Network Performance Management: A Look Back to 2006
by Team AppNeta on February 13, 2012
Remember when data, video and voice networks were separate? Before cloud was a buzz word, cellphones were only for calls, virtualization was in its infancy and what’s an iPad? Considering all of these changes, it seems an impossible job for network teams to maintain high functioning networks and applications with the flood of changes we have experienced throughout the IT industry. As we are now well into 2012, let’s take a quick visit to what has changed during the past six years.
Managing large, complex, distributed networks has never been easy. But with the dramatic changes that have taken place in both the technology and business landscapes in the past few years, it’s arguably never been harder. Let’s take a look at some of the key trends that have radically transformed our networks – and what it takes to manage their performance – just since 2006.
The convergence of all forms of communication from texting, to gaming, to streaming onto IP networks has become ubiquitous since 2006, due to the convenience, flexibility, manageability and cost reductions that it offers in comparison with separate infrastructures.
But as users exchange increasingly rich and varied content over the Internet and corporate networks, the demand for more and more bandwidth has grown exponentially. At the same time, IP-based applications have grown increasingly sophisticated and hence increasingly dependent on “clean” network capacity to function optimally. Network managers must not only provide enough capacity, but also ensure that latency, jitter and packet loss are minimized.
In July of 2006, VMware released its VMware Server. Virtualization technologies of all kinds have continued to proliferate and mature as businesses look to save money and streamline their physical IT infrastructures. Of course, an offset of increased virtualization is massively increased network traffic. Distributed virtual server instances must communicate over the network, for example.
Cloud and SaaS
Cloud computing leverages virtualization but takes it to a higher level. There was essentially no buzz about cloud in 2006 – now it has become mainstream.
The key impact of cloud on networks is increased utilization – and a greater need for end-to-end monitoring – as more users interact remotely with performance-sensitive, distributed cloud services. Increased network utilization due to deployment of “private” cloud-based applications has become a major reason why organizations need to pre-assess their network capacity before moving mission-critical business procedures to the cloud.
Six years ago, smart phones were much less prevalent than today and tablet computers like the now-omnipresent iPad barely existed. Users are demanding the right to bring their own devices (BYOD) to work, and to be constantly connected to the corporate network.
For the network administrator of 2006, network devices were much better controlled. Healthy devices generally meant a healthy network, and SNMP was available to help ensure visibility. Today SNMP data alone is inadequate to monitor and manage complex, distributed network paths and the proliferation of mobile devices that roam free across corporate WANs.
Back in 2006 the US economy was still ticking along nicely. There were concerns about a slowdown, but nothing like the seismic shocks experienced since.
In comparison, today’s IT departments have weathered layoffs, outsourcing, budget cuts and limited resources. The result is a “do more with less” mindset, with a focus on leveraging current investments to the max, and finding the cheapest way possible to implement anything new.
As a result, IT departments are moving away from complex monitoring solutions and embracing small-footprint, inexpensive tools. Many organizations now lack the headcount to manage sophisticated network instrumentation. Today’s multi-tasking administrator handling network monitoring is most concerned with straightforward visibility into the network so he or she can quickly pinpoint and resolve problems, and move on.
A further ramification of this trend is an emphasis on proactively preventing network performance issues. Forward-looking network managers are seeking ways to streamline and fine-tune their networks for optimal efficiency, as opposed to simply “adding more bandwidth.”
As we look back on the network performance management challenges of 2012 from, say, 2018, what will we see? Very possibly an even more complete reliance on high capacity, high performance, low latency, widely distributed networks to convey ever-greater volumes of all forms of data.
End-to-end insight into key performance metrics like capacity, latency, jitter and packet loss along the entire network path from source to destination and back will probably be more important than ever. So will the ability to monitor network health remotely, and to deploy network performance management tools quickly and cheaply as the network morphs and technology changes.
For network administrators, network engineers and other networking professionals, these are interesting times.