By Steve Brar
The enterprise network and all the applications that run on it serve as the foundation for how employees get their work done. Email, Microsoft SharePoint, VoIP, customer relationship management, enterprise resource planning, and every other custom or off-the-shelf application runs on the network. In order to provide these applications to end users, more enterprises are adopting a hybrid IT model that incorporates a mix of on premises and cloud-hosted applications, and of networks comprised of private—hosted, public—Internet, and cloud infrastructure.
While this model can provide significant IT and business benefits, new research shows users now struggle with poor application performance issues on an alarmingly regular basis. The result is a major performance gap between the business’ needs and expectations and IT’s ability to deliver. Fair or not, executives will blame IT for this performance gap that can lead directly to angry customers, low employee morale, damage to a brand’s reputation, and lost revenue.
Optimal Application Performance
The Riverbed Global Application Performance Survey 2015 finds that 98 percent of executives agree that optimal enterprise application performance is critical to achieving optimal business performance. And yet, 89 percent of executives say the poor performance of enterprise applications has negatively impacted their work on a regular basis. 58 percent say it impacts their work at least weekly, and 36 percent say it affects them every day.
The primary culprit is the hybrid IT architecture that has become the dominant model for enterprises. Even as enterprises continue migrating applications to the cloud, they likely also keep those that provide access to highly sensitive information and systems on premise. This makes monitoring application and network performance more complex, time consuming, and costly for IT. Add in the need to maintain security even as users increasingly access those applications on their mobile devices, and effective monitoring becomes a Herculean task. You no longer have full control or visibility over what’s going on across the entire network.
And it’s not going to get any easier. In fact, monitoring the network will only become more difficult as cloud adoption rates continue to skyrocket. Nearly all—96 percent—of survey respondents reportedly use some cloud-based enterprise applications in their work today, and 84 percent expect their company’s use of cloud-based enterprise applications to increase over the next two years. When those applications deliver the expected user experience, companies cited they realized improved employee productivity—51 percent, time savings—50 percent, cost savings—47 percent, improved customer satisfaction—43 percent, and faster delivery of products to market—33 percent.
It’s when those applications fail to meet performance expectations that problems arise, and it happens all too frequently.Respondents report that when an application is slow, crashes, or is not available, that interruption and resulting productivity loss lead to a number of problems that impact the company’s bottom line, including dissatisfied clients or customers—41 percent, contract delays—40 percent, missed critical deadlines—35 percent, lost client, or customers—33 percent, and negative impact on brand—32 percent.
Users’ frustration levels are not only a result of poor performing applications, but also a lack of understanding of why IT cannot ensure their expectations are met. Globally, 71 percent of survey respondents say they have frequently felt “in the dark” about why their enterprise applications are running slowly.
Executives can then compound the problem by trying to work around it. 37 percent of respondents admit they have used unsupported applications when corporate applications run slowly or stop working altogether, thus adding to infrastructure complexity with more “shadow IT.”
Achieving Full Visibility
The solution is to establish end-to-end visibility into application performance across the entire network. Immediately identifying the cause of a performance issue is essential to fixing it before users even notice. A majority of the executives surveyed agree that better visibility by IT staff into application performance would result in increased productivity—56 percent, improved customer service—54 percent, improved product quality—49 percent, improved employee engagement—46 percent, and increased revenue—43 percent.
This leads us to a key question you must ask yourself when determining whether you have the necessary visibility, “do I know what I need to monitor on my software-defined data center?”
If you don’t know the answer, you’re not alone. In its July 2012 report, Realizing Business Value and Return on Investment with Application-Aware Network Performance Management, IDC finds that most organizations simply don’t know the types of applications, number of devices, or traffic sources on their enterprise networks.
Overcoming this problem requires implementing a solution that provides multiple unified views of the network, application traffic, and actual end-user experience, and also conducts its own discovery, dependency mapping, and behavioral analysis. The goal is to answer questions including, what’s on your network, who’s using it, how are they using it, where are they accessing it, and when did this all take place?
Only after you gathered this information will you be able to address performance issues. The pervasive virtualization of data center resources, combined with the availability of APIs to control those resources, makes the software-defined data center a possibility. This combination of virtualization and APIs allow for greater agility and improved efficiency, as data centers can now deliver the right resources at the right time.
Ensuring applications perform well requires an understanding of three key requirements, visibility into layers of virtualization, application performance infrastructure must also be virtualized, and API access to application performance infrastructure.
Visibility into Layers of Virtualization
Virtualization introduces layers of abstraction that can hide the details of what’s happening to an application. As physical systems get carved up into logical units, information about the physical system alone is insufficient. You need the ability to isolate performance issues within virtualized and physical environments.
Application Performance Infrastructure Must also be Virtualized
Pervasive virtualization is at the foundation of the software-defined data center. This improves utilization and reduces capital and operating costs. To maximize efficiency across your data center, you need virtual application delivery controllers, storage delivery controllers, WAN optimization controllers, and other application performance infrastructure.
API Access to Application Performance Infrastructure
In a software-defined data center, infrastructure is accessible and configurable through lines of code. That requires all components of your data center—including application performance infrastructure—to have APIs. APIs allow programmers to define what services are needed in their code, as well as integrate infrastructure with orchestration systems.
The days of keeping application, information stores, and users within the confines of the office walls are over. As more applications migrate to the cloud, and as a growing number of employees work in branch offices and other remote locations, the work of optimizing and delivering great application performance grows increasingly difficult. After all, you can’t control what you can’t see. Closing the application performance gap is now a business imperative, and requires enterprises have a clear view of how all applications are performing, whether they are on premises or in the cloud. SW
Steve Brar is the director of platform and solutions marketing for Riverbed Technology.
Feb2016, Software Magazine