My question is two parts really. I have a Windows 2000 AD environment. Almost all clients are statically IP'd, though there are a few DHCP clients out there but not many. I just noticed that the DNS records in my DNS servers don't necessarily match up with the true records for a number of desktops. As desktops have been replaced over time, the process to name them and keep the same old IP address was lax.
Correct me if I am wrong, but it would appear to be that XP & Win2kPro desktops do NOT dynamically update their DNS entry if they are statically IP'd. From what I have read so far, you need them to be using DHCP for addressing, and then it should work. So first part of the question is, I am right or wrong with this paragraph?
Assuming I am right, I thought of creating a more automated method to update the DNS server entries to match the real values on the client. I was thinking of having a script to ping a machine, then run the 'hostname' command on it. Then I would take the correct IP and hostname and use the DNSCMD command to create the proper entry. In theory, I don't see why this shouldn't work.
However, the DNSCMD errors out on me every time, from my Domain Controller/ DNS server itself. (actually from both of my DNS servers).
Starting simply, from an RDP on the DNS server, I try:
dnscmd . /Info which gives Command Failed: Unknown 1717
dnscmd hostname_of_dns_server gives same error 1717
dnscmd IP_of_dns_server /Info gives same but error 1753
Further commands such as:
dnscmd . /RecordAdd company.com newhost A 10.10.10.10 give me the same errors above, depending on the entry for my DNS server, either a 1717 or 1753 error.
It would appear that the DNS command fails when communicating to the DNS server. Clearly, doing these commands on the DNS server itself shouldnt be an issue. Is there some configuration or Group Policy setting that would block this? I am beginning to think I somehow misconfigured integrated AD-DNS.
Any help would be appreciated.