If you go back to about 1994 or 1995, when Netscape 1.0 became available and the Web started taking off, the network performance needs of the average data center were fairly modest and targeted. Today, keeping networks up and running is not a major concern because connection technologies such as Ethernet are so stable, says Ted Ho, CEO of Gigamon. In addition, the Internet itself is largely resistant to failure, says Jim Melvin, CEO of Apparent Networks. “IP networking is so dynamic that regardless of what happens, it maintains connectivity—one of its greatest features,” Melvin says.
Although basic network connectivity can be described as durable, making sure that the network runs efficiently with a minimum of bottlenecks and outages is a knottier task. Motti Tal, executive vice president of marketing and business development at OpTier (www.optier.com), says that it’s pretty much a given that IT environments have become increasingly complex over the years. “A data center today is composed of so many different moving parts, technologies, and technology expertise areas that [it’s important] to assure that transactions flow smoothly within applications and the infrastructure [in order] to improve compatibility and the end-user experience while reducing costs,” says Tal.
According to Ho at Gigamon (www.gigamon.com), network monitoring and performance analysis in the data center are too often an afterthought. Most network tools purchased over the past decade do not have the capacity to monitor 2010-era networks, many of which employ virtualization, cloud computing, distributed computing, and high-speed connectivity. Given that networks have become so much more complex, how do you analyze network performance effectively? Here are some tips to keep in mind.
Complete Article: Know Your Network Performance: Put A Finger On The Pulse Of Your Data Center’s Flow Of Information
Post a Comment