At Interop our interest was piqued by a new managed service from Hughes Network Systems that we see as a harbinger of services to come. Using home-grown software installed in a DSL modem, Hughes bonds (i.e. inverse multiplexes) up to four DSL lines to deliver a maximum of 12 Mb of bandwidth to distributed branch offices. Hughes touts performance comparable to Ethernet without the accompanying price tag, thus positioning itself to unseat incumbent MPLS providers. Read more
RIPE NCC (Réseaux IP Européens Network Coordination Center) in the Netherlands is in the early stages of rolling out a distributed Internet measurement system called Atlas. The system collects data that it incorporates into a master "atlas" of Internet information available in condensed form to the public, and in more detailed form to researchers and participants. The Atlas system, which launched in November of 2010, now has about 1500 USB-powered probes (a.k.a. nodes), which volunteers install on their own premises. Read more
We may be witnessing the ants in the Internet ant hill starting to pull together to compete against Goliaths like Akamai and even Amazon Web Services. After a decade of status quo in the CDN space, we're seeing an innovative new business model--one that federates cloud service providers to offer content delivery services over participating cloud hosters' infrastructure . . . anywhere. A UK company named OnApp has created a federated CDN platform for cloud hosters, which enables them to rent unused computing capacity to and from each other to deliver CDN services. Read more
With Akamai's uncharacteristic collaboration with Riverbed to embed Steelhead WAN optimization software into its CDN overlay network, the tide has officially turned--WAN optimization as a service is going main stream. It is unprecedented that Akamai, known for a "not invented here" culture, chose an established vendor to deliver its first bona fide WAN optimization service, Steelhead® Cloud Accelerator. Meanwhile, upstart Aryaka is following in Akamai's original CDN footsteps and building a WAN optimization service on its own proprietary appliances. Read more
Do you know how much traffic the act of streaming video to a user's device generates? Chances are you don't, and you don't particularly care. But you should, because lively video surfing by lots of users can affect your work and home life. When you multiply the traffic that video surfing generates over the course of a month by the number of users doing the surfing, it can add up to a helluva lot! Read more
As iPads, iPhones, Nooks, Kindles, and other mobile devices multiply like rabbits and join the established population of desktop and laptop computers, the nature of traffic to and from all device types is changing. The key reason for this change is that users are downloading more apps, which replace the browser for completing most tasks. Mobile device users generally rely more heavily on apps to do their bidding, while computer users continue to rely more on browsers. Read more
The term Intercloud describes the future interconnectedness--or federation--of clouds, similar to the network of networks that is today's Internet. Like the electrical grid, in which a utility is shared based on supply and demand, the Intercloud will be a mesh of cloud computing resources owned by many, and interconnected and shared via open standards. Read more
The FCC (or someone) should implement an Energy Star equivalent for bandwidth consumption by Internet applications such as Google and Facebook to educate consumers about demand placed on their broadband connections by the prodigious growth in resource consumption by major websites, browsers, and behind-the-scenes services. The Internet is an economic ecosystem. It is impossible to promote its health by focusing solely on either the supply or the demand sides of the system. Until now, all eyes have been on the supply side but not the demand side. Read more
NetForecast examined the FCC's Internet connection performance data to determine if consumers are receiving the bandwidth they pay for. We found that the higher the bandwidth purchased, the better the chance that the bandwidth delivered meets or exceeds advertised levels. At lower service tiers, consumers generally receive less bandwidth capacity than advertised, while at higher tiers consumers tend to receive more than the advertised bandwidth capacity, except during peak usage periods Read more
Like most systems, the Internet is subject to the laws of resource supply and demand. Demand for Internet resources that affect speed is most heavily influenced and driven by the way websites are designed and operated. The primary factors influencing the demand for resources that affect page load time are payload size and application turn count. Payload consumes bandwidth resources, and application turns consume latency. Read more
In its recently published Measuring Broadband America report, the FCC erroneously uses the terms 'bandwidth' and 'speed' interchangeably. The FCC should know better. ISPs claim to deliver high speeds if you buy their higher bandwidth services. This misleads consumers into assuming that a higher bandwidth connection will automatically deliver a faster user experience. In fact, bandwidth is but one of a half dozen factors that affect user response time (a.k.a. speed). Read more
Based on NetForecast's analysis of the test methodology underlying the FCC's Measuring Broadband America report and our experience evaluating alternative consumer broadband performance measurement methodologies, we conclude that the SamKnows consumer premises measurement boxes used produce high-quality test results. That said, however, some of the test servers introduced data inconsistencies and errors, and the data analysis was insufficient transparent to fully validate the results. Read more
The recently published FCC Measuring Broadband America report, which publicizes measurements of US ISPs' broadband performance, covers only half of the broadband performance equation--the supply half. NetForecast believes it is time for the FCC to expand its investigation to include the demand half of the equation--the prodigious growth in resource consumption by major websites, browsers, and behind-the-scenes services. Read more
This week NetForecast publishes a free report detailing the findings of our recent APM best practices benchmarking survey. The results document how, compared to enterprises with poor APM best practices benchmark scores, top-scoring enterprises experience 75 percent fewer critical application problems, are 75 percent more likely to learn about performance problems proactively rather than from user complaints, and typically spend four hours resolving an application performance problem, compared to 16 hours for low scorers. Read more
As is the fashion, application performance management vendors are now speaking "cloud". But each tool is unique, and it is challenging to see the big picture of how the current crop of APM tools fits into the often complex cloud environment. So, to identify the right application performance measurement toolkit for different situations, we have taken a stab at placing APM vendor tools into the cloud services "geography" framework we introduced in our last posting. Read more
When it comes to measuring and managing application performance in cloud environments, you need to look at where your cloud services are located within your application delivery system. This cloud service "geography" is critical because whether your cloud service(s) are at the back end and/or at the front end of your application delivery system will determine where and how you need to measure performance--and where you measure will dictate what application performance measurement toolkit is right for you. Read more
If you want the performance of your networked applications to place you at the top of the charts you should have two or three performance management tools as we described last week. But which vendors are likely to get you the best results? Our APM enterprise benchmarking survey results shed light on that question. Read more
You may be yearning to buy management tools from yet another vendor, but NetForecast's recent survey of performance management best practices and their results shows that using more management tools is not necessarily better. Most high-performing enterprises use two or at most three vendors, while low-performing enterprises use one or two. The takeaway is that most enterprises need at least two and no more than three vendors to achieve good application performance results. And if you have stellar best practices, you can get by with one vendor. Read more
Many enterprises zero in on watching either infrastructure or application performance. Based on results from NetForecast's latest application performance management benchmarking survey, we can unequivocally say that those with a "stereoscopic" view of infrastructure as well as application performance achieve dramatically better results than those with a "monoscopic" view of just one or the other. So if you are focusing on only one view, we suggest you get with the program and expand to a more holistic, dual view of performance. Here's the data that shows why this is smart. Read more
If you work in a big organization replete with financial and human resources, ITIL may be your best bet for structuring your application performance management. But if your organization is like most, and you don't have the wherewithal and time to implement full-blown ITIL, we suggest a much more practical and less onerous approach to managing application performance. At NetForecast we call it "ITIL-Lite" for APM. Read more