Hi All, We currently have a Exchange 2010 SP1 RU6 Dual Datacentre and related Ad Sites Site A and Site B. Our AD is setup as a Hub Spoke not a full mesh and AD sites X, Y Z have AD Site IP links to Site A but not to Site B. As we load balance across both datacentre's we have users in AD sites X, Y, Z which currently have performace issues connecting to the CAS Array in Site B. But we also have users in site A which have no issues connecting to Site B. Could there be a AD related issue with this desgin? All the doco I have read seems to show that this should work as the AD lookup should be completed by the DC where the mailbox is located not by the site DC. Thanks Josh

You have how many Physical Sites ? How many AD sites ? -- Exchange is in Site B Users are also in Sit A -- Now Which Site users are having performace issues ? What exactly they are experiencing - They go disconnected for some time or ?Satheshwaran Manoharan | Exchange 2003/2007/2010 | Blog:http://www.careexchange.in | Please mark it as an answer if it really helps you

[PS] C:\>(Get-ADSite).count 34 Users in all sites apart from A, B (Yet to confirm every site I have confirmed 6 so far) exchange in Site A and Site B only. Users outlook hanging and popup box states disconnected and then restored after 10 sec. Users in Site A have no issues getting to the CAS array in Site B. The issue screams network and I think it is ALG related on one FW. But also wondered about the current AD desgin given they all point to Site A.

The AD design would have no affect, all your Exchange servers are in site A and B, the user sites have no exchange so you're strickly speaking about AD site links (not exchange site links) for AD replication this has no affect on exchange routing. I would investigate the FW, get help for your network admin to do some netmon captures. James Chong MCITP | EA | EMA; MCSE | M+, S+ Security+, Project+, ITIL msexchangetips.blogspot.com

Yep that's my thoughts exactly. The AD lookup's for excahnge should be completed where the mailbox lives not the remote site. Our netmon traces show no server disconnect but a reconnect. Junos 10.1 does some funky stuff with the ALG. We are disabling it on friday and changing the keepalaive time later

The AD lookups for the Exchange server (dcaccess) should happen against the GCs in the same site and domain as the Exchange. The AD lookups for the clients (dsproxy) should happen for the GCs in the same site as the users (supposed to not always the case) James Chong MCITP | EA | EMA; MCSE | M+, S+ Security+, Project+, ITIL msexchangetips.blogspot.com