Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

On a Vista x64 PC with a wired connection to a Server 2003 Domain environment, a DHCP-assigned IP, and an empty hosts file, intermittently any attempt to ping, telnet or otherwise resolve a particular machine name (or FQDN) (and not always the same machine) will begin to fail with an "unable to resolve" error. The trigger that causes a machine name to stop resolving has not been identified.

However, when this situation occurs, nslookup is still able to successfully resolve the name, and access to the machine via IP address (ping, telnet, etc) is also successful.

UPDATE: "ipconfig /displaydns" returns:

myserver.mydomain.local
----------------------------------------
Name does not exist.

So it seems a lookup failure has been recorded. The lack of TTL normally present on all other entries returned is also concerning.

The simplest way I've found to recover this situation is to release and renew the DHCP lease; "ipconfig /flushdns" does not work.

How can I resolve this or how can I extract more diagnostic information to find the root cause?

Thank you

share|improve this question
add comment

4 Answers

This sounds like name resolution is failing. Is the DNS server also set via DHCP? When you do the nslookup do you explicitly give the DNS server IP? There are issues with some name resolution servers because of the IPv6 stack on Vista so you could see if ping -4 works when ping does not (the Vista IP stack attempts name resolution using IPv6 first and then moves to IPv4 but some resolvers reject the subsequent IPv4 request: see http://technet.microsoft.com/en-gb/library/bb878121.aspx for more information on this). I'm not a windows expert so my usual line of attack would be to look at the packets on the LAN, you might prefer to explore the windows side of things.

share|improve this answer
    
The DNS server is also specified by DHCP. When I use nslookup I do not specify which DNS server to use. When the system is working and when it is not, the same DNS server IP is used. –  Jason Stangroome Jul 1 '09 at 7:03
add comment

From an elevated (Run as Administrator) command prompt you could try resetting winsock/tcpip

netsh winsock reset catalog
netsh int ipv4 reset resetlog.txt
netsh int ipv6 reset resetlog.txt
share|improve this answer
    
I've now tried this and rebooted as suggested by the output of those commands. I expect this is supposed to be a one-off solution to prevent the issue from re-occurring. I will now watch to see if the problem happens again. –  Jason Stangroome Jul 1 '09 at 22:09
add comment

I fixed a similar issue by removing all DNS servers from the DHCP config except the domain controller.

The clues were the advice to try ping -4 in previous replies, and this techarena forum post

share|improve this answer
add comment

If it weren't for the NSLookup working, I would definitely say this is a DNS issue. It still maybe, the first thing I would check when it stops working is that you can ping the DNS server of your network by IP address. If this works then you know its at least contactable.

You can also try pinging the name of a machine using the ping -4 to ensure it uses IPv4, I've known Vista have a similar problem where it trys to use IPv6 for some reason.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.