Troubleshooting Hogs Bandwidth (by Steve Brown)

Vendor Profile - Network Instruments provides in-depth network intelligence and continuous network availability through innovative analysis solutions. Enterprise network professionals depend on Network Instruments’ Observer product line for unparalleled network visibility to efficiently solve network problems and manage deployments. By combining a powerful management console with high-performance analysis appliances, Observer simplifies problem resolution and optimizes network and application performance. The company continues to lead the industry in ROI with its advanced Distributed Network Analysis (NI-DNA™) architecture, which successfully integrates comprehensive analysis functionality across heterogeneous networks through a single monitoring interface. Network Instruments is headquartered in Minneapolis with sales offices worldwide and distributors in over 50 countries.

Stephen Brown is the Product Marketing Manager for Network Instruments with nearly a decade of experience in network management and security. Steve is also the head geek in charge of their company blog, Network Observations. Steve likes to "get his geek on" anything network security related, VoIP stuff, and anything music and biking related (the kind that requires push power).

Between January and February 2008, we have communicated with 592 network engineers, IT directors, and CIOs scattered around North America, Asia, Europe, Africa, and South America. In addition to being geographically diverse, the population was also evenly distributed among different sized networks. This study was a follow-up of a similar exercise we performed a year ago and the goal is to gauge the changing attitudes and concerns of network professionals, to calculate the average time spent by an IT staff troubleshooting network problems, and to determine adoption rates of new technologies and topologies.

As a leading provider of innovative network analysis solutions, helping organizations and enterprises – including 70 of the Global Fortune 100 companies, it is obvious why Network Instruments would want to know what networking experts think. But we have learned that most of all, technologists also want to hear from other technologists. As Mike Barlow, coauthor of the book “Partnering With the CIO”, has said, “[CIOs] want to listen to other CIOs, not the VP of marketing … they want to hear from someone who's had the same experiences they have.”

What we have learned from this global survey is that despite the many new tools for monitoring and optimizing application performance, almost 75 percent of network professionals continue to cite “identifying the source of a problem” as their primary troubleshooting concern, marking a 25 percent increase from 2007. Performance problems are also on the rise with over two-thirds of respondents spending at least 25 days per year determining the cause of these issues.

Our research indicates that network professionals faced the following challenges:

• Continued lack of troubleshooting information: 31 percent cited this as their major network concern
  
• Ensuring application delivery: 25 percent felt this was the greatest network challenge
  
• Application performance headaches: Over one-third cited bandwidth consumption issues as the chief offender, while 32 percent selected application latency and delay issues as the second most common
  
• Problems investigating intermittent errors: 32 percent felt their organization needed to improve its ability to troubleshoot sporadic performance errors
  
• Security and compliance problems: These issues continue to be a headache for three-quarters of respondents
  

As my colleague, Charles Thompson - Manager of Systems Engineering, says,

“Over the last two years, while IT staffs purchased new tools to optimize applications and traffic, the amount of time spent troubleshooting performance problems actually increased. It's clear that relying on new tools or increasing bandwidth doesn't address the performance problems. These problems will continue to grow as companies implement new technologies and applications on their networks. Without visibility into these applications, performance will continue to suffer.”

The following are some of the key statistics from our report …

Major Troubleshooting Statistics

• 73% of network professionals cited “identifying the source of a problem” as their primary troubleshooting concern (increase of nearly 25% over last year's study)
  
• Over two-thirds (71%) of respondents spent at least 25 days per year attempting to identify the source of performance problems as either network or application
  
• Application Performance – Most common problems cited – (35%) monitoring bandwidth consumption, (32%) measuring latency and delay, (29%) handling use complaints, and (19%) managing application patches

VoIP

• 66 percent of organizations had or will implement VoIP by the end of 2008 (increase of 5% over last year's study)
  
• 62 percent did not have visibility into VoIP traffic on their network

10 Gb and MPLS Adoption

• 35 percent of organizations had or will migrate to MPLS in the next year
  
• 24 percent of organizations had implemented or will implement 10 Gb in the next year

Now let’s drill down on the details of what our customers have told us ...

Time-Consuming Troubleshooting

With new applications and technologies being implemented on the network, the State of the Network Global Study 2008 by Network Instruments sought to determine the biggest challenges organizations faced in troubleshooting application performance as well as the amount of time spent at various points in the process.

As shown in the graph titled Primary Application Performance Problems, 73 percent of respondents cited “identifying the source of a problem” as their primary troubleshooting concern. Managing bandwidth consumption was the second most common problem identified, followed closely by measuring latency and delay issues.

Going beyond quantifying how network professionals spent their time troubleshooting, the survey attempted to gauge the time spent in specific aspects of troubleshooting. First, respondents were asked to identify the amount of time they spent replicating problems as a means of diagnosis. Displayed in the Days Per Year Replicating Problems chart, 41 percent of respondents spent less than 25 days per year replicating network problems. Twenty-seven percent estimated spending between 26-50 days annually replicating issues. The remaining 32 percent said they spent at least 50 days annually recreating network issues.

Another key step in the troubleshooting process is isolating the source of the problem to the network, application, or system. As noted in the Days Per Year Identifying Problem Source chart, 29 percent indicated spending less than 25 days per year isolating the source of a problem. Thirty percent indicated spending 26-50 days annually determining the source of a network issue, while 41 percent of IT staffs spend more than 50 days annually isolating the problem’s cause.

Two trends of interest when comparing the results from 2008 with those of 2007:

1) The number of respondents stating “identifying the source of a problem” as their primary application performance challenge rose by nearly 25percent.

2) Most organizations spent at least 25 days each year determining the cause of performance problems. The number of companies remained steady from 2007 to 2008. A partial explanation for the large amount of hours spent troubleshooting may be found in the relatively low number of respondents that used tools and technologies to monitor application performance and health.

Deploying and Managing VoIP

The global adoption of VoIP continued through 2008 with 66 percent of organizations having implemented or looking to implement VoIP within the next twelve months. As shown in the chart titled Global VoIP Implementations, 55 percent of respondents worldwide have implemented VoIP, while 11 percent will implement VoIP in 2008.

Looking at the Primary VoIP Concerns of organizations, most respondents were concerned about quality of VoIP service. Almost half were concerned with the application’s impact on other network applications, and an almost equal number were worried about the general reliability of VoIP within a business environment.

While many organizations had concerns over aspects of VoIP performance like call quality and the application’s impact on the network, many did not have the visibility necessary to monitor VoIP. Sixty-three percent indicated they did not have hardware or software in place to monitor the quality of VoIP service.

In comparing the results of the 2008 Global Study with 2007, three VoIP trends emerged.

1) More organizations are concerned with the call quality of VoIP rather than the application’s reliability.

2) The number of companies expressing confidence in their VoIP application rose significantly.

3) The percentage of organizations relying on VoIP monitoring tools remained the same
even as adoption rates increased.

MPLS and 10 Gb Implementations

Migration to MPLS networks at least on a global basis appears to be steady, with most organizations still in the early stages of adoption. Twenty-seven percent of organizations have implemented MPLS networks with an additional 8 percent of organizations migrating to MPLS in 2008. Despite the immense media coverage of MPLS, 60 percent of firms have no plans to implement MPLS in the next 12 months.

The study also took a look at 10 Gigabit (10 Gb) implementations, which have remained slow. Thirteen percent of global organizations have deployed 10 Gb networks, while another 11 percent expect to roll out 10 Gb in the next 12 months. Seventy-one percent had no intention of investing in 10 Gb this year.

Continue reading other LoveMyTool posts on Network Instruments »