Evaluating high-availability solutions for Microsoft Exchange.

Your network topology is excellent. You have all of the latest, high quality hardware and your monthly software maintenance and update schedule is planned and executed with military precision. However, you still have to schedule downtime for server maintenance, notify the entire company, and listen to their collective groans. And, as always, you have more IT tasks than hours in the day to complete them in.

This is a typical scenario within many IT environments; maintenance, upgrades and the addition of new hardware and users are mapped out with care to minimize interruptions and make the most of your staff's time and resources. Then, out of the blue, the mail store gets corrupted and your Exchange server crashes.

This is not atypical of an IT professional's work environment. These talented, dedicated professionals work long hours behind the scenes to assure that the local and global corporate services stay up and running, smoothly, efficiently and consistently.

Continuity of mission-critical infrastructure and applications is one of the biggest challenges they face, and the number one mission-critical application is email. When it fails, access to an estimated 60 percent of corporate data is cut off and everyone from the CEO to the front-desk receptionist is immediately aware of the problem and starts yelling. IT is then typically burdened by numerous calls from users unable to access their email, further compounding the problem and limiting the capacity of the team working to resolve the issue.

Companies today rely heavily on email to conduct their business and can lose thousands, even millions of dollars in productivity by a single day without access to this essential communications tool. Implementing an email continuity solution has been proven cost effective compared to the very real costs of even a brief loss in connectivity.

So, with the variety of marketing promises for 24 X 7 faultless continuity solutions, a fixed budget, and no time for added tasks, how do you evaluate the options for dependable, cost effective, mission critical application continuity?

There are six key questions to ask that are critically important when considering a continuity solution for Microsoft Exchange. These questions are important because many of the so-called high availability and continuity products actually destabilize your network and proliferate data corruption when your Exchange server fails, adding to the workload of the already overburdened IT staff.

The questions are simple and the details are important so make sure you get past the marketing hype to the honest detailed technical answers. Look your vendor in the eye and ask:

1. Does the product place intrusive software agents on my Exchange server? (For your peace of mind and the stability of your network the answer must be no!)

Many continuity products install drivers, operating system services and/or other software on the Exchange server. The drivers are installed to intercept and capture disk block level writes. The services have write access to and perform operations on the Exchange mail store and configuration.

2. Does the product make ongoing modifications and configuration changes to my Active Directory and DNS? (Again, for the stability of your network, the answer must be no!)

Most continuity solutions make continual, automatic modifications to your Active Directory during both failover and failback. These third-party modifications are extremely risky to the stability of all of the Microsoft infrastructure applications running on your network and dependent on Active Directory. Destabilization of Active Directory can cause these applications to crash.

3. Will Exchange database corruption be replicated on my Exchange server? (If you do not want to explain to everyone, from the receptionist to the CEO, why their email has disappeared from the Exchange server, the answer must be no!)

Most Exchange continuity solutions replicate database corruption to the backup server. Furthermore, backup servers that share a data or mail store device are completely disabled in the event of database corruption. These products copy at the block, file or byte level and cannot detect Exchange database corruption when running in real time. Therefore, the vendors of these products recommend that you run the secondary server data replication a half hour behind the primary server to minimize the possibility of data corruption. The half hour delay produces a data loss window during which time email can be truly lost and irretrievable. And, despite the delay, corruption can still be replicated on the secondary server.

4. Is Failover truly instant or a 10-30 minute downtime window for my end users? (If you don't want to spend budget money and your precious time installing a continuity solution that "almost" provides continuity, and then still have to cope with all of the employees screaming at you that their email is down, then the answer must be instant failover!)

Continuity product vendors claim to have instant failover. However, most solutions take anywhere between 10 and 30 minutes to complete failover. During this failover process when the services on the primary Exchange server are shutting down and restarting on the secondary server, the end users are down and/or email is lost. Furthermore, the DNS entries cached on each client desktop must expire before clients can connect to the secondary server. End users experience downtime on each failover and failback.

5. Do I have the added burden of substantial ongoing maintenance tasks for a secondary server? (If there are simply no more hours in your day or days in your week and you want to sleep peacefully when you finally go to bed after a long day, then the answer must be no!)

Most continuity solutions require IT administration to maintain the secondary server, including regular maintenance and applying patches, updates and security fixes. These secondary servers need to be in "lock step" with the primary Exchange server, meaning that the servers must always be at the same software version and patch levels making the IT tasks more complex and time consuming. Therefore, having a secondary server doubles the Exchange server maintenance and support hours for IT administration unless you choose a solution that includes remotely managed updates and regular maintenance.

6. Will setup & installation take hours or days out of my already overflowing schedule? (If the only time you want to spend hours or days setting up or installing something is when you order that present for your child that says "some assembly required," then the answer must be no!)

Because most continuity solutions require setting up and maintaining a secondary server that exactly matches the version and patch level of your Exchange server, and you have to install software agents on both your Exchange server and the secondary server and configure those software agents on each server, these solutions require many complex, manual installation steps and are prone to error. This situation destabilizes your Exchange environment and increases the likelihood of Exchange server failure and the resulting end user downtime--here we go again.

So you have read the literature, had the hard discussions, and wrenched the detailed answers to these critical questions from numerous vendors. Choose wisely because you now know which products have "some assembly required" in the fine print and which ones actually do what they claim. Remember, with all of your hard work, you deserve the peace of mind in knowing that your Exchange continuity solution really is always on and available, so you don't have to be.

Andrea Skov is vice president marketing & business development at Teneros, Mountain View. CA

www.teneros.com