Is it possible to have too much of a good thing? Well, it depends who you ask. If you ask the West Virginia winner of a $315 Million Powerball jackpot from several years ago, then he might say “yes”. Shortly following his new windfall, bandits broke into his car and absconded with $500,000 in cash. On the other hand, country music icon Alan Jackson penned a song on the theme. He says “Too much of a good thing … IS a good thing”. One presumes Jackson wasn’t crooning about grand theft larceny. Of course, this blog is about systems and network management, not love songs and the law. I assume readers of this missive all agree Network Management System (NMS) tools fall into the “Good Thing” bucket. However, that begs many questions: If one tool is good, then wouldn’t two be better? How about three? How many tools are too many and what is that tipping point?
The answer to this question isn’t as straightforward as it sounds. Instead it lays somewhere on a spectrum. At one end are the more frugal IT organizations we all know and love. On the other end are the operations helmed by super-geek perfectionists that we also know and love. I would posit that the best way to ferret out that tipping point is to employ a good, old-fashioned cost/benefit analysis. I’m not talking about anything particularly rigorous here, but rather a “back-of-napkin” thought exercise.
Recall that I assumed anybody reading likely agrees that an NMS tool in your environment is good thing. Implicit in that assumption are that the benefits of such tools are obvious. Primary among them are greater control of your infrastructure and higher service levels to your customers. Therefore, I won’t expound further. However, if you’re still not convinced let me know. I’ll make sure one of our account reps calls you tomorrow (and the next day … and the next day … and the day after that).
The more telling part of this analysis is the cost aspect. The NMS tools market is considerably more crowded than it used to be. If you look hard enough you can probably find a tool that will give you the exact feature/benefit you need. However, it’s clearly important to ask about the costs associated with those benefits:
Monetary Costs: I mention this cost first because it’s the most obvious. How much would this new tool cost to implement? Does it require the purchase of hardware? Is there a support or subscription fee? These are just a few questions, but there are undoubtedly others.
Time Well Spent?
People Costs: For a moment forget the potential monetary cost of consultants brought in to implement your fancy new NMS tool. What about day-to-day technical expertise? Do you have the right people in your operation to maintain the tool? If yes, then how much time will it take them to configure it and keep it relevant? Keep in mind this is time NOT spent on other crucial IT maintenance tasks. This cost is one that’s overlooked more often than you might think.
Two’s Company, Three’s a Crowd.
Technology Costs: The last “cost” is something I would refer to as a “technology” cost. Unlike the people and monetary aspects, which are applicable to all environments, this cost really only comes to the fore in IT infrastructures when there is more than one tool deployed. To solidify this idea of “technology” cost let’s consider the following, not-inconceivable scenario:
Acme, Inc. has one NMS tool used by the server team, two tools are used by the Network Engineers (one for traffic analysis and another for availability/performance tracking), the applications group is watching web performance in a fourth tool, and Acme’s managed services provider uses a fifth tool for monitoring IT resources from an external NOC. There’s a partially cross-trained internal IT staff of 20 and a ticketing system deployed to track user requests.
This scenario is technology cost brought to life. Regardless of the silo’ed nature of each functional team, every tool likely has to monitor common subsets of the same objects for complete visibility. Therefore, unless there is some kind of integration between all the tools, then five disparate device/inventory lists need to be maintained. Does each tool send out email alerts to contacts? Now there are five separate contact lists to maintain as well. Your technology costs are the time, money, and expertise that must be expended specifically to get all of the above components integrated and running as a cohesive system.
Less is More … Maybe
This long-winded exposition aside I still haven’t answered the question; “How many tools are too many?” Clearly, the answer to this question depends on the point at which the costs of the various tools deployed equal or exceed the benefits provided. Coming from the perspective of an NMS tool vendor I’ll go out on a limb and say “You only need one tool. Netreo OmniCenter!” However, coming from the perspective of a recovering systems administrator and NMS consultant, I’ll simply say you need enough tools to make sure all of your monitoring needs are met at as low a cost (monetary, people, and technology) as possible.
Network and systems management is hardly a “one-size-fits-all” proposition. Different industries, environments, and architectures all have their own unique requirements. The real trick in answering this question is to do your homework and figure out that tipping point for yourself.