This “non-resolvable IP address slows down the process” might also explain the high CPU utilization we’ve experienced on yet another node - we cmcdump all of the 30,000 services of our main instance to another Checkmk VM in our DMZ in order to be able to access Checkmk results without need for dial-in VPN. With 2.0, this was slow but stable, with 2.1 it’s really really taxing on the “slave” instance - constantly starved of resources and Nagstamon times out. In this case there are hundreds of occurrences of “Cannot lookup IP address of” every minute, every time we cmcdump the current state of the main instance.