where I lose access to the Proxmox VE and all LXCs, even though I can ping them. I have also enabled backups. I haven't looked at the logs yet, will update when I do.
 

Attachments

  • Screenshot_20250519-040111.png
    Screenshot_20250519-040111.png
    105.7 KB · Views: 11
Subscribing to this.

I also have a HP Elite Mini (don't have all the details on hand at the moment) and I'm experiencing almost everything OP has experienced with his build on this first page.

I'm going to use this thread for some of the tips that were provided. I thought it was a headless issue (apparently linux? doesn't like monitor-less machines operating), so I added one of those HDMI adaptor that acts like a monitor, but the issue still continues.

When I lose connection to the machine that's hosting proxmox, I cannot connect to proxmox nor any of the services I have via VM/Container through web UI. I'll try pinging them though. However, when I go to my equipment (I've recently added a mini monitor to my proxmox machine), the machine is up and operating fine, at least at a glance.

Anyway, all of this vague on my part, hopefully I'll have time today or tomorrow to look into this and resolve it once and for all.
 
_gabriel said:
try disabling offloading :
ethtool -K enXX tso off gso off
Replace enXX with your real NIC interface name.
Click to expand...
Since I've been struggling with the same issue for a while since some update, I also tried the same setting and observed it. Unfortunately, it doesn't solve the problem. At this point, I'm not even sure if it's actually the network card's fault. It also often happens when backing up to internal disks. Strangely, I have identical machines that aren't affected at all. I just have the feeling it's due to system load.
 
Last thing I tried to find what's causing this issue was to try and monitor its temperature to figure out if the nvme ssd overheats for some reason, because when I lose access I get some EXT4 filesystem errors (see attached logs). I did some tests where I downloaded some huge (60GB) files through one container, noticed the temperatures increasing, but as soon as the file was downloaded they went back to normal.

I now have 15 days uptime and hadn't had the issue since and I'm thinking if the script I have set up to monitor the ssd temperatures is actually preventing the ssd from going into some kind of deep power saving state. My SSD is a Samsung 990 Pro SSD - 1TB and the command I have running every 5" in the background is nvme smart-log /dev/nvme0. Now I understand this might be a coincidence and the issue might return, but maybe if more people give it a go we have verify if it's the solution easier. Then we could find a proper way to prevent deep power-saving states (e.g.
GRUB_CMDLINE_LINUX_DEFAULT="quiet nvme_core.default_ps_max_latency_us=0" in /etc/default/grub). This will probably increase the idle power consumption a bit, but ensure stability in the system.
 

Attachments

  • Screenshot_20250821-194352~2.jpg
    Screenshot_20250821-194352~2.jpg
    481.9 KB · Views: 4
  • Screenshot_20250906-103311.png
    Screenshot_20250906-103311.png
    211 KB · Views: 4
Last edited:
werty89 said:
Last thing I tried to find what's causing this issue was to try and monitor its temperature to figure out if the nvme ssd overheats for some reason, because when I lose access I get some EXT4 filesystem errors (see attached logs). I did some tests where I downloaded some huge (60GB) files through one container, noticed the temperatures increasing, but as soon as the file was downloaded they went back to normal.

I now have 15 days uptime and hadn't had the issue since and I'm thinking if the script I have set up to monitor the ssd temperatures is actually preventing the ssd from going into some kind of deep power saving state. My SSD is a Samsung 990 Pro SSD - 1TB and the command I have running every 5" in the background is nvme smart-log /dev/nvme0. Now I understand this might be a coincidence and the issue might return, but maybe if more people give it a go we have verify if it's the solution easier. Then we could find a proper way to prevent deep power-saving states (e.g.
GRUB_CMDLINE_LINUX_DEFAULT="quiet nvme_core.default_ps_max_latency_us=0" in /etc/default/grub). This will probably increase the idle power consumption a bit, but ensure stability in the system.
Click to expand...

Is this still working for you by chance?

I've been battling this issue for a while now. I was convinced it was the NIC... I did the disable offloading mentioned above and even replaced the NIC altogether. Issue still persists.

But then I read your post, and I'm also running a Samsung 990 1TB in my machine, so now I'm really curious if I've been chasing down the wrong device.

When I get home to reboot it, I'm going to try fwupdmgr as soon as it comes up to see if it has anything for the NVME.

Anyway, would appreciate any follow up notes you may have.
 
You must log in or register to reply here.
Top Bottom