snmp monitoring issue

[radarcs]

Hello,
We are running a proxmox cluster with following version: pve-manager 7.3-6 on Debian 10
After last package upgrade:

pve-firmware
zfs-zed
libnvpair3linux
lxcfs
libuutil3linux
libpve-storage-perl
libzpool5linux
lxc-pve
qemu-server
pve-i18n
libpve-common-perl
libzfs4linux
pve-kernel-helper
zfsutils-linux

we receive following Alarm from all proxmox containers in our monitoring tool:
CPU Load: No such object (SNMP error # 222)

We also experienced following.
When host reboots and starts the container CPU Load is properly shown.
But when we restart the container CPU Load will show above error.

thx for help in advance

 

[djdonnerwolke]

Hello,

I join this problem.
I have the same problem since the update from my proxmox host. For me the message looks a little different:
For me it shows: No Such Name (SNMP error # 2).



I don't know which update in Proxmox caused the problem.
I am glad to get some help.

 

[Eschin]

Just noticed this problem on my proxmox lxc snmp daemons also. Sometime around early March 2023, all the snmp CPU graphs for my Proxmox LXC containers started working intermittently. Either returning invalid data(NaN) or an actual metric.

Here's what they all started looking like around early March 2023;



Saw this thread with similar timestamps, and thinking we're seeing the same issue. Wondering what changed early 2023 to cause this.

Going to try rebooting my hosts with containers starting up at boot time to see if that resolves the snmp issues, until the lxc itself is rebooted, as mentioned above.

Anyone else seeing snmp data pertaining to cpu load going sporadic like this?