Hi,

I read through this article by Ross:

blogs.technet.com/b/exchange/archive/2012/05/30/rpc-client-access-cross-site-connectivity-changes.aspx

I am having trouble understanding the effect this would have in a datacenter outage scenario.

Using the example environment in the article:

Previously (before RU3) if the primary site went offline (mailbox and CAS), the standby copies would become active, but clients would lose connectivity until  RPCClientAccessServer attribute for each failed over DAG database was updated and replicated through AD to point to the new CAS server.

Now after RU3, if a datacenter outage occurred, would the clients find the standby CAS using autodiscover, and the standby CAS would accept the connections due to AllowCrossSiteRPCClientAccess being $false for the DAG, therefore, no need to update RPCClientAccessServer? and no need to wait for AD replication to get client connectivity back?

I also assume due to  this AllowCrossSiteRPCClientAccess attribute being $false, when patching is occuring in the primary, and DB's are failed over (the primary CAS is still online), users are now being prompted to restart outlook even though the primary CAS is available, whereas previously it was unnoticeable due to cross-site CAS? This would be an acceptable compromise if the behavior for datacenter outages is correct!

Sorry this is in 2013 discussion, I could not find a forum for it in exchange server forum category

Hi,

I think we can refer to the following article:

http://blogs.technet.com/b/mspfe/archive/2013/05/15/exchange-2010-why-does-allowcrosssiterpcclientaccess-revert-to-false.aspx

Thanks,

Hi,

Thanks for your reply. Although the article you quote is regarding an issue with the AllowCrossSiteRPCClientAccess value defaulting to $false. This does not explain the implications for DR in my scenario. I just want to confirm that if it is set to $false we would no longer have to set RPCClientAccessServer to the secondary CAS and wait for replication to get access back for the databases that failed over to the secondary datacenter. In fact, this would be a beneficial change if I understand it correctly?

Thanks