Where Bottlenecks Commonly Occur and How You Can Predict Them

    December 11, 2015

    By Wyndham Sellers

    When a bottleneck occurs within a dense, extensive IT infrastructure, it can be hard to identify where exactly the blockage is located. A good sense of where these bottlenecks tend to pop up will help you avoid downtime and outages.

    IT maintenance is work that is often done best when it goes unnoticed, but as IT gets more highly integrated into every facet of business, that’s becoming a harder standard to meet. With new and sudden growth putting added pressure on IT infrastructure, how can these added complexities and demands be successfully managed?

    The first step is to get out ahead of the problem by addressing the conditions that might cause future issues rather than only reacting to bottlenecks after they’ve already happened. Preventing network disruptions that impede service for your customers, instead of merely responding to problems as they arise, is a mark of IT Service Optimization (ITSO) Maturity.

    The Bottleneck Obstacle

    When a major retailer launches a new clothing line (as The Data Center Journal explains), or doctors and nurses need to share a vast array of patient information within a hospital (as this study on PubsOnline details), or a marketing campaign hits its peak, there are customers depending on quality IT service. A bottleneck could mean a loss of both the purchase at hand and the long term trust of your customers — and in some industries, it could even put lives in jeopardy.

    Yet, according to a 2014 study commissioned by TeamQuest and reported on by Network World, network slowdowns are the single greatest challenge facing IT at large companies. In general, bottlenecks occur when traffic exceeds the 100% threshold of a system's capacity, as this study by Tsinghua University notes. Just like an influx of cars merging onto a highway, an overload of data will slow the flow of an entire system.

    Some increases in demand on your systems are predictable. Seasonality will always be a factor in the strain on your servers. A major shopping holiday will spur activity for retail companies, for instance, while companies with an elderly consumer base will see more traffic during the period of the month when social security checks arrive.

    But what if a marketing campaign goes viral on a major shopping holiday? What if one client plans a half-off promotion the same day that millions of customers receive their social security checks? What are the limits of seasonality in predicting the demand on your servers? In an increasingly complex world, sophisticated systems are needed to predict the multiple factors that might impede service. TeamQuest Predictor software can help lift your ITSO maturity from Chaotic to Value.

    Predicting Problems Before the Bottleneck


    Rather than operating according to a “break/fix mentality,” (as defined by IT Business Edge) TeamQuest's predictive analytic software allows for problems to be detected before a system can be compromised. The software functions as a critical supplement to a system’s additional capacity that exists in case of an emergency such as a bottleneck. A business that wants to reach the Service or Value levels of ITSO Maturity should integrate this software to ensure the reliability of servers even while running multiple high-demand applications.

    “Convergence of social, mobile, cloud, and big data technologies presents new requirements,” Gartner analyst Svetlana Sicular has observed in TechRadar, “getting the right information to the consumer quickly, ensuring reliability of external data you don't have control over, validating the relationships among data elements, looking for data synergies and gaps, creating provenance of the data you provide to others, spotting skewed and biased data.”

    The flip side of these new demands on IT systems is that big data technology is available for IT to improve service. Knowing your capacity, and when it might be reached, is essential for IT. TeamQuest offers exciting new tools for predicting and avoiding potential bottlenecks. It is a capability that can now be done automatically, so that predictive software works to prevent those bottlenecks that even the most vigilant IT staffer might not notice.

    (Main image credit: Ignatius Win Tanuwidjaja/flickr; 2nd image credit: Andrew Hart/flickr)