Findings from a recent survey of more than 175 Network Engineers
“Organizations no longer see WLAN as just a cost-cutting tool, but as a means to change how they operate through the use of technology, including ubiquitous access to the network and the adoption of new types of devices. 1"
Taken from IHS’ Wireless LAN Strategies and Vendor Leadership survey of North American enterprises last October, the above quote shows just how essential wireless has become. No longer a nice to have, the wireless local area network (WLAN) is now a critical component to network infrastructure.
The same IHS survey reports investment in wireless continues to grow faster than any other part of enterprise access networks, as organizations move from providing wireless as a convenience to counting on it to drive digital business transformation. 2
All of which means that in a hyper-connected, always-on world increasingly dominated by wireless access, Network Ops is more than a gatekeeping function; it’s absolutely essential. It can either dramatically help an organization move forward or greatly hinder its progress. In short, the Network Ops team truly holds the keys to the kingdom.
Which is all very well and good—unless, of course, you’re the network engineer expected to maintain perfect uptime and no slowdowns. Then it’s no longer the stuff of future legend; it’s just an ever-present challenge.
So, how are today’s Network Ops pros managing an increasingly wireless world? NETSCOUT recently surveyed more than 175 network engineers from mid- to large-size organizations in North America across a variety of industries to find out.
모든 조사 데이터 및 중요점을 고려한 결과, 비즈니스 변신에 대한 WLAN의 중요한 성질 및 그의 관리에 있어 직면하게 되는 상당히 어려운 네트워크 운영 조건에서, 표준화된 프로세스와 확인 목록이 네트워크 성능을 향상시키면서 또한 네트워크 엔지니어의 작업을 쉽게 해주는 것으로 결론지었습니다.
In the pages that follow, we share the survey results, examine the key takeaways and explain how WLANs, a legendary World War II airplane and critical-care processes in an intensive care ward share a common need.
Key Data Points:
- 63% of network ops teams comprise less than 10 people, but 45% of all teams manage 20 or more sites, the majority of which have at least five access points
- 80% of network ops groups have no dedicated WLAN team, despite investments in WLAN growing faster than any other part of enterprise network access networks3
- 10% of network engineers spend more than 25% percent of their time fixing WLAN networks, with another 29% spending at least 12% of their time on troubleshooting, instead of planning projects or configuring/upgrading their network
- As a group, respondents reported using 21 different equipment vendors and 27 troubleshooting tools
네트워크가 점점 복잡해짐에 따라서, 네트워크 운영자가 관리하는 것이 점점 힘들어지고 있습니다.
When Norman Abramson, a professor at the University of Hawaii, invented the first wireless computer communication network in 1971, it was a fairly rudimentary network of eight computers deployed over four islands and was based on ham-like radios4.
Since then, WLANs have circled the globe, adding billions of users and millions of access points (APs), offering greater performance, reliability, convenience and—of course—complexity. Today’s network engineer is spread pretty thin. The majority of survey respondents (58%) reported supporting more than 10 geographically disbursed sites, with 45% supporting 20 or more. This is compounded by the number of APs at each site, usually more than five. That number of APs will only increase, according to the IHS survey, in which organizations said they expect to increase the number of APs by 15% annually over the next two years5.
From a technology perspective, it doesn’t get any easier. Two-thirds of respondents (67%) manage WLANs running on a hybrid of network standards. Given the cost of upgrades, the move to 802.11ac will likely take place over an extended period of time. Of the organizations represented in the survey that haven’t already migrated to the new standard, only 21% will have migrated before the end of the year, while 36% will have only partially migrated, and 43% are further out.
Considering the increasing complexity and investment in WLAN equipment, it would make sense to have a corresponding increase in networking pros dedicated to wireless, but that’s not the case, according to survey respondents. Roughly 80% said their organization had no dedicated wireless team.
Figure 1a: Number of sites supported
Figure 1b: Average number of APs per site
네트워크 운영 전문가가 어떻게 시간을 사용하는지와 그들 부서의 우려/우선 순위에는 분명한 단절이 있습니다.
While network pros cite many of the same concerns as their employers, the way they spend their time shows greater focus on reactive issues, rather than proactive efforts. In our survey, 39% of network engineers cited performance and reliability as their top concern, followed closely by security (26%) and then capacity/coverage (21%), the latter probably the result of being closer to individual user complaints.
These priorities dovetail with the IHS survey, where organizations said security was their chief concern (46%), followed by reliability (31%) and performance (26%). If the IHS survey had combined reliability and performance into one category (as the NETSCOUT survey did), the order of concerns would be a perfect mirror.
Interestingly, 2 respondents reported having “no concerns” whatsoever. We’ll let readers make what they will of that particular statistic.
However, day-to-day activities show a decided tilt toward reactive firefighting. Roughly 39% of network engineers spend at least five hours (or 12%) every week fixing WLAN networks. And 10% spend a whopping 25% or more on troubleshooting.
While some could argue a portion of that time is spent fixing security problems or performance/reliability issues, the statistics are troubling from the standpoint of not anticipating those same issues and problems so they can be avoided in the first place. The point is, no IT organization wants users to serve as an early warning system.
그림 2: Top concerns
There is no single standard for how a WLAN is configured and managed, which means network engineers must learn and apply new processes, tools and gear as they move from company to company, or even job to job.
Whether asked about technologies, processes, policies or practices, respondents’ answers in every case covered an incredibly wide range, leading us to conclude there is no one perfect way to design, configure, manage, and upgrade a WLAN.
While that conclusion seems like a no-brainer, the impact on a network engineering pro’s career progression isn’t. As noted earlier, only 20% of organizations represented in the survey have dedicated wireless teams; meaning, everyone on the team is expected to “know” wireless networking. But the ways different organizations approach WLANs are so varied, it puts individual network engineers at a disadvantage when they start a new job or take on more responsibility in an existing one.
For example, respondents as a group named more than 20 different vendors who supply their organizations with network infrastructure equipment. And when asked what vendors supply their troubleshooting tools, the group named more than 25 different companies. The implication for network engineers is that they will likely be required to learn all-new tools and technologies several times over their careers.
Compounding things further, when asked about the mix of Wi-Fi clients in their network environment, responses were all over the map. While just over one-third of respondents had no fixed clients, the other two-thirds did. Portable and mobile clients were more lopsided in terms of whether organizations had them or not, but the percentages varied greatly, affecting how network engineers approach their day-to-day responsibilities.
BYOD is another area of variation. When asked about how their organization’s BYOD policy affected their network, nearly one-quarter (23%) said they had no BYOD policy, while others alluded to specific security policies, a list of approved devices, use of a segmented network, or no personal devices allowed.
Either as a result of BYOD or equipping employees with multiple devices to access the network, many organizations will see the number of clients accessing the network climb above the number of employees at a particular site. However, respondents to the survey indicated a wide range of approaches to planning for this situation. While 7% plan for just one device per employee, a significant percentage (41%) plan for two or three devices. And at opposite ends of the spectrum, 10% plan for more than three devices, while 42% either don’t know how they plan for multiple devices per person, or they don’t plan at all.
The above data on BYOD practices and policies is worrisome, given the data from the IHS survey, in which responding organizations indicated they expected the percentage of user-owned devices accessing the network to grow from 39% to 47% percent (in addition to the increase from 29% in late 2014)7. This shows how just how pervasive BYOD has become.
Outsourcing practices are just as varied. For example, while 46% of organizations represented in the survey don’t outsource any work, the remaining 54% outsource at least one key function that contributes to institutional knowledge that would help internal teams better manage their WLANs, such as design (10%), verification and site surveys (16%), installation (23%), and troubleshooting (6%).
Among survey respondents, use of the cloud for managed Wi-Fi broke roughly 2:1 in favor of not currently or planning to use the cloud (68% vs. 32%). And for those who wouldn’t be using the cloud, 82% said it had nothing to do with any “no-cloud” policy.
그림 3: Impact of BYOD
그림 4: Devices per employee
Given the critical nature of WLANs to business transformation and the expectation of highly variable environments, standardized processes and checklists would make network engineers’ jobs easier while improving network performance
For anyone reading this document who is now shaking their head over the value of using something as simple as a checklist, a brief history lesson.
The year was 1935. The place was Wright Field in Dayton, Ohio. Three aircraft manufacturers were competing for a major military contract to supply more than 100 planes, but it was pretty much a done deal. The Boeing entry had already flown circles around the other two entries, and the military pilots who had tested it couldn’t wait to get their hands on the plane. This final demonstration was considered window dressing.
Except for one thing: when the airplane taxied down the runway with two highly experienced military pilots at the controls, it rose smoothly into the air, but then suddenly turned on one wing, fell and burst into flames on impact, killing two of the four people on board and one of their rescuers.
So what happened? The lead pilot neglected to release the elevator lock prior to take off, a simple thing to remember for such an experienced pilot. The military wasn’t quite ready to give up on the plane, so a group of pilots put their heads together and decided something was needed to make sure all steps for takeoff, in-flight changes and landing were followed in the correct order. And thus, the aviation checklist was born, saving the Boeing Airplane Company (as it was known then) from bankruptcy and launching the legend of the “Flying Fortress” – the airplane given much of the credit for helping end WWII in the European theater8.
Commercial aviation has followed suit, creating literally a notebook of checklists for each airplane model and then reviewing them after every major airplane incident. It’s no exaggeration when test pilots say there’s a dead body behind every step on the checklists they follow.
Another example of highly skilled professionals needing checklists to save lives: doctors.
In his book, “The Checklist Manifesto,” surgeon Atul Gawande writes how intensive care doctors and nurses have used checklists to save trauma victims, nearly eliminate hospital-acquired infections and improve the quality of care.
Every day, intensive care nurses (and doctors) perform an average of 178 actions on each patient under their care9. The potential for error is immense, with deadly results. The disparity in outcomes between hospitals that use checklists and hospitals that don’t has been proven through multiple studies. Hospitals testing the use of checklists saw dramatically improved survival rates, as well as greatly reduced infection rates and other hospital errors.
So why aren’t we hearing more about this miracle? Because so many hospitals and the medical professionals who run them resist the idea that something so simple would be required by the brilliant minds (and large egos) of surgeons and other doctors. It’s human nature to dislike being told what to do, how to do it and in what order, but the data is irrefutable.
If we were to apply the same lesson to the ever-changing WLAN environment and the network engineering pros who oversee them, the use of standard processes and checklists could potentially make their jobs easier, less stressful and less costly to their organizations.
For example, let’s go back to all those functions that may or may not be outsourced. If enterprises required their vendors to follow standard processes and guidelines for design, verification, installation and troubleshooting, the network ops team would have a firm grasp on their WLAN environment just by reading the documentation.
Given the large number of network infrastructure and troubleshooting tool options in use by different organizations (and not likely to change anytime soon), the need for standardized processes and checklists looms even larger. Newly hired network engineers would get up to speed more quickly. Recently promoted IT managers would have a better sense of the entire environment, not just the areas where he or she previously worked. All of these would combine to save the organization money and improve user experience at the same time.
That last point bears repeating. If network ops teams were to adopt the same approach as pilots and surgeons use to save lives; that is, using checklists to ensure processes are followed to the letter and in the right order, network engineers could do the same thing for their enterprise users who claim they would die without perfect uptime and performance.
Notes on Methodology
The survey was conducted in late March 2016 among more than 175 network engineers, representing mid to large enterprises in financial services, retail, healthcare, manufacturing, education, public sector, services and other industries. To qualify, respondents were required to be actively involved in managing their organization’s wireless networks.
The survey results are not based on a probability sample; therefore, no estimate of theoretical sampling error can be calculated. All decimals in this report are rounded to the nearest percentage point, which may result in certain numerical totals adding up to slightly more or slightly less than 100 percent. Results expressed as percentages are based on the number of respondents who answered that specific question.
1 IHS, Inc. “Wireless LAN Strategies and Vendor Leadership: North American Enterprise Survey” by Matthias Machowinski, October 2015
4 Wikipedia entry. https://en.wikipedia.org/wiki/Wireless_LAN, accessed 2016년 5월 28일.
5 IHS, Inc. “Wireless LAN Strategies and Vendor Leadership: North American Enterprise Survey” by Matthias Machowinski, October 2015.
8 New Yorker magazine. “The Checklist” http://www.newyorker.com/magazine/2007/12/10/the-checklist, accessed 2016년 5월 31일.