Okay serverfault, please help me figure this one out...
I have four locations, each connected to each location with Juniper Netscreen point to point vpns.
I have a file server in the main office, my main production file server.
About once a month one of my locations cannot get to this server, and it lasts for about 2 or 3 hours. No ping, no rpc, no remote desktop, no shares, no nothing! And it's only this one server. They can still get to all others servers, all other shares.
The office where the server is and the other two locations can still get to this server (file shares), no problem.
The only thing that fixes it is rebooting the server, not ideal as the other locations need to get to it at all times. Restarting the server service doesn't help, I've tried restarting some of the other services too, nothing.
There is absolutely nothing in the log files to explain this, I have been over and over them. Nothing shows up in the log files of the machines at the remote location either. And, of course, nothing in the firewall logs.
A tracert, gets to the server, then times out.
About the only thing I haven't done is set up netmon, or other packet sniffer on this machine. It couldn't be the network interface could it? I'm about ready to just replace the server...
I am notorious for not being able to say the words, "I don't know", but I am saying it now.
Any thoughts?
Thanks to anyone who reads or responds to this!
What is the os of the fileserver? I would check several things, when it occurs next do a netstat on the fileserver and look for the ips of the location you are having issues with and look at the status I.e. Close wait etc. I would do a nmap scan from the site in question to the fileserver to see if you can reach the host at all. I would look at latency from the site to the fileserver. Finally i would do a packet capture and generate some traffic to the fileserver.