Posts tagged Systems Engineering
Poll: Where Should You Start With Network Design?
May 22nd
Today, our Information Thought of the Day (ITTOD) is another poll subject.
The question concerns the systems engineering of network design. Specifically, I am interested in your thoughts on what the first concern of Information Professionals should be when they do design. Obviously, every project will have different prioritization of the considered factors, but everyone has to start somewhere. Where do you start?
I think that paradigms are important. I see most Information Professionals biased towards one of the factors above in their work. I think that it is important to realize the lenses that we see our professional world through.
Do you like polls as a daily topic? Do you have a recommended one for another week? Please let me know.
That is my Information Technology Thought of the Day (ITTOD) for May 22, 2009 ©Scott Coughlin .
A New “Standard” Every Day
May 18th
Today’s IT Thought of the Day is simple. What is it that drives the Information Industry to create new "standards" every time they want to do something new? Do we really need a new standard every day?
Today’s thought was inspired by two recent stories from Engadget concerning two examples from within only the narrow realm of display to computer interfaces.
- Engadget.com: DisplayLink makes Linux source code available… finally! .
- Engadget.com: The next generation of HDMI actually adds some welcomed features
The same situation is found everywhere in the Information Technology world. Whether you are addressing programming, hardware, or software, we, as an industry, seem to actually enjoy complicating our own lives. We create consumer confusion, technological complexity, requirements for new investment, and troubleshooting chaos by our own actions and then have to pay the price ourselves in dealing with it for our stakeholders. The end user is not the one who makes this all work in the Enterprise — we are.
The real question is whether this constant tweaking of interfaces actually delivers more benefit than it does in the costs mentioned above.
I fully appreciate the need for progress. I like the fact that our "hacker culture" make us the "special sauce" that brings agility, unexpected capabilities, and increase productivity. I even welcome the cool new things that every new interface permits our industry to accomplish.
I just think that we all have a box in our closet full of cables that are no longer usable, cards that are no longer supported, software that is incompatible, and books that tell of "the next big programming language" and have to ask myself: Does the benefit of supporting a new standard every time we turn around really deliver more than it costs?
What do you think? Am I on track? What were your favorite interfaces and standards of 2008? Is it getting better or worse? Please share.
That is my Information Technology Thought of the Day (ITTOD) for May 18, 2009 ©Scott Coughlin .
Network Design: Conformance vs. Customization
May 4th
Whenever you approach a new computer network project where you are replacing an existing system with something new you run into one of the classic conundrums of the Information Professional and Information Technology fields: The argument between conformance and customization.
There is a very human tendency to want to provide a customized, tailored, and unique solution to every situation, This I call a customized network solution. I am convinced that there is something in our DNA that makes that feel like a natural course of action. The problem is that it can be expensive, detrimental to security, and drastically increase your long tail costs for maintenance, support, and administration as you attempt to operate a “one-of-a-kind” solution.
On the other side of the coin, there is strong arguments for forcing your network instantiation to conform to a standard solution set. Whether this is using only industry standard equipment, forbidding any custom software, only using web protocols, or strictly following some off-the-shelf model, the befits to standardization are many. First, support and administration are less expensive because you can easily replace and scale your team without being beholden to intellectual property or corporate knowledge dependencies. Second, your long term maintenance costs will be minimized since you can easily replace components and upgrade equipment from wide palettes of available equipment and never ask, “Will this work with our setup?” Finally, in the computer security business every customization or deviation from the standard is just a vulnerability waiting to be exploited by bad guys or worse forgotten by your own support folks.
In my experience some network engineers are of one camp and some are of the other. So if you are in the market for one or the other types of setups, choose your architect wisely. The challenge comes because nearly all users are usually for customization unless they are also responsible for paying for it. Since the upfront costs are not too different, it can sometime be difficult to explain to a non-tech savvy executive that the difference in sustainment expense for a customized solution can easily fun 5-10 times greater over the standard 3-7 year lifecycle of a computer network.
I believe that the best practice is to formally force the customer to define their requirements for the new network up front. Make sure that you have all of their concerns addressed and cost estimated. Being able to show, with numbers, the business effects of each customization decision is key to a satisfied network owner. Once requirements have been agreed to, make sure that the network design is truly aligned to those requirements and include funding estimates on one, three, and five year horizons in the deliverable of the installation.
The pros and cons of each approach are clear. In reality, you cannot have one or the other in total, but rather have to ensure that you manage the polarity between customization and compliance carefully to deliver the required solution set at the expected cost.
Does this adequately describe this challenge? Do you have any stories how one or the other becoming unacceptable dominant. Do you know of any times that the wrong one was emphasized?
That is my Information Technology Thought of the Day (ITTOD) for May 4, 2009 ©Scott Coughlin.
