There are several handy sayings and rules of thumb that experienced troubleshooters use to cut through whatever problem they’re trying to overcome. And while not all of them are specific to computers, they can usually be applied to computer-related troubleshooting tasks. Here are a few that might help nudge you in the right direction if you’re trying to troubleshoot past some pesky nuisance without success.
Check Your Cables
Well, this one is computer — or at least electronics — related. When weirdo problems arise, check your cables. Reseat your cables, too. You don’t always have to do this for internal cables because it usually takes a good physical jolt to mess those up, but external cables come loose quite often, and are usually the ones where experienced administrators thump their heads and say, ‘how stupid!’ which leads to the next rule…
Occam’s Razor is another one of those scientific/philosophical approaches to problem solving that you may have heard about, and it follows rules that, above all, seem to appeal to the basic tenets of common sense. Basically, it tells us that, when you’re considering two different theories, the simplest one is most likely to be the correct one. This is a contorted way of saying “Follow the path of least resistance.”
Think in His Shoes
Even if he is an it (like a computer), you can easily use this tactic to follow the path of least resistance. The goal is to determine how the computer followed the easiest path and landed with your problem. This requires some existential thinking. See the computer, network, hard disk, virus; be the computer, network, hard disk, virus. This is definitely a skill that takes some practice, but if you can learn when and how to do this, you’ll find solutions to your problems so fast that your peers will think you’re a genius.
Consider the Consequences of What You’re Doing
Of course, you realize that bringing down the server will solve the problem, but that also creates a bigger problem. Think through every possible solution that you have to a problem (once you’ve determined what the problem really is) and consider a few things: Does it solve the problem? Does it create a problem? Does it solve multiple problems? Keep in mind that a single problem can produce multiple symptoms, so a ‘yes’ answer to the question ‘Does it solve multiple problems?’ definitely puts you on the path of least resistance.
Even while you’re troubleshooting, and before you move to actually eliminate a problem, consider the consequences of what you’re doing. Schrödinger’s cat is a famous thought experiment in quantum mechanics that involves a cat in a sealed black box. The idea is that, in order to observe the activities (or lack thereof because the poor kitty would suffocate), you alter the results. After all, an otherwise suffocating kitty would get air from the peephole you poked in the box. Observation always affects outcome in the same way a guy with a TV camera and a journalism badge somehow attracts idiots (myself included). The same principle applies to your server. If you use Performance Monitor to track a problem, you might make it worse because of the overhead and resources that are necessary to run Performance Monitor to begin with. This doesn’t necessarily mean that you shouldn’t run Performance Monitor any more than you should simply wait for the server to keel over. What it means is that you need to consider the intended — and unintended — results of what you do. Determine if these are acceptable. If they’re not, can they be reduced or eliminated, or would this knock you off of the path of least resistance?
Don’t Be Stubborn
When you’re solving a problem, the goal is to solve the problem — not become a hero (although this can’t hurt if it’s a by-product of success and not the goal itself)! There’s nothing wrong with having an idea and abandoning it. There’s also nothing wrong with proving yourself wrong. The only wrong here is if the results don’t solve the problem and the consequences get you fired. Somewhere along the line, you might encounter a problem where you realize that it’ll take you longer to troubleshoot and solve the problem (assuming that it’s not a human-related issue) than it would to simply reinstall Windows and rebuild the server. A good troubleshooter is always curious to know who or what caused the problem, and sometimes it’s easy to forget that the goal is not finding the problem, but solving it. Sometimes the time factor associated with the whole process demands that you throw in the towel and start over. Hopefully you have a good disaster recovery plan in place that will enable you to do this. If not, you’re stuck, and you’ll quickly learn the lesson…
Prepare for the worst, and hope for the best. Make sure that you have a plan in place that is appropriate for your environment in the event that a disaster occurs. If you need to rebuild the server, make sure that backups are available, current, and valid. Help yourself in a crisis by pre-thinking as much as possible. At what point should you choose to rebuild the server? Remember that troubleshooting is solving problems, even those that aren’t computer related, but do have an impact upon your systems, such as power outages and floods. While you can’t really solve these problems, you can overcome them by moving the systems, sending employees home, purchasing new equipment, or even relocating the business to a temporary facility. Consider the ability of the company to conduct business, also.
So these are just a few ways to get through the process of troubleshooting with, it is hoped, your sanity intact and your problems solved. Of course, these make up just the tip of a potential troubleshooting iceberg; what suggestions do you have that could be added to this list of troubleshooting tips?