Hi everybody,

we are implementing a MX Exchange 2013 (CU 2) as the first Exchange server in the organization.

A ReceiveConnector for connections from our LAN network should allow authenticated users to send mails as any sender string. The solution I found here and elsewhere on the net seems to be simple:

Add-ADPermission -identity smtp-lan-relay -user "nt-autoritt\anonymous-anmeldung" -extendedrights "ms-Exch-SMTP-Accept-Any-Sender"
Add-ADPermission -identity smtp-lan-relay -user "nt-autoritt\anonymous-anmeldung" -extendedrights "ms-Exch-SMTP-Accept-Authoritative-Domain-Sender"

Unfortunately it is not working with FrontendTransports!
If I delete the connector and set it up as HubTransport it works!

Now my questions:

• is that a known issue / bug or a wanted behaviour?
• is there any disadvantage if I am running the connector as Hubtransport instead of FrontendTransport (as it should be)?

Thanks a lot for your help!

Hi

It might be a known issue and I Found a similar thread

http://social.technet.microsoft.com/Forums/exchange/en-US/44b67f45-25cb-4496-b8e0-c94d55c98281/authentication-changes-2013-on-receive-connector

It used the same workaround as yours

To create a new receive connector and define it as a Hub Transport and not a Front-End Transport.

I dont think there will be any disadvantage.

Cheers

Hi Zi Feng,

thanks for your confirmation...known issues sounds as "bug" that should / will be fixed in the future?
So I will change all my connectors to hub-transport in the meantime.

What I dont understand: if there is no disadvantage for me to set it as hub-transport I dont get the reason why there are different transport-roles at all...?!?

Thanks,

Matthias

Not sure it will be fixed or not, please wait for the released notes.

In my point of view, change to hub transport is just a workaround due to this known issue, and it will take any bad effect, But these two settings should be different. 

... and it will take any bad effect, But these two settings should be different. 

Garen,

did you men will not take and bad effectt?

Because if there are bad effect, I would maybe not implement the workaround ;-)

Thanks,

Matthias