I've inherited an issue with 2 servers, one running SBS 2003 and one running WS2008 (non-R2). Periodically Google and other popular sites will appear down. Now obviously we know it's not a problem with the remote sites, it's become an internal joke such as "Google's down!" which of course it isn't.
It's resolved 99% of the time by simply restarting the DNS Server and DNS Client services on both server machines, then it works again. What is causing this and why? My logs don't appear to show anything relevant at the time of the incidents occuring, just notifications that the service has start/stopped (and yes I have clicked "Show all Events").
I'm trying to setup either Google DNS or the ISP's DNS so if our DNS fails the network can still access the internet, but am unsure of the best way of doing it. It's a live network in a business environment so I can't do too much experimenting. I'm thinking of just turning the DNS services off on the SBS machine and just retaining them on the WS2008 machine. Is this a good idea or not advised?
All the client PC's are on dynamic IP's from a pool of 100 IP's with plenty spare, and the several Mac's we have on-site are fixed IP's, but the problem appears more on Mac's than it does PC's and it's not all of them at, just specific client machines at a time, restarting the server's services fixes the issue. I need to rule out any DNS issues as we "think" we have an internet connection issue as well, with 1% packet loss on a fibre optic line, and naturally the ISP are saying it's our internal DNS and we're blaming them etc.