AMP has lost its licence key due to a hardware or system configuration change

OS Name/Version: Ubuntu 20.04.5

Product Name/Version: AMP Release “Halimede” v2.4.0.10, built 27

Problem Description:

At every system restart I get this message if I login to amp:

AMP has lost its licence key due to a hardware or system configuration change. Please re-enter your licence key under Configuration -> New Instance Defaults. You may also need to reactivate your existing instances on this host.

Then I need to re-enter my key and re-activate all instances.

I need to restart my server a few times per month, so just not doing it is not an option.

Steps to reproduce:

  • Restart the server…

Actions taken to resolve so far:
I upgraded every instance.

There were no issues in the past. This started happening after the last amp release.

1 Like

This is a known issue. You’ll have to wait for a future release that fixes it.

Thank you for the information. It is just annoying, because the usage in the License-Manager is very high. A reset button would be helpful there.

Disregard anything that license manager shows. It purposefully is on a delay. But the system knows what you’re really using. If you ever get an e-mail (you won’t) then you can reach out about it.

Similar issue. Happened 3 times so far. Looking forward to a fix.

I guess this isn’t a priority, and I’m tired of having to put the license back in. Dumping my sub for now till a fix is in. Thanks.

Idk, it is extremely annoying to always do.

1 Like

It is a priority to fix but it’s difficult to track down because I can’t reproduce it. It’s been well known for some time that VPN software can screw with AMPs licencing but beyond that there’s nothing else known to cause issues where AMP can’t reliably establish a machine ID.

Can I help you somehow? Because it is still happening on my server.

Check your list of Network interfaces using ip addr and note down all the mac addresses. Reboot and do the same again. See which if any change ignoring the docker interface if you have one.

Just did that. There are no changes in the loopback and enp7s0(main ethernet) interface.
The interfaces that do change are: docker0, br-mailcow and many interfaces that start with veth.

I am also getting the same problem on every reboot. The hardware for the machine that I was running the server on recently died, and I had to migrate the drive to a VM running on Proxmox. There’s also a large number of docker instances on the same host. The Mac address stays the same though.

Please let me know what you’d like me to look at and I can pass on any information that you need.


We’ve got a change coming in this upcoming update that should fix the issue.

Part of the problem has been that the licence system wasn’t able to reliably determine which network adapter was the systems ‘primary’ adapter. There’s new logic in place now where it does extra tests in order to figure out which exact adapter is the primary adapter and use that.

The network adapter isn’t the only thing the licencing system uses to establish a machine ID but it’s the one that has been failing in most situations due to things like virtual adapters whose MAC address changes each boot.