Voicemail Preview
Exchange 2013 now uses the Microsoft Speech Platform version 11.0. This helps to provide a more accurate Voice Mail Preview experience for subscribers. It also helps with the global address list (GAL) grammar generation that has been changed with Exchange 2013. This GAL grammar is used to help determine spoken commands and names of contacts in the users mailbox and in the organization.
In Exchange Server 2013, to address ASR grammar generation scalability issues for UM, the speech GAL grammar generation no longer happens on the server with the Unified Messaging server role installed. Instead, it happens periodically using the Mailbox Assistant, on the Mailbox server running the Microsoft Exchange Unified Messaging service that hosts the organization’s arbitration mailbox. The GAL speech grammar file is stored in the arbitration mailbox for an organization and then later downloaded to all Mailbox servers in that Exchange organization. By default, the Mailbox
Assistant runs every 24 hours
Protect Voicemail using AD RMS
- Prevent forwarding of voicemail
- Integration with AD RMS and Exchange unified messaging
- Permissions designated by sender (by marking the message as private) or by administrative policy
Exchange 2013 UM Architecture
If you notice here, the connection points from CAS to MBX is always going to be the same protocol that the client uses to connect to CAS.
CAS will proxy requests from clients on the appropriate protocol to the appropriate back-end MBX server.
For SMTP- once the CAS decides to accept the message, the SMTP session is then proxied.
One unique difference is Exchange UM. The nature of real time traffic in the RTP media stream does not lend itself to a proxy scenario. The signaling traffic (SIP) is used to redirect
the RTP stream to the back-end mailbox server
