SolutionBase: Even more new features in Exchange 2007

Exchange 2007 has so many new features that we couldn't discuss them all in <a href="">one article</a>. Here's Brien Posey's follow-up explaining even more about what you'll find in Exchange 2007.

This article is also available as a TechRepublic download.

Exchange 2007 represents a major change for Microsoft and the entire Exchange product line. It's as different from previous versions of Exchange as Exchange itself is from Microsoft Mail. Here are some of the key new features you'll find in Exchange 2007.

Built-in antispam

Fighting spam is an ongoing battle. Although it isn't a replacement for a third-party, enterprise-class, antispam solution, Exchange 2007 offers several mechanisms for defending your Exchange Server organization against spam.

Exchange Server's built-in spam filtering uses multiple inspection techniques to determine whether or not a message should be considered spam. One such mechanism is the various allow/block lists that Exchange Server maintains. Administrators can define allow/block lists that provide filtering against specific domains, senders, IP addresses, and recipients.

In addition, Exchange 2007 uses other filtering techniques, such as SmartScreen filtering, local keyword filtering, and locally-determined dynamic sender reputation. In case you’re wondering, dynamic sender reputation helps determine whether or not a message should be classified as spam based on whether the sender has sent spam to you in the past.

When messages are determined to be spam, Exchange can be configured to take various actions. For example, Exchange can refuse to accept the message, issue a non-delivery report to the sender, or remove the message to the Outlook Junk E-mail folder.

For the first time, antispam protection that’s built into Exchange Server also protects against phishing scams. Exchange can scan inbound content and use heuristics, the SmartScreen filter, or any included URLs to check to see if the message is a phishing scam.

Hosted filtering

Hosted filtering isn’t so much a new feature as it is a service. The idea behind hosted filtering is that, for a price, Microsoft will filter out all of the inbound spam and viruses for you. Wondering why you should bother paying Microsoft to do something that you could do just as easily on your own? Let's look into why hosted filtering is the better option.

In most organizations, spam and viruses make up anywhere from 70–95 percent of the inbound e-mail. These unwanted messages can consume a disproportionately high percentage of the company’s Internet bandwidth. Once the messages arrive, additional resources such as CPU time, disk space, disk time, and memory are used to process the unwanted messages.

When hosted filtering is used, all of the traffic that would normally be destined for your Exchange organization is sent to a filtering server owned by Microsoft instead. The server filters out any spam and viruses, and then passes the legitimate messages on to your Exchange Organization. Having unwanted messages filtered out ahead of time allows you to conserve a significant amount of bandwidth and server resources.

Local continuous replication and cluster continuous replication

In my humble opinion, local continuous replication and cluster continuous replication are some of the best new features to make their way into Exchange Server 2007. It has been my experience that the vast majority of Exchange Server related problems that I have had to troubleshoot over the years have had to do with database problems. The nature of these problems varies, and might include anything from database corruption to a database backup that doesn't want to restore correctly.

Local continuous replication and cluster continuous replication are both automated, near real time, backup mechanisms. Before I show you how these replication mechanisms work, I should mention that they should not be used as a substitute for conventional backups, but rather as an additional safeguard for your data.

In order to understand how local continuous replication and cluster continuous replication work, you need to know a little bit about how the Exchange Server databases work. Exchange 2007 stores messages in an Extensible Storage Engine database (previously known as a jet database). When new messages need to be added to the database, Exchange Server does not place them directly into the database. Instead, the messages are written to transaction log files that are 1 MB in size. Whenever a transaction log file fills up, Exchange Server creates a new transaction log file. When a conventional backup is run, the contents of the transaction log files are committed to the Exchange Server database.

Local continuous replication is a mechanism through which two copies of a database are maintained on the Exchange Server. Typically, these copies would be stored on separate hard drives for fault-tolerant reasons. Local continuous replication works by using a technique called log shipping. The idea is that any time a transaction log is filled up, a copy of the log is sent to the database replica. This means the primary database and the replica database will never be completely synchronized, but they will always be within 1 MB of being consistent with each other.

The benefits of local continuous replication are obvious: you'll be provided with a database backup that is always available (as opposed to having to retrieve a tape from offsite storage), and it's always up to date. There are a couple of negative aspects of local continuous replication, however.

One of the downsides to local continuous replication is that it must be implemented at the storage group level. You can not enable local continuous replication on individual databases within a storage group. This is because all of the databases within a storage group share a common set of transaction log files. Another disadvantage is that it tends to be resource intensive. It's fairly obvious that local continuous replication would consume a lot of disk space. After all, if you're creating a backup of your primary database, the backup will be at least as large as the original database. However, local continuous replication consumes other types of resources during the log shipping process. The actual amount of resources consumed varies widely depending upon how busy the server is, and on the efficiency of the underlying hardware.

Cluster continuous replication is very similar to local continuous replication except that rather than storing two copies of the database on the same server, log shipping occurs from one server to another. Cluster continuous replication is a little bit more difficult to configure than local continuous replication, but it does have its advantages. Cluster continuous replication offers full redundancy of both hardware and software. It also performs data validation and allows symmetric fail over. Furthermore, it doesn’t really matter where the two servers are located, so long as there is a high-speed network link between them. This means the two servers could theoretically be located in separate data centers.

Transport rules

My second favorite new feature in Exchange 2007 is transport rules. Transport rules allow you to implement controls on messages as they flow through your Exchange Server organization. I could probably write an entire book on the subject, but I will try to give you an idea of what transport rules are capable of and what they are used for within the space of this article.

One way transport rules are used is to gain control over the messages leaving your organization. Under normal circumstances, a user could type absolutely anything into an e-mail and send it to anyone in the world. If that message happens to contain something inappropriate, it is often your company that will pay the price for the message in the form of bad publicity, or even litigation.

As such, transport rules are typically used to append legal disclaimers to the end of messages or to filter messages based on content. For example, you could create a transport rule that scans messages for profanity and then it prevents the message from leaving the organization. Such a rule could even be designed to send a copy of the message to the human resources Department. A common variation of this type of transport rule involves scanning for certain keywords that would indicate that confidential information is being sent through e-mail.

Another common use for transport rules is to enforce compliance with various regulations. For example, suppose that your organization was required to retain messages related to a certain project for five years time. In such a situation, you could create a transport rule that scans messages for certain keywords related to that project. When such messages are found, you could place a copy of the message into a dedicated public folder that is configured with a five-year retention policy.

In certain regulated industries, employees in one department may be forbidden from e-mailing employees in another department. In situations like these, transport rules can be used to create ethical firewalls. The basic idea is that transport rules can be created that look for messages sent from specific users to other specific users. If such messages are detected, then the message can be blocked, and a message can be sent back to the original sender indicating that they are prohibited from sending e-mail messages to users in the specified department. Keyword-based exceptions can be created to these rules. For example, you could create an exception in which messages containing the phrase "company picnic" were allowed to pass through the ethical firewall.

Improved Outlook Web Access

Numerous improvements have also been made it in Outlook Web Access (OWA). Alert has been redesigned to have reduced bandwidth requirements and improved response time.

The most impressive new feature in OWA is auto-complete. When you begin typing an e-mail address into the To or CC field, auto-complete will fill in the rest of the address automatically, just as it does in Outlook. This is possible because auto-complete information is now stored in the Active Directory.

Other improvements include: a new search engine that allows you to quickly search through your Inbox, LinkAccess based access to document libraries, and built-in Unified Messaging support.

New mailbox types

You probably won't see this feature listed on any of Microsoft's marketing materials: Exchange 2007 now supports multiple mailbox types. In previous versions of Exchange Server, there was really only one type of mailbox, and that was the mailbox associated with a user account. In Exchange 2007, there are four different types of mailboxes.

User mailboxes still exist, and function very similarly to the way that they did in Exchange 2003. A variation of the user mailbox is a linked mailbox. A linked mailbox is a mailbox that is not connected to a domain user account. Instead, a linked mailbox can be associated with a user account from a different forest. This means that people who do not work for your company can have an e-mail address on your Exchange Server without you having to create a user account for them.

The last two types of mailboxes are resource mailboxes. More specifically, you can create room mailboxes and equipment mailboxes. These types of mailboxes are used only for scheduling purposes. For example, if you could create room mailboxes that correspond to each conference room in your office. By doing so, users could schedule a conference room at the time that they are planning meetings.

Granted, it is possible to simply use a user mailbox for conference room scheduling; in fact, that was common practice in Exchange 2003. There are two things that make resource mailboxes special. The first advantage is that you can associate relevant attributes with those mailboxes. For example, if you are creating a room mailbox, you can specify the number of people that the room will accommodate, among other things.

Scheduling assistant

The second advantage of creating resource mailboxes is that they are recognized by Exchange Server's new scheduling assistant. The new scheduling assistant differentiates between users, rooms, and equipment. When planning a meeting, you can simply choose the people that you want to invite, the desired location, and any equipment that you’re going to need.

Another nice thing about the new scheduling assistant is that it is designed to take working hours into account. Each employee’s work schedule is stored as an Active Directory attribute. The scheduling assistant can be configured to take working hours into account when recommending meeting times. That way, the scheduling assistant won't be recommending you have a meeting at 3 a.m. on a Saturday night. Additionally, proposed meeting times are now color-coded to reflect whether the proposed time is a good, fair, or poor choice based on attendee availability.

Loaded with features

As you can see, Exchange 2007 is loaded with new features. I honestly believe Exchange 2007 is probably the most significant Exchange Server release since Exchange 2000.