If someone asked you to identify the single biggest weakness in your Exchange configuration, what would your response be? Some administrators would cite their hardware; others would point the finger at their network or Internet connection, and some would put the onus on their software. However, in a surprising number of cases, the true weak link is lack of knowledge about and planning for your environment. The fastest, least-expensive route to better uptime for your Exchange systems is to become smarter about how you do your job.
Now, before you hammer me with email, let me make clear that I know administrators have a tough job and that products are often to blame for operational failures. We all have horror stories about unreliable hardware or poorly designed software that either did something wrong or made it too hard to figure out how to do things right. However, an unfortunate number of sites are forgoing some simple steps that can help raise their uptimes. In the spirit of helping banish unscheduled downtime (and productively filling what for some people might be a slow holiday period), let me present a few simple suggestions for improving your messaging availability.
1. Know your SPOFs. That Exchange server, laboring away in a corner, that has only one disk and one power supply is a SPOF, or single point of failure. Any component or service whose failure can knock out your Exchange server is a potential SPOF. Some SPOFs, such as electrical service, might be outside your control. Others, such as how much redundancy you specify when you buy hardware, are within your control, although you might not be able to change them now. But what you can certainly do is review your network and Exchange design to make sure you know what SPOFs exist and what to do in case one or more of them fails. In other words, you should know which Global Catalogs (GCs) your Exchange servers routinely talk to (use the Dsadiag tool), what kind of redundancy your hardware has, where you'd get spare parts in case of a failure, and so on.
2. Revisit your procedures. Do you know what to do in case of a virus outbreak on your servers? How about a failure of your RAID controller? What would you do if your office flooded over the weekend? You should have a clear plan of action for every failure or situation that can keep your users from getting their email. And you should write down those plans. Then, if you eat too much holiday turkey, another team member can follow the appropriate plan to restore service while you're sleeping off your overindulgence.
3. Poke around. Are you confident that your performance and availability monitoring is doing a good job? Dust off the Performance Monitor and use it to see what's really happening on your servers. While you're at it, make sure that your server and message monitors are functional—they won't do you any good if they silently fail.
4. Play "If I Had a Million Dollars." If the budget fairy paid you an unexpected visit, what would you spend the money on? New servers? More bandwidth? A team vacation to Cancun? Although you might not be able to actually follow through on these plans, it's always wise to know ahead of time what you need in the future.
5. Get smart. 2002 is drawing to a close; now is a good time to evaluate your personal knowledge level. If your year-end schedule and budget allow it, why not take a training class to bolster your knowledge in some Exchange-related area? If that's not possible, spend a few hours each week from now until the end of the year reading a good Exchange book or browsing the dozens of technical white papers that Microsoft makes available on its site. If you really want to challenge yourself, set a goal to take and pass the two Microsoft Certified Professional (MCP) exams for Exchange: Exam 70-224, "Installing, Configuring, and Administering Microsoft Exchange 2000 Server," and Exam 70-225, "Designing and Deploying a Messaging Infrastructure with Microsoft Exchange 2000 Server."
These steps aren't magic, but if you take the time to carefully assess your SPOFs so that you have a better picture of the real weaknesses behind your network, you'll be much better prepared to fix them. Likewise, my other suggestions call for you to evaluate what you know, what you have, and what you do so that you can make the necessary changes. Happy tweaking!
