Local DNS on the VM do not resolves after 10 minutes

[dune356]

Hello,
I am facing an issue which drives me crazy.
I have a Proxmox installation which runs an Almalinux virtual server whic runs Plesk.
When i boot the VM the local DNS work fin and then after 10 minute does not resoves. The sequence is just after rebooting the Almalinux VM using the teminal:
1) PING google.com -> Works, returns the traditional data
2) PING 8.8.8.8 -> Works, returns the traditional data
3) AFTER 10 minuntes (approx.)
4) PING google.com -> freezed, no return
5) PING 8.8.8.8 -> Works, returns the traditional data
It is anyoying because some services on th website could not operate correctly when the DNS is not resolving.
I have tried a lot of conf options but not one work (stop the firewall at proxmox level and VM level etc..
Does any body faced this issue ?
Thanks for your help.
Best regards.
John

 

[Onslow]

Hello, start from finding what is the DNS server for this VM (when resolving works and when doesn't).

Edit: for instance execute
nslookup google.com
and see what the IP address is in the lines Server: and Address: (the Address: just below Server:, not below "answer" section).

Like in an example from my PC:

Server: 1.1.1.1
Address:        1.1.1.1#53

 

[dune356]

Hello, start from finding what is the DNS server for this VM (when resolving works and when doesn't).

Edit: for instance execute
nslookup google.com
and see what the IP address is in the lines Server: and Address: (the Address: just below Server:, not below "answer" section).

Like in an example from my PC:

Server: 1.1.1.1
Address:        1.1.1.1#53

Hello,
Thanks for your reply.
Here are the result when it is working and when not.
WHEN NOT WORKING:
[root@monterosa ~]# nslookup google.com
Server: 1.1.1.1
Address: 1.1.1.1#53

Non-authoritative answer:
Name: google.com
Address: 142.250.181.238
Name: google.com
Address: 2a00:1450:4001:828::200e

WHEN WORKING:
Server: 1.1.1.1
Address: 1.1.1.1#53

Non-authoritative answer:
Name: google.com
Address: 142.250.185.78
Name: google.com
Address: 2a00:1450:4001:800::200e

It seems to be the same.

 

[Onslow]

Now I don't understand. Both when "working" and when "not working" you successfully receive a good answer.
So in both situations it resolves OK.

Could you describe the facts in more details?

 

[cwt]

I would check the VM‘s logs for possible errors during the „does not work“ period:

systemctl list-units --failed
journalctl -b | grep -i dns

 

[Amgems]

Could be similar to what I am finding -- vmbr0 stops receiving packets sent to VMs.

vmbr0 is used as basis for vlan04. ioT is a bridge based on vlan04. I have a vm which has ioT as the network. I reboot proxmox, and DHCP works. After a while, I cannot log into the VM, and on investigation, discover that the lease has expired, the DHCP requests are going out, but the DHCP responses are not received by vmbr0. I verified the DHCP server is sending the response. I verified the switch has a MAC table entry for the VM addressed to the correct port. I also verify via mirroring traffic that the response is sent. The proxmox host seems to lose the knowledge that the VM macs are to be received. Because the VM's DHCP requests are not replied to, it deletes the default route, and the VM no longer responds to HTTPS tcp connection requests, which are not on the ioT lan segment, so require a default route.

Just did an update, and there was a lot, so I am hoping to find it fixed. Was not fixed prior to reboot, and reboot fixes the MAC filters for all the VMs, but eventually, they are lost, and connectivity to VMs breaks.