WEB-interface does not have information about nodes and VM/LXC status
- Thread starter SilentStorm
- Start date
The question marks in the tree panel on the left signify a miscommunication with "pvestatd" daemon which is responsible for statistics.
Drop down to PVE shell and examine:
- available space
- log information
- status of pvestad service
- have you done anything unusual?
- have you rebooted?
good luck
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Drop down to PVE shell and examine:
- available space
- log information
- status of pvestad service
- have you done anything unusual?
- have you rebooted?
good luck
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
>available space
I think node has enough free space.
>log information
I dont see any logs.
>status of pvestad service
See screenshot
>have you done anything unusual?
I dont think so
>have you rebooted?
Yes, already rebooted, now it works fine
Attachments
Last edited:
It repeated today, but now question marks only on storage, and I see very high load average on node.
Command "systemctl status pvestatd" send error.
Also I see many messages on local screen (via ip kvm), but cant read (because very low quality - see screenshot).
After reboot, which file contains these messages?
Attachments
IDW your exact problem or your exact HW/config- but my guess is your running some mini pc (or miniscule) with that midget 9w i3-N305 Processor.
Note that CPU is rated by Intel for a max of 16gb as per their docs, you have 32gb. I see this time & time again of vendors (mostly far-east) speccing out their matchbox PCs against MF own specs. You can surely expect trouble.
I see your running (at once) 4 VM's (including a CCTV one which I guess is pretty intensive) + one database LXC
IDW your passthrough situation - but I imagine it exists.
I think its time to actually purchase a PC.
I will point out one other disturbing note: In your original image your running in Legacy BIOS boot mode - however in the second image (with a kernel update) you've gone over to EFI. If this was unintentional - its probably disturbing.
Note that CPU is rated by Intel for a max of 16gb as per their docs, you have 32gb. I see this time & time again of vendors (mostly far-east) speccing out their matchbox PCs against MF own specs. You can surely expect trouble.
I see your running (at once) 4 VM's (including a CCTV one which I guess is pretty intensive) + one database LXC
IDW your passthrough situation - but I imagine it exists.
I think its time to actually purchase a PC.
I will point out one other disturbing note: In your original image your running in Legacy BIOS boot mode - however in the second image (with a kernel update) you've gone over to EFI. If this was unintentional - its probably disturbing.
Thank you.
I checked RAM (original Kingston, what i bought separate from PC-box) with Memtest and checked SSD (also (original Kingston, what i bought separate from PC-box) with Victoria and CrystalDiskInfo and I dont saw any problems.
CCTV VM runs only for watching, no for recording and in this mode CPU load less than 20%.
PC-box has active cooling, and I changer thermal interface, temperatures are good.
I disassembled PC and erase contact for RAM and SSD.
If problem will repeat, next step I will replace step by step SSD for other model and RAM for 16Gb and will check. I think its good way.
And maybe I need external rsyslog server for collecting logs from PVE?
Last edited:
>What was the Legacy BIOS/EFI issue?
SSD is Kingston HyperX Predator 480G (old model with cool MLC NAND, its have m.2 pci-e 2.0 x4 interface and AHCI mode, non NVMe). And I checked how it works with "only UEFI" mode and "CSM enable mode" in BIOS, its not issue, its my experiment. I have new SSD with NVMe, but its too hot for this small PC-box.
I dont know how I can check HW fast. Now I can only run VMs and wait for a problem.
Last edited: