Clients also interface with Global Catalog servers in order to access their Exchange mailboxes. So that begs the question - are there DC's in each remote site? Are they global catlogs? "Republicans in Congress are moving to ratify a constitutional amendment to ban flag burning, thus ending the Iraq insurgency."

0

Please see my response #2. Only one DC at all, in the main store. Only one network, only one forest, only one domain name, only one Exchange Server, so only one Global Catalog. The only difference for the branch stores is that they are on different subnets, but they all have the same default gateway.

0

How can they all have the same default gateway if they're on different subnets? Back to the discussion on DC's/GC's, maybe it's possible the one DC you have which is obviously your sole GC is overwhelmed. Next time it happens, try connecting to the DC using LDP in support tools on port 3268. "Republicans in Congress are moving to ratify a constitutional amendment to ban flag burning, thus ending the Iraq insurgency."

0

You're right, they are NOT all on the same DG, sorry! We only have about 150 users in a non-tech environment, so it shouldn't be overloaded, but I have run load graphs too, and they look OK. I'm leaning towards possible WINS or DNS errors maybe, and am looking into that right now... Thanks-

0

Hi, We have the same problem. It started after we removed a Checkpoint firewall from in between the server and the Cisco Router. When mail stops working, RDP also stops functioning from off certain machines on a remote subnet, to the Exchange 2003 server. But, other machines on the same remote subnet can still connect trough RDP. Ping works fine from all machines in the remote subnets. I do see a lot of routes in the routing table (route print)on the exchange server, from every connecting remote VPN-LAN. As this might be the problem, we´ve turned of the propagation of routes on the Cisco Router but that doesn´t stop the routing information to appear in the servers routing table. So still, twice every week, at random times routing to the exchange server starts malfunctioning on application level. We are at a loss. I know this is no resolution, but maybe the extra infromation might leed to one. R, Maik

0

Thanks for the note Mark; I was sure that others are having the problem! Do you get any notices in the Event Viewer when it goes down? We don't... Ping works fine during this time here as well. It seems to be twice a week consistently too, I don't know why that would be. Thanks again- -Rick-

0

Rick, We get no messages in event log because of the simple fact that Exchange is still running correctly. I am convinced that it is an application level routing problem. Could be that it has been initiated bij an update. When did trouble start at your company? R. Maik

0

Rick, Can you tell me what service pack (and patch level) are on the Exchange 2003 Server? R. Maik

0

We're experiencing the same problem. We have one Exchange 2003 SP1(enterprise) server with 70 mailboxes set up at our HQ. Approx every 36-48 hours the users at our 10 remote branches (connected via IPSec VPN) lose connectivity to the Exchange server. The only thing that cures this is to reboot the Exchange server. We have two DCs that are acting as GCs at the HQ. All local Outlook clients remain connected when the branches go down. Netstat shows numerous open connections from each remote client on the Exchange server when this problem occurs. The clients can ping the host name and IP of the Exchange server during the outages. This is driving us crazy!

0