Can anyone provide direction as to why my e-mail server depends on a particular domain controller? It must be up or no one can access their e-mail. I am running Exchange 2003 on a Windows 2000 member server. I have two DCs and each are global catalogs, wins, and DNS servers. Name resolution seems fine. The older Windows 2000 DC is the one that must be running for e-mail to work. The newer DC holds the master FSMO roles and is a Windows 2003 R2 server... exceot for Schema master which resides on the older server. I'm not sure of the history of the environment since I recently started with this company. I ran best practices analyzer for exchange and can't get a clue from it... the exchange org is in native mode... nothing seems wrong.

Event ID 2080 of DSAccess gives you more detail about the communication of Exchange with Active Directory and you can find out why it is giving problem.Go through below article and compare the event id 2080 of your server to look out the cause.Event ID 2080 from MSExchangeDSAccesshttp://support.microsoft.com/kb/316300Additionally check the Directory Access tab of Server properties and make sure that "Automatically Discover Servers" option is selected.A Closer Look At Directory Service Access (DSAccess) - Part 2http://www.msexchange.org/tutorials/Closer-Look-Directory-Service-Access-DSAccess-Part2.htmlAmit Tank | MVP - Exchange | MCITP:EMA MCSA:M | http://ExchangeShare.WordPress.com

Yes, please check the setting in the Directory Access tab. Also raise up the diagnostic logging is a good start for troubleshooting, which is a DSAccess event that can help you diagnose most topology-related problems Issue description: Users cant access their mailbox when a specific DC downs Explanation: By default, exchange server will use the DC and GC servers which are in the same AD site. Meanwhile, exchange server uses Config DC to get its configuration information. If the Config DC is down, exchange server will become unavailable; however Exchange should pick another DC server from working DC list Check info: 1. Whether exchange server had been restarted after the windows 2003 DC added? 2. Whether exchange and outlook has registry settings for specific DC/GC? 3. Whether both DCs are in the same site with exchange server? 4. Have you rebooted the windows 2003 DC after its promoted to GC? As a new GC, NSPI is disabled by default, which is used for outlook address book lookup Troubleshooting: 1. Please check the connectivity between exchange server and DNS server a. Please run ipconfig /all on the exchange server to check if new windows 2003 DC has been entered DNS field b. In the DNS manager of 2003 DC, verify that exchange servers A record exists 2. Please verify the configuration of AD site on the exchange server a. Please run NLTEST /DSGETSITE. If a valid site name is returned, then exchange server can determine which site it belongs to b. Please run NLTEST /DSGETDC:<domain-name>, see if the communication between exchange server and DC is running normally 3. Please verify the health of the domain a. Please run dcdiag >dcdiag.txt and check the output b. Please run POLICYTEST tool (In the exchange installation CD, support\Exdeploy\i386 folder) to confirm that all DCs have exchange-specific parts of the Group policy. Notes: Each DC should return SeSecurityPrivilege. If one or more domain controllers return an error, you should rerun SETUP /DomainPrep in the domain 4. To verify the health of both DCs, please use ldp.exe to confirm if both DCs are responding to both port 389 (DC) and 3268 (GC) connections (KB 252335) Resources: Description of the Policytest.exe Utility Browsing and Querying Using the LDP Utility

One of our DCs had an onboard NIC not being used and not connected, with add in NICs which were teamed but advertised not only thier own address but also the microsoft address from the unused NIC to WINS. When the 1st DC in the list and primary wins server went down then this secondary picked up the WINS requests and returned the invalid address. I simply disabled the NIC,deleted the 1Ch records from WINS, and rebooted each DC to recreate the 1Ch records. Thanks for the responses.Todd