Web Interface

[Johann]

Hello! I'm new to the Proxmox community and I have Proxmox 6 installed. The web interface has worked fine for a month now and suddenly I can't access it from 8006 anymore. I am still able to ssh to my 2 nodes using their ips.

Thanks

 

[sg90]

Try

"service pveproxy restart"

If not check the output of "service pveproxy status"

 

[Johann]

Running the status command gets me this output:

Loaded: loaded (/lib/systemd/system/pveproxy.service; enabled; vendor preset:
Active: active (running) since Sat 2019-12-28 18:43:59 EST; 1 day 6h ago
Process: 3382 ExecStart=/usr/bin/pveproxy start (code=exited, status=0/SUCCESS
Process: 24972 ExecReload=/usr/bin/pveproxy restart (code=exited, status=0/SUC
Main PID: 3386 (pveproxy)
Tasks: 4 (limit: 4915)
Memory: 18.6M
CGroup: /system.slice/pveproxy.service
├─ 3386 pveproxy
├─25221 pveproxy worker
├─25222 pveproxy worker
└─25223 pveproxy worker

Dec 29 00:00:36 pve pveproxy[3386]: starting 3 worker(s)
Dec 29 00:00:36 pve pveproxy[3386]: worker 25221 started
Dec 29 00:00:36 pve pveproxy[3386]: worker 25222 started
Dec 29 00:00:36 pve pveproxy[3386]: worker 25223 started
Dec 29 00:00:43 pve pveproxy[3388]: worker exit
Dec 29 00:00:43 pve pveproxy[3387]: worker exit
Dec 29 00:00:43 pve pveproxy[3389]: worker exit
Dec 29 00:00:46 pve pveproxy[3386]: worker 3387 finished
Dec 29 00:00:46 pve pveproxy[3386]: worker 3389 finished

I tried accessing the WEB ui from safari and chrome and Firefox on a Mac and edge on windows

 

[sg90]

Running the status command gets me this output:

Loaded: loaded (/lib/systemd/system/pveproxy.service; enabled; vendor preset:
Active: active (running) since Sat 2019-12-28 18:43:59 EST; 1 day 6h ago
Process: 3382 ExecStart=/usr/bin/pveproxy start (code=exited, status=0/SUCCESS
Process: 24972 ExecReload=/usr/bin/pveproxy restart (code=exited, status=0/SUC
Main PID: 3386 (pveproxy)
Tasks: 4 (limit: 4915)
Memory: 18.6M
CGroup: /system.slice/pveproxy.service
├─ 3386 pveproxy
├─25221 pveproxy worker
├─25222 pveproxy worker
└─25223 pveproxy worker

Dec 29 00:00:36 pve pveproxy[3386]: starting 3 worker(s)
Dec 29 00:00:36 pve pveproxy[3386]: worker 25221 started
Dec 29 00:00:36 pve pveproxy[3386]: worker 25222 started
Dec 29 00:00:36 pve pveproxy[3386]: worker 25223 started
Dec 29 00:00:43 pve pveproxy[3388]: worker exit
Dec 29 00:00:43 pve pveproxy[3387]: worker exit
Dec 29 00:00:43 pve pveproxy[3389]: worker exit
Dec 29 00:00:46 pve pveproxy[3386]: worker 3387 finished
Dec 29 00:00:46 pve pveproxy[3386]: worker 3389 finished

I tried accessing the WEB ui from safari and chrome and Firefox on a Mac and edge on windows

Did you try the restart command?

What is the exact error your getting in the browser?

 

[Johann]

Okay so I forgot, it works on Firefox, but takes a while to load, on the other browsers it says it can't open the page.

Also maybe its related but I can't remote into my windows VMs, everything was fine a week ago and suddenly any UIs on my Proxmox nodes are slow or unresponsive. My second node which is in the same cluster, can't be access through the IP.

 

[Johann]

I think it was accessing it through http instead of https that caused it to fail btw and a combination of the restart command working

 

[Johann]

So just now it broke again and restarting the service doesn't help. I made sure to use https and still nothing. It says it cannot connect to the server

 

[sg90]

So just now it broke again and restarting the service doesn't help. I made sure to use https and still nothing. It says it cannot connect to the server

The fact you also said you're unable to connect to your Windows VM, have you checked network connectivity? Make sure you don't have some form of network issue/outgoing or incoming attack e.t.c

 

[Johann]

My other servers are fine tho, the windows VM is in the Proxmox clusters and everything that isn't connecting is in Proxmox. There doesn't seem to be any network issues. Is there a way to reset Proxmox but keep my containers and VMs?

 

[sg90]

My other servers are fine tho, the windows VM is in the Proxmox clusters and everything that isn't connecting is in Proxmox. There doesn't seem to be any network issues. Is there a way to reset Proxmox but keep my containers and VMs?

Have you tried just restarting the physical server?

 

[Johann]

Yeah, I've been restarting it every time I can't connect to it. Thank you by the way for helping out. I'm really new to Proxmox

 

[sg90]

What do the OS logs show?
Any resource constraints? What does top and free -m show?
Are you using ZFS or anything or just a single disk / mdadm Raid?

 

[Johann]

I'm on a R710 and R510 with hardware raid. They are in a cluster, which honestly I don't need but I don't know how to separate them.

free -m


total used free shared buff/cache available


Mem: 11992 11569 132 21 290 111


Swap: 8191 6562 1629

 

[sg90]

That looks to be your problem them, your running very low on RAM only 111MB free. You probably have processes being killed by OOM to keep the whole server crashing.

You need to adjust the RAM you have allocated to the VMs to leave more available for the core system and host OS.

 

[Johann]

Oh, how much ram should I allocate for Proxmox?

 

[Johann]

When I allocate the ram, should I put the same thing in the hardware settings for the swap and memory? Whats the difference?

 

[sg90]

If your not using ZFS or anything like that then I'd say atleast 1GB min. By looks of it your also heavily using your SWAP so looks like you are very heavily over committing on RAM.

 

[Johann]

Ok, I'll adjust the settings and see how it runs. Thank you for helping out!