Recent Articles

IP Subnetting Part 1 - IPv4

Despite the large number of jokes in high school about never needing to use mathematics after graduation, the art form known as IP Subnetting requires a solid grasp of math based on binary numbers (0, 1).  In this first part of my devoted attention on IP Subnetting, we'll focus on Internet Protocol version 4 (IPv4) addressing, some history, and the methods available to make IPv4 addressing go farther.

Understanding the "network" command

Many network engineers seem to misunderstand the use of the network command in routing protocols on Cisco IOS, especially once they see the usage in BGP.  This occurrence happens more frequently with those studying to become a CCNA, but may not be obvious to experienced engineers either.  Based on my own experience, I have very distinct memories of confusion when the concept I understood was first corrected by a more experienced engineer.  So let's breakdown the confusion.

Troubleshooting with OSI Model

At Layers 1-4, there are several tools or tricks available to quickly verify normal operations.  I have routinely used each of these tools for many years and expect many of them to be familiar.  Many of these tools are technology-specific, meaning that one tool has an application with one technology but not another.  I should also caution that the list below is not intended to be exhaustive or complete.  It's just a collection of my thoughts on what I find really useful.

backnext

Info Point

Building Bridges from Today to Tomorrow™

Primacore® offers consulting services tailored to your business needs, whether that requires process automation and strict, environmental standardization or if you keep an open mind to new technologies and how they can support your business functions.  Invite us into your world and, together, we can start "Building Bridges from Today to Tomorrow™"

Learn More

OSI Model in Practice

OSI Model in Practice

In the previous article, I summarized the seven layers of the OSI Reference Model and promised to show why it should be relevant.  While the model can be (and usually is) a vague concept, it has a tremendously practical application in isolating problems with system-to-system communication.  The OSI Model in practice can be approached in (at least) two different ways: linear vs.  divide-and-conquer approaches.

Linear Approach

The linear approach is a method to begin at either the physical layer or application layer and work through each layer in a linear fashion until the problem is isolated.  I like to refer to each starting point as either top-down (for starting with Layer 7) or bottom-up (for beginning at Layer 1).  For example, if you are unable to reach Routerlabprep.com from your home computer, you could start by examining your physical network connection, verifying each component on a layer-by-layer basis.  You could also start by verifying the operation of your web browser and working your way down.  I generally don't prefer or recommend the Linear approach in most circumstances, since the algorithm used to isolate the problem may not always provide the quickest resolution.  There are times where it pays to troubleshoot with a bottom-up strategy.  A friend of mine is very fond of saying, "It's always the Physical Layer." I have wished many times that he weren't right.  :/

Divide-and-Conquer Approach

The divide-and-conquer approach divides task of troubleshooting the problem in two and eliminates one half with each step, narrowing the potential source of the problem down to a specific layer or function.  For me, the first step in a divide-and-conquer approach generally involves begining at Layer 3 to verify IP reachability.  If the end-system is reachable, move up the stack (Layers 4-7).  If not, then look closer at Layer 3 with a traceroute before proceeding down the stack (Layers 1-2).

Linear vs.  Divide-and-Conquer

The numerous steps required for troubleshooting system-to-system communication dictate that any troubleshooting exercise can be (and many times, will be) a lengthy process.  In comparison to the linear approach, the beauty of the divide-and-conquer approach manifests itself by shortening the time involved for most instances.  Without getting too deep into the theorhetical mathematics behind this phenomenon, the divide-and-conquer method uses a type of recursion to find a match among the set of potential problems.  With the sorts of complex problems that occur in networks, no one has the time to approach them in a linear fashion (unless it's the physical layer or a browser configuration problem, natch).

Practice Makes Perfect

Putting the OSI Model into action requires a lot of practice.  As a network engineer, the reward for such hard work can be attained regardless of the nature of your expertise (whether operational, design, implementation or architectural).  Take every opportunity you can to practice, even if only using your home computer to troubleshoot problems using the Internet and its many wonderous resources.  In the next article, I'll discuss some common tools for troubleshooting Layers 1-4 of the OSI model, including a few that are invaluable to me, but seem to be neglected by some.