Resolved IPv6 configuration lost across the entire server, reinstallation of IPv6 fails

[W4ru]

Server operating system version

Ubuntu 22.04.5 LTS

Plesk version and microupdate number

Plesk Obsidian 18.0.72 Update 3

Hello!
I have read a lot here and on the web, but I am not getting anywhere.

Brief summary:
I have two separate monitors for ipv4 and ipv6 in platform360.io. On Tuesday, I received a notification that ipv6 was not accessible. I ignored it, but yesterday it became clear that the entire ipv6 integration on the server was no longer working.

I tested many suggestions and ultimately removed the IPv6 addresses used everywhere, including deleting the IPv6 from the server itself (V-Server at Ionos). Restarts.

Then I reassigned IPv6 in the Ionos backend, reintegrated it in Plesk (it was recognized correctly), ran Troubleshooter and plesk repair --web.

After all these steps, the status remains that the websites are not being delivered via ipv6, but also that a “ping6 google.de” is not being resolved from the server.

I have no idea what else I could do.
I am thankful for any advice and would be pleased to send details of the configuration or log file.

 

[Sebahat.hadzhi]

Hello, @W4ru . Is the correct IP address selected in "Domains -> domain.com -> Hosting & DNS -> Web Hosting Access" for the website you are unable to access?

 

[W4ru]

Hello @Sebahat.hadzhi ,

I tested the following scenarios:

1) If I have not set up IPv6 in the V-Server backend and query the interfaces on Plesk/Ubuntu, only IPv4 is there and works (obviously).

2) If I create an IPv6 in the backend, but only set it up at the Ubuntu level or in Plesk via “IP reload” (but not on any website), then I cannot resolve Ping6 externally via SSH from the server or even reach the server itself via IPv6.

3) If I set up IPv6 for websites, nothing changes in terms of the above; the website itself is also not accessible via IPv6 and (e.g., Wordpress) cannot access services via IPv6 using API.

In the meantime, I have gone through everything twice (i.e., removing and reconfiguring all IPv6 settings). Out of necessity, I have now moved important websites to another server and am considering setting up the server again.

Or, since the server is still running Ubuntu 22.04 and is currently not productive, I am considering upgrading to 24.04 and hoping that the error will be “fixed” in the process.

What other details can I provide?

Thank you very much!

 

[W4ru]

Additional information:

The server has been running for 1.5 years. There have been no changes that are in any way “system-relevant.”

The crash happened in the middle of the day. I usually do restarts late at night. But it's possible that I did a restart on that Tuesday. It only takes 2-3 seconds, so no one notices.
About a week earlier, I deleted a subscription because it had been copied to another server via Migrator. Otherwise, just updates from Plesk and Ubuntu. No other “ gimmicks” because a premium customer has their website there.

I have five V-servers running Ubuntu and Plesk.

 

[W4ru]

Upgrade to Ubuntu 24.04.3 LTS went smoothly.

However, status remains unchanged:

ping6 google.de from server remains unsuccessful, everything else based on ip4 as well.

 

[Sebahat.hadzhi]

Thank you for the update. Do you have Firewall enabled on the server in question? If yes, could you please confirm the status of the "IPv6 Neighbor Discovery" rule? Additionally, could you please run the following command and provide us with the output:

ip6tables -L -n -v

 

[W4ru]

The firewall from Plesk was and is not in use.

The output is attached

I use the firewall in the backend of the V-Server. Identical to what I use on the other servers. This remains unchanged and applies identically to IPv4 and IPv6. But I have considered several times sending a support request to IONOS to ask whether the FW there is buggy... I will do that now.

 

[GuidoDe]

I have exactly the same problem. VServer from Ionos. The server has been running for years. To find the cause, I temporarily disabled the firewall and WAF in Plesk and re-read the IP addresses in Plesk. I also reinitialized the Ionos network firewall.

Nothing helps. Connections via IPv6 fail.

 

[W4ru]

I have opened a ticket at IONOS.
I will send the details to their support team and post here when I receive feedback from them.

 

[GuidoDe]

"ip -6 neigh" currently shows
fe80::1 dev ens192 lladdr 82:01:cf:____:____:3e router STALE
2a01:239:____:____::1 dev ens192 FAILED

(____ replaced for real IPv6 addresses)

I've also opened a ticket.

 

[W4ru]

"ip -6 neigh" currently shows
fe80::1 dev ens192 lladdr 82:01:cf:____:____:3e router STALE
2a01:239:____:____::1 dev ens192 FAILED

Thanks for your reply!

I get the following output:

root@mx:~# ip -6 neigh
fe80::1 dev ens6 lladdr 82:01:d2:87:xx:xx router REACHABLE
root@mx:~#  ip -6 neigh
fe80::1 dev ens6 lladdr 82:01:d2:87:xx:xx router REACHABLE

This would suggest that there may be a different cause in your case.

What does netplan try or netplan --debug try return?

I had to correct the permissions here (!), but that only resolved the error message regarding the permissions. The error “ping6 google.de” or external access remains unchanged.

And this is the netplan config here:

ls -l /etc/netplan/
total 4
-rw------- 1 root root 485 Sep 29 11:55 50-cloud-init.yaml
root@v2:~# cat /etc/netplan/50-cloud-init.yaml
# This file is generated from information provided by the datasource.  Changes
# to it will not persist across an instance reboot.  To disable cloud-init's
# network configuration capabilities, write a file
# /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
# network: {config: disabled}
network:
    ethernets:
        all-en:
            dhcp4: true
            dhcp6: true
            match:
                name: en*
    renderer: networkd
    version: 2

 

[W4ru]

Correction & addition:

I wanted to send for IPv4 and IPv6, but the output above was from a different server... here is the correct output:

root@v2:~# ip -6 neigh
fe80::1 dev ens192 lladdr 82:01:1e:9d:xx:xx router STALE
root@v2:~# ip -4 neigh
10.xxx.xxx.xxx dev ens192 lladdr 82:01:1e:9d:xx:xx REACHABLE

So for IPv6/IPv4, I only have one line each in the output.

 

[Sebahat.hadzhi]

@W4ru , to me it looks like the IP address is not present in the config file at all. What you can try is to add it manually:

  dhcp4: true
      addresses:
        - 2001:8d8:abcd:1234::2/64 (where 2001:8d8:abcd:1234::2/64 stands for your IP and mask )

However, if you already opened a ticket with IONOS they should be able to handle the task for you as your service provider.

 

[W4ru]

I received an automatic reply from Ionos (after sending a scope log file yesterday) and my issue was marked as resolved.

Ping6 from server to goole.com and ping6 to the server confirms: all OK like one week ago.

I hope to get a little more background information over the phone...

It now looks very much like there was an error on the side of the V-Server that I don't have access to, or at IONOS itself.

Learning curve: Just check things out yourself quickly and get the service provider early on board ;-)