Security alerts and OWA connectivity issue after reboot of Exchange Server and DC's.

[nelsonsk2]

I've got a 2003 SP2 Domain with primary and replicatin DC and Exchange Server 2010 SP2 on 2008R2 server. All have been working fine for the 9 months since bringing up the exchange server. After routine updates and reboot some of my outlook users are getting security alerts for the autodiscovery and OWA users are getting errors when logging in. If the outlook users click yes to the security alert or re-install the cert they connect and may or may not get the alert hours later. The OWA users can just refresh their browser from the error and in 99.9% of the cases their mailbox will load fine. I'm also getting periodic Initialization Failed errors when opening EMC.

I've run ExBPA no problems are found. I've run dcdiag on both DC's and from Exchange against both DC's and all pass. I've also restarted MSExchange AD Topology service and the others required by that restarting that service. When I review the event log on the server though I'm seeing a number of errors. These are not all of them but they seem to all indicate communication issue between DC's and Exchange.

MSExchange Mailbox Replication Event ID 1005
Description:
The Mailbox Replication service was unable to determine the list of mailbox databases hosted in the local Active Directory site.
Error: Could not find any available Domain Controller.

MSExchange ADAccess Event ID 2102
Description:
Process MAD.EXE (PID=4708). All Domain Controller Servers in use are not responding:
dc1.domain.local
dcr.domain.local

MSExchangeMailSubmission Event ID 1009
Description:
The Microsoft Exchange Mail Submission service is currently unable to contact any Hub Transport servers in the local Active Directory site. The servers may be too busy to accept new connections at this time.

 

[ShackDaddy]

Can you check your directory services event logs? It could be that you have journal wrap, USN Rollback or some other AD issue that's causing the Exchange server LDAP queries to fail intermittently. Also on each of your DC's run "repadmin /showreps" and see if there are any unsuccessful replications recently.

I'm pretty certain all your issues are going to boil down to AD/DNS issues.

Are both your DC's global catalog servers? Does the DC that Exchange is primarily configured to communicate with also hold the FSMO roles?

Dave Shackelford
ThirdTier.net
TrainSignal.com

 

[nelsonsk2]

The directory services event log only has a couple NTDS Replication warnings Event ID 1083 over the past several months. Everything else is Info only events. The last warning was on 5/23 but this problem just started on Friday after reboot.

I ran the repadmin /showreps on both dc's and everything shows successfull.

Both DC's are global catalog servers and the exchange server is pointing to my primary dc with holds the FSMO roll.

Also, as an update when I logged into my EMC this morning to investigate why my printers will no longer send to email I received several Intialization failed errors but I'm in now. I'm unable to connect to the Queue viewer and the transport service won't start. I'm going to reboot the server.

I agree it sounds like a AD/DC issue but All the errors seem to be on the exchange server rather than either DC.

 

[nelsonsk2]

After the reboot this morning the Exchange server found my DC's and seems to be working normally. No errors for exchange are showing any longer. Unfortunately the reboot seems to have broken my spam filter now but I'm working with their support on that one.

 

[nelsonsk2]

I'm refreshing this thread from last week as my server has begun exhibiting the same problems again. This morning I've had three instances where the exchange server can't find my DC's and the transport service just stops without any error other than the "Can not find any available Domain Controllers". I'm receiving a number of MSExchangeApplicationLogic Event ID 9106 and MSExchange ADAccess Event ID 2102 and some MSExchange EdgeSync Event ID 1025 errors. I did have to reboot the server again yesterday morning but everything worked fine all morning.

I've read on a on a couple of forums that some .NET Framework updates released last week were known to cause some issues with Exchange Server 2010 and required multiple reboots to get all the services working properly.

Has anyone else experienced this? ExBPA, DCDiag, etc all say everything is working properly but the event log on my Exchange server and performance on some mailboxe connections speak to the contrary.

Any thoughts?