Microsoft Exchange Server 2013: E-mail improved

The new release of Microsoft Exchange Server has a host of enhancements, mostly focused on architecture, flexibility and integration.

Alan Maddison

Thousands of organizations rely on Exchange Server to manage e-mail communications. As a mature platform, there have been high expectations for this latest release. Exchange Server 2013 offers a compelling mix of improvements and enhancements that provide significant value and incentive to upgrade. 

In developing Exchange 2013, the Exchange Product Group focused on a number of key areas. The changes are wide-ranging and include everything from how you deploy Exchange to Exchange management tools. For example, Exchange 2013 introduces a new Web-based management tool called the Exchange Administration Center. This replaces both the Exchange Management Console and the Exchange Control Panel that were introduced with Exchange 2007.

At the other end of the spectrum, there are the most significant changes to Public Folders (PF) since their inception. Now both the PF hierarchy and the content are stored in mailboxes. This lets you include them as part of a database availability group (DAG) for high availability (HA). Most of the other enhancements have been in the areas of architecture, storage, HA and site resiliency; compliance and eDiscovery; Lync 2013 integration; and Outlook Web Access (OWA).

Advanced architecture

Perhaps the single biggest change to the Exchange Server architecture in the 2013 release is the consolidation of roles that were introduced in Exchange 2007. With 2013 only offering two roles—Client Access server (CAS) and Mailbox server—this change is a welcome simplification from a deployment perspective. This also affects other areas including simplifying namespace planning for site resiliency and load balancing (Exchange 2013 only requires Layer-4 load balancing). Together with improvements in recovery times for DAGs, these changes will result in greater operational efficiency.

While the CAS in Exchange 2013 provides all the functionality found in earlier releases, it has also been significantly redesigned and simplified. The CAS now provides stateless client authentication and session control. It even supports mid-session changes in CAS connectivity.

The new Mailbox server role encompasses much of the functionality previously found in the Hub Transport and Unified Messaging server roles. It’s now entirely responsible for all data rendering. This new topology also means the CAS and Mailbox roles are now loosely coupled. This provides a great deal more flexibility in how you design and deploy an Exchange environment. For example, you no longer need an administrator to deploy a CAS in the same site as a Mailbox server.

One important point to note is that Exchange 2013 doesn’t include an Edge Transport role. If your organization wishes to maintain its existing Edge environment, you can route Exchange 2013 through an existing Exchange 2010 Edge server (or servers).

As part of the role consolidation in Exchange 2013, the Unified Messaging role is now split between the CAS role that runs the Unified Messaging Call Router server and the Mailbox role that runs the Unified Messaging Service. This change means the CAS role now acts as the entry point for Unified Messaging and Session Initiation Protocol (SIP) requests. It’s also responsible for proxying these requests to the Unified Messaging Service running on the Mailbox server.

This new routing topology is different, however, in a Lync-integrated environment. The reliance of Lync on its Inbound Routing component means it considers an Exchange 2013 Mailbox server a trusted peer, just as it did with Exchange 2010. It will route directly to the Unified Messaging service on the Mailbox server.

In addition to topology changes, there are some feature enhancements such as Voice Mail preview and Caller ID. The speech-to-text accuracy in Voice Mail preview has been improved by using version 11.0 of the Microsoft Speech Engine and version 4.0 of the Unified Communications Messaging API.

Caller ID reliability improvements are based on changes in contact management and aggregation. In addition to searching the default Contact folder in a user’s mailbox, Exchange 2013 will search across other contact folders the user may have created. Exchange 2013 also supports aggregating contact data from other sources such as Lync 2013.

Storage support

As with earlier versions of Exchange, storage and its impact on performance remains a critical consideration. Starting with Exchange 2007, Microsoft made many significant advances related to storage. These included broadening support for different types of storage, moving from a SAN requirement to recommending Just a Bunch of Drives (JBOD), redesigning the internal structures of the Exchange databases and reducing the IOPS requirements of the databases.

The Exchange team has continued to make storage-related improvements. Exchange 2013 has improvements in the following areas:

  • Reduction in IOPS, with deeper checkpoint depth to reduce IO from passive databases
  • Multiple databases per volume that optimize use of large disks
  • Automatic Reseed for rapid recovery from disk failures
  • Automatic recovery from storage failures, including failures related to long IO times and excessive memory consumption by the replication service
  • Lagged copy enhancements for automatic log play down

Exchange 2013 also introduces a new Managed Store process. While the Information Store service has always been a critical part of Exchange, the Exchange 2013 code has been rewritten in C# and is now known as the Managed Store. This consists of two services, one of which is a dedicated worker process.

As a result of this improved granularity, there’s a dedicated worker process for each database. This ensures that in the event of an issue with a single process, it will only impact a single database. As part of this redesign, the Managed Store is also tightly integrated with the Exchange Replication service. It now uses the FAST search engine for better indexing and search performance.

Site resiliency and HA

While Exchange 2013 continues to rely on DAGs and Windows clustering to provide site resiliency and HA, there have been a number of improvements to both functions.

In Exchange 2013, the site failover process is significantly simplified. CAS and Mailbox recovery can occur independent of each other and automatically. There’s also namespace redundancy through the ability to assign multiple IP addresses to a specific Fully Qualified Domain Name (FQDN), as nearly all client access is via HTTP in the event of failure of one of those IP addresses. For example, if the virtual IP address (VIP) of a load balancer is unresponsive, the client will automatically try one of the other IP addresses configured in DNS.

The Microsoft Exchange 2013 development team saw an opportunity to create an automated, integrated health-monitoring solution that could improve availability. Comprised of two services that run on each CAS and Mailbox server, Managed Availability is implemented as a workflow. That workflow has three components: the Probe engine, the Monitor and the Responder engine.

Health data collected from synthetic transactions is fed to the Monitor by the Probe engine. The Monitor analyzes the data and uses a predefined rule set to determine the health of the various Exchange services. Based on the result of this assessment, the Monitor can initiate a recovery process through the Responder engine or simply create an event log entry.

In Exchange 2010, Active Manager used several database properties—including copy queue length and replay queue length—to determine which database copy it should select in a failover situation. This internal algorithm has been extended in Exchange 2013. It now includes additional health checks made possible by the Managed Availability processes. These new constraints mean that in a failover situation, Exchange 2013 is actively considering not only the database status, but also the status of additional server components.

Compliance and eDiscovery

Compliance and eDiscovery have become important considerations for many organizations, so Microsoft has made significant advances in archiving and discovery. Exchange 2013 has a number of important enhancements including in-place holds, in-place discovery and unified search across both the primary and archived mailboxes. It also has new data loss prevention (DLP) capabilities.

DLP is a method of protecting sensitive information and ensuring that users adhere to corporate policies. Exchange 2013 implements DLP via Transport rules and actions. To accommodate this DLP capability, Microsoft has extended Transport rules and implemented new predicates such as “MessageContainsDataClassifications.”

Exchange 2013 performs actual content analysis and classification through techniques such as dictionary matches, keyword matches and parsing data using regular expressions. The new DLP capabilities also help you define Policy Tips. These are an extension of Mail Tips and you can use them to notify users of possible compliance issues with data in their e-mails. Exchange ships with a number of predefined templates designed to comply with existing requirements, including those defined by Personally Identifiable Information (PII), Payment Card Industry Data Security Standards (PCI-DSS) and financial regulations in the United States, United Kingdom and Japan.

Exchange 2013 also introduces a number of archiving and discovery enhancements. These include a unified approach to placing legal holds (in-place hold) and searching (in-place discovery). In-place hold lets you place query-based holds often used in eDiscovery searches or time-based holds to meet governance or regulatory requirements.

The tight integration between these two features lets you simultaneously search and hold content using the same interface and query. Keyword Statistics also helps you refine your eDiscovery approach. This provides critical data on the effectiveness and breadth of your results.

You can also archive Lync content via Exchange and control the disposition of that data with the Exchange retention policies and discovery capabilities. This integration across applications extends to using SharePoint 2013 eDiscovery Center. You can search data archived in Exchange and SharePoint and place holds on that data.

Lync Server 2013 integration

Continuing the Microsoft tradition of “better together,” there are a number of new Exchange 2013 features that integrate directly with Lync 2013. These features include archiving integration and a unified contact store.

When you enable archiving integration, all archived data from a user’s Lync client is written to a hidden folder in that user’s mailbox. This folder is also used by the Exchange litigation hold feature. It’s indexed and discoverable by either a mailbox discovery search or the SharePoint 2013 Preview Discovery Center. The policy that determines whether or not to archive content is controlled by the Exchange archiving policy in place on the user’s mailbox.

The unified contact store is new to Lync 2013 and Exchange 2013. Earlier versions of Exchange and Lync maintained lists of contacts separately, even though some of the data was amalgamated when presented as part of a user’s Contact card.

Lync 2013 and Exchange 2013 can now share a common storage location for contact data. The unified contact store lets Lync read and write data to the Contacts folder in the user’s mailbox. When you first enable this feature for a user, all their existing Lync contacts are automatically migrated to Exchange 2013, thereby ensuring a seamless experience.

You can also now use high-resolution photos in the Lync client and the Lync Web App. Either you or the user can upload photos and store them at 648 x 648 resolution. While stored in the user’s mailbox, they aren’t actually viewable by the user outside of the Lync client.

OWA enhancements

In Exchange 2013, OWA has gone through some significant changes, including extending OWA across all major client devices such as desktops, tablets and phones. The development team redesigned the interface to ensure a consistent UX across all devices. More importantly, the underlying core application logic is also consistent across devices.

Beyond the interface changes, one of the most significant is offline support. With Internet Explorer 10, Safari 5 or Chrome 16, users can now use OWA even while not connected to the Web. While this functionality doesn’t quite match the capabilities of Outlook, a user can still perform common tasks such as sending e-mail and meeting invites, and working with items in frequently accessed folders.

Another new feature in Outlook and OWA is the ability to extend functionality through the Office cloud app model. Based on platform-independent JavaScript APIs, this new model will help developers easily extend OWA.

Microsoft Exchange Server 2013 adds a lot of features, functionality and flexibility to an already strong and stable platform. The changes affect nearly every aspect of Exchange, from server architecture to revamped PFs. Now is a good time to start evaluating this new e-mail platform to see how it might improve communication and productivity in your organization.

Alan Maddison

Alan Maddison is an 18-year veteran of the IT industry, focusing primarily on Microsoft technologies. For the last seven years he has worked as a consultant focusing on delivering professional services. He is currently a senior consultant with Microsoft Consulting Services.