Disclaimer:

All the information in this blog post is subject to change as Exchange Server 2013 is still under construction. Although it’s not very likely that big changes will occur between now an RTM, it might.

Over the past few months there have been quite some speculations about how Exchange Server 2013 would look like and what changes would be included.

Today, we can finally start taking a look at the latest member of Microsoft’s messaging platform: Exchange Server 2013.

This article only touches a subset of the things that changed. Over the upcoming days and week, I’ll be releasing more information so don’t forget to check back regularly!

Architecture

Back in the days when Exchange Server 2007 was released, Microsoft “split” Exchange Server functionalities into multiple roles because – at that time – hardware was a limiting factor: you didn’t always have as much processing power as you could possibly need (at least not in an relative affordable way).

At first, this recommendation has been transferred over to Exchange Server 2010 only to be changed somewhere mid-lifecycle, simply because hardware now was powerful enough.

Now, In Exchange Server 2013 the number of available roles has drastically been reduced. There’s only a Client Access Server and a Mailbox Server role. The Client Access Server acts purely as a proxy (which means that it doesn’t process data!) while the latter is more like a combination of all the other roles (MBX, CAS, HT, UM). In a way you are now always deploying multi-role servers.

Because of these changes, the affinity between the Client Access Server and the Mailbox Server becomes less important whereas before, some protocols required you to connect through the same CAS. Because of some limitations that would still exist with the RPC protocol, it has been dropped entirely for all direct client connectivity. Yes, this means that your clients will only be connecting using HTTP (RPC over HTTP).

This change in behavior introduces some additional benefits:

  • You’re not necessarily required to deploy Hardware Load Balancers. DNS Round-Robin could do the trick.
  • The number of required namespaces is drastically reduced
    In particular, I especially like the reduction of the amount of namespaces that is required. Will make deployments a lot easier!

Store improvements

It seems that the store has been rewritten, AGAIN. Only this time, it has been rewritten in C# making it entirely future-proof Winking smileAlong with the new source code comes a new name: “Managed Store”. It’s designed to allow more granular management. In contrast to what some have been saying (perhaps even hoping), it continues to leverage ESE as the underlying engine. Why? Well, to quote Ross Smith IV on this topic: “SQL squeels like a pig where ESE is easy”… My guess is that he means that ESE still outperforms SQL when it comes to the typical transactions that are performed against an Exchange Database.Also new is that each database now runs within it’s own dedicated worker process: each mount-request will create a new worker process which exits when a database is successfully dismounted. This means that the process of one database does not necessarily impact another process/database when e.g. it hangs.

Indexing & Searching

As expected, and predicted: FAST is now integrated with the Managed Store to provide a better search & indexing experience.

High Availability

The Database Availability Group (DAG) is still responsible for providing high availability. However, some “under the hood”-changes like code improvements and a deeper checkpoint on passive nodes allow for faster fail-over times.

Management

The Exchange Management Console is no more. It has been replaced by the Exchange Administrative Console (EAC), a web-based management interface that allows you to perform most of the common management tasks. Since the EAC resides on the server, it’s pretty much accessible from virtually anywhere and doesn’t require you to install the Management Tools like before. As with OWA, Safari, Firefox and Internet Explorer will be supported.

(Modern) Public Folders

For years, Microsoft has been deemphasizing the use of Public Folders. However, companies still continue to use them. Let’s forget some of the issues for a moment: they’re relatively easy to setup and are very easy to use.

That’s also one of the reasons why a lot of companies are still using them. One of the problems with Public Folders though, is that they cannot benefit from the Database Availability Group and therefore lack any real form of High Availability.

This changes to the better! Public Folders are now transferred into Public Folder mailboxes. This means that they reside in regular mailbox databases and therefore can benefit from a DAG’s protection. Look at it like this. Public Folders are now a new type of mailbox. To setup a new Public Folder, you’d use New-Mailbox –PublicFolder.

Of course there’s much more to it than this. For starters, I find the migration process – as it currently exists – a bit too complicated and am curious to see whether that will change. Check back over the upcoming days/weeks as I will be going more into detail about the new features and changes!

Site Mailboxes

Team Mailboxes are a great example of how the different products within the Office Suite come together. The idea behind Site Mailboxes is simple: it’s a “joint-venture” between SharePoint and Exchange Server allowing you to store and access items in SharePoint directly from Outlook, OWA or SharePoint itself.

This feature is huge but will require you to deploy Exchange 2013, SharePoint 2013 in combination with Outlook 2013.

Outlook Web App

A lot of effort has been put in OWA. It has completely been redesigned (and rewritten). The GUI is now entirely according the Metro-style you find anywhere else throughout Microsoft’s products and is available in different flavors. For each type of device there’s a different presentation: 1-column (smart phones), 2-column (tablets) or 3-column (mouse-based UI).

Along with the graphical changes & improvements, some features have been added as well:

  • Offline Access which will allow you to store a limited number of items in the browser’s offline cache.
  • People-centric: pretty similar to the “People Hub“ in Windows Phone
  • OWA Extensibility will allow 3rd-party applications to integrate with OWA, therefore enhancing the experience within.

What about ActiveSync?

ActiveSync is very popular: almost every device that’s out there supports the use of it. Unfortunately, the quality of the implementation depended on the licensee rather than the protocol itself. This could cause quite a lot of confusion and doesn’t improve either an administrator’s or end user’s experience: some devices would not support all A/S policies or “behave” badly.

Compliance & Discovery

That compliancy has become an important part of Exchange Server shouldn’t come as a surprise. Recent licensing changes already announced that multi-mailbox searches would no longer required an Enterprise CAL.

The importance of compliancy and discovery certainly shines through in the many changes and improvements that have been made:

  • You can now perform searches across the primary mailbox and archives in OWA
  • You can apply personal tags to default folders in OWA
  • Improved Discovery (e.g. exporting to PST is now possible from the eDiscovery Console; messages can now be previewed in the console)
  • “Federated Discovery” allowing to search through Exchange, Lync and SharePoint using a eDiscovery Console in SharePoint 15.
  • Improvements to litigation hold feature: you can now place a mailbox on hold using a query (query-based hold) or a defined timeframe (time-based hold).

So, what’s still the same?

Basically, almost everything that wasn’t mentioned above. They haven’t changed in that the way how these features work did not change. However, you should keep in mind that now the Exchange Management Console has been removed, they might be managed differently (either using EAC or EMS):

For example:

  • Address Book Policies,
  • Information Rights Management (IRM)
  • Personal Archives,
  • Federated Sharing
  • …

My personal thoughts…

To me, Exchange Server 2013 feels like a natural evolution of its predecessor Exchange Server 2010: there are no groundbreaking features, at least none that you couldn’t expect. It has already been predicted a long time ago that the management of Exchange would shift from the Exchange Management Console (EMC) and PowerShell (EMS) to a more web-based management portal (EAC) and EMS.

It seems that Microsoft has aligned the product more with what we see in the real world and I look forward experiencing how these changes will have a positive impact on both deployment and management of your Exchange server environment.

However, if there were a single item I had to choose from that I welcome the most, then I would most probably not be something in the Exchange Server’s architecture, but rather the unified end-user experience. I think this feature will be one of the key reasons that will drive the deployment of Exchange Server 2013 along with it’s deep(er) integration with products like SharePoint and Lync.

Disclaimer:

Please note that the information in the blog post is based on the current available build (public beta). The final version (RTM) might be different from the current one in that Microsoft can decided to add or remove certain features between now and RTM.

For more information, have a look at the Exchange Server 2013 Preview Documentation