Since upgrading to LMS3.2 there are approximately 20 devices that are losing the entered host and display names each day. I have reset these, but they are cleared during one of the processes running overnight. How can I resolve this issue?
Yes, this could explain it. I have been trying to reproduce, and I cannot. One thing to try would be to add some static entries for the problem devices to the server's local hosts file. Then, use NMSROOT/bin/resolver.pl to test that IP-to-name resolution is working:
NMSROOT/bin/perl NMSROOT/bin/resolver.pl IP
If that works, run a new Discovery, and see if the same problem occurs.
We had a similar problem here. What it turned out to be for us was a mismatch between DNS, Reverse DNS, and the hosts file on the LMS server. Once they were synchronized properly the name loss stopped. We are beginners with LMS though.