Take pride in knowing that your company depends on its help desk. And when it comes to information technology problem solving, few skill sets are more valuable to ensure business continuity. Yep, even marketing execs and similar partners pale in comparison to those whose workload is defined by troubleshooting.
Like most professions, though, not everyone is created equal and the skills of IT staff vary from company to company (along with their salaries). What distinguishes run-of-the-mill IT staffing solutions from the real professionals — i.e. the ones consulted in a crisis?
It Could Be Zebras
Some use industry certifications, requirements, employment history or career accomplishments as a guideline. Throwing the resume to one side, how do you really know? By testing their IT problem-solving skills. IT ticket resolution is a skill built in time, but contrary to popular belief, the best practitioners do not rely on years of experience (although it does help) or gut instincts on where the problem lies.
Commonly called "confirmation bias," this refers to situations wherein the problem is identified without analysis and before an actual fix is found. This includes the tendency to interpret analysis to fit the problem. Such an ad hoc approach isn't your best bet, especially in an area that business-critical processes depend on to remain operational.
"If you hear hoof beats, assume horses not zebras." Not so in IT. Certain parts of Africa find zebras to be more likely, and by the same token, support sees threats other users may not expect their machine to suffer from. If you're serious about information technology problem solving, a process is necessary. Call it troubleshooting or diagnostics, but in all cases a structured and logical methodology is recommended. While turning it off and on again has its place, there is no substitute for logic — no matter how volatile — when tracking down a fault.
IT tickets are varied and problems sometimes have several possible solutions, but all of them may work to some degree. Standard operating procedure (SOP) documents are a little more flexible to make diagnostics easier without being restricted to a set process flow. Unlike other departments, it's impossible to set processes and procedures in stone; troubleshooting, by definition, is a trip through clues that may very well lead to a brand new source — which in turn allows the provision of a solution you can call up if the issue recurs.
Really, Sherlock?
Like investigators, IT pros are presented with a case, usually a user complaint or perhaps an observation that the network is unusually congested. In all cases, however, solving the problem involves five main elements, each of which must be completed. No space for impromptu patches in this process. The five elements start when a problem is reported and takes the following route:
- Collection of all available information or "how does the problem manifest?"
- Information analysis. What does this actually mean, technically?
- Process of elimination. If the wireless network is slow, it's unlikely to involve the coffeemaker (which is on a separate network).
- Speculate on the cause. "What could be the problem?" Several possible options exist, and these should be ranked from highest to lowest in probability.
- Test the solution. Depending on time of day or if doing so will cause downtime, further analysis may be necessary before implementation.
Diagnostic Methods
Process is one thing, but how about methodologies? Unfortunately, when it comes to information technology problem solving, there are several in common usage. Selection often depends on the problem involved. The most common are:
- Following the traffic. Network monitoring tools allow the tracing of routes, ping, traffic analysis and more to simplify diagnostics.
- Spotting the difference. Compare devices or processes that are working with those that are failing.
- Moving the problem. If you move a device or swap cables, does the problem move with it?
More complex methodologies involve starting at one of the seven layers on the Open Systems Interconnection (OSI) model. Layer selection will depend on the problem symptoms.
IT diagnostics is a specialist's protocol, and while common problems may well be fixed instinctively or on the obvious, it is best to record all problems in a structured manner. A ticketing system is perhaps the most visible form of this, affording your company a tangible record of all issues, how long they took to solve and the methods used to rectify them. Surely, this is a better way than relying on gut feelings or preconceptions of certain devices or software.