By Cassandra Balentine
Enterprises rely on many complex processes for daily operations. The ability to utilize data to create a holistic view of an organization, pinpoint root issues, and predict potential concerns before they become problematic is increasingly important. Enterprises need the ability to integrate IT and business operations to provide a better overall customer experience. The challenge IT operations face with managing data makes this a difficult goal to attain.
The complexity of IT environments requires a solution that enables the ability to analyze all types of data from all sources within IT operations, an area where traditional tools seem to fall short.
Today, a variety of solutions are being introduced. This includes IT operations analytics (ITOA) and advanced operations analytics (AOA), which applies machine learning over multiple sources on enterprise networks—including devices, routers, and application interfaces—to correlate similar data patterns and recognize patterns to flag as possible risks. Although some of these ITOA solutions continue to be introduced to solve these challenges that IT operations face, the search for the ultimate solution is still ongoing.
Sancha Norris, senior product manager, operations analytics, Hewlett-Packard (HP), says these ITOA solutions should be able to derive meaningful insights that can be used to improve IT performance and service levels, IT worker productivity, the end-user experience, and overall business performance.
In a 2014 report published by Enterprise Management Associates (EMA), The Many Faces of Advanced Operations Analytics, the firm describes AOA as big data used by IT for a variety of use cases. “These include optimizing service performance, minimizing security issues, managing change, and optimizing capacity across internal IT and the extended enterprise—including partners, service providers, suppliers, and ultimately governing IT more effectively as a business in support of the business it serves.”
EMA notes that while these are separate areas with separate constitutes and traditionally separate markets, its premise is that, “effective AOA investments are beginning to blend those values and promote more effective ways of working across IT.”
Sasha Gilenson, CEO, Evolven, explains that while today’s ITOA solutions can reveal important information about specific areas of data, ITOA tools are still limited, constrained by operating in narrow silos—application performance management (APM), log, network, etc., concentrating on just the symptoms that surround issues and hindered by their own limited analytics capabilities.
“A single pane of glass has turned into a correlation at the visual level but not the data level. Now is the time for the next step,” says Gilenson, “to finally realize the benefits and promise that ITOA offered. IT stakeholders, like those responsible for stability, performance, and security of business systems, need to overcome the limitations of their silos and apply a blended analytics approach that collects data from the various dedicated silo tools across the environment, then correlates and analyzes this variety of data to extract useful insights. A blended analytics solution can proactively detect potential performance, availability, and security issues to provide automated root cause analysis for quick and effective prevention and resolution of issues.”
The amount of data generated, collected, and analyzed is constantly increasing. This offers the potential for valuable insight, but also presents operations challenges for IT.
“Not all data sources are created equal. That means that in order to understand the overall picture we need to start looking at data with relation to symptoms, IT context, and changes, and the role each data stream plays within the overall picture,” says Gilenson.
Setu Kulkarni, director, product management and strategy, TIBCO, sees the use of ITOA/AOA information generally falling into three categories, driving real-time customer engagement, optimizing IT platforms for service level agreements (SLAs) and cost efficiency, and making innovation happen faster.
Two primary drivers for investment into ITOA include the need to improve IT efficiency in a dynamic world and the demand for a more timely view into business activities and results.
Kulkarni explains that old-school ITOM tools do not address the dynamic and elastic world adequately. “IT organizations are looking at what industry leaders like Netflix are doing and try to adopt similar techniques for their organization.”
While IT analytics have existed for some time, the new challenge is to address the growth of machine data and unstructured data—such as social media and the Internet of Things (IoT), which now play a large role in transforming IT, explains Norris. “There is a tremendous opportunity to combine IOTA with the power of big data technologies that are capable of harnessing and analyzing all types of IT data, both structured and unstructured. This is the new ITOA and the market for it is just beginning.”
The potential of this data, when properly leveraged, is a differentiator for enterprises. “Companies with more data and insights will easily outdo their competition. Hardware and software vendors have made technology advancements that allow enterprises to easily and cost effectively collect and analyze data. Running a software-defined business and operational visibility is akin to running a physical retail business not knowing what’s on the shelves, what’s in inventory, and where the shipment trucks are and their estimated time of arrival. Not capturing every possible data point and not making data-driven decisions is just a recipe for disaster,” says Maneesh Joshi, senior director of product marketing, AppDynamics.
Kulkarni says ITOA/AOA tools are already starting to replace some data warehousing infrastructure components, especially for marketing and sales analytics.
Digital business is transforming the way IT organizations need to support business objectives. Norris explains that while downtimes have always been costly, in a digital world it becomes more impactful to the business in terms of both magnitude and speed. “Digital business comes with an increasing amount and complexity of data—traditional structured data will now be supplemented with unstructured data. This means more data will be accessible to IT than ever before.”
Large communication gaps between IT and the business ITOA/AOA hinders the ability of IT to share actionable insight in business terms. “Today’s tools lack the depth needed to help bridge the gap and remove departmental silos to ensure everyone has the necessary context to make the best business decision,” says John Miecielica, director of product management, TeamQuest.
“Businesses are run by software—it’s become a critical backbone to deliver value for businesses old and new,” adds Bruno Kurtic, founding VP of product and strategy, Sumo Logic. All of these applications generate data, and applications now touch customers more than any other vehicle. “To monitor and assure that critical software is available, ITOA tools are a must have, not a nice to have.”
“After many years and billions spent on various tools, IT organizations still suffer from performance and availability issues, wasting valuable resources in complex and lengthy troubleshooting,” says Gilenson. “IT still relies extensively on a war rooms practice to respond to these severe issues. So why do these challenges still persist in spite of all the investment and attention? The answer boils down to a critical roadblock that IT has not succeeded in removing—lack of situational awareness.”
Gilenson laments that one of today’s best practices for investigating incidents is to get as many of the smartest IT specialists around, together into one room. But they then each go to work on their separate silo areas. “This is the situation where ‘I watch application performance metrics, you dig the infrastructure logs, and she investigates network KPIs.’ The experts analyze these symptoms around the issue and try to reverse engineer from these symptoms. To find their way to the actual root cause, they attempt to manually crunch the available data and iteratively try educated guesses. But they are missing the true root cause—the changes.”
Patrick Hubbard, technical product marketing director, SolarWinds, says businesses have discovered that first to market with new services is more important to success than traditional IT cost management. “Once cost was the primary driver, but increasingly, flexibility is the chief advantage,” he explains. “Smart businesses are moving to transform IT into opportunity hubs and ITOA is central to realizing that opportunity,” he adds.
Toufic Boubez, founder/CTO, Metafor Software, predicts the next few years will mark the convergence of metrics analysis, log analysis, and the advanced machine learning frameworks that have—until now—been restricted to academia. “This convergence will bring context aware, intelligence monitoring that automatically adapts to dynamic environments, making guesswork and static thresholds—along with the corresponding false positives and alarm fatigue—a thing of the past.”
This will free IT operations from the common scenario of having to keep an eye on hundreds of graphs on screens in an attempt to stay on top of all the data they are collecting. He says it also brings automated alert correlation capabilities that will end up replacing much of the investigation work.
Drit Sulijoti, co-founder/chief product officer, Catchpoint Systems, suggests the biggest driver for enterprises selling their services and products is increasing profit margins by lowering the cost of goods sold, including IT operations costs. “For other enterprises, it’s simply increasing productivity by eliminating the issues that impact it—like outages or slowness in an application key to the employee. With smart, strategic monitoring, IT departments can drastically reduce the amount of time they spend identifying and resolving problems, thus increasing the company’s profit margins. At the end of the day, it’s all about changing the perception of IT being a cost center to a value center.”
IT for the Business Case
With the continued availability of automatically collected and correlated real-time data, enterprises are presented with an unlimited number of possibilities. Miecielica notes that a core function not everyone gets correct is extracting the right data from the infrastructure to provide an ROI for the business. “How does it give context to data that is captured, reported, and acted on by the business? I’m talking about tying IT metrics with business metrics in a meaningful way and applying analytics across a broad set of data to transform data into knowledge and wisdom so the business can make the best decisions possible. The better the data, the more accurate the business analysis, and the more successful the business outcome.”
Joshi points out that enterprises can leverage this data to easily answer questions that connect the dots between business and performance data.
In a business case scenario, Joshi offers the example of an online retailer’s ability to segment data to understand which products or features are the most popular. They can also quickly identify how much revenue was lost in the last hour due to failed in-flight transactions that may not even have hit the back-end database. They can go a step further to identify which specific users were impacted and individually offer them a discount as part of the win-back campaign.
Customer-facing applications are more prevalent in the digital business era. Therefore, there is more on the line in terms of customer usability that could make or break a sale.
With ITOA/AOA, IT can tie business performance data to better prioritize trouble shooting of IT performance bottlenecks. “The ones having higher revenue impact obviously go to the top of the priority queue. IT can also use system performance data for better capacity planning or better resource management,” says Joshi.
This information is valuable across a variety of industries, and not limited to ecommerce applications.
“ITOA truly gives enterprises the information they need to have an accurate view into their organization’s IT operations, enabling them to detect issues and solve or mitigate them before they affect the company or their customers,” says Kevin Conklin, VP of strategy and marketing, Prelert.
Enterprises struggle to identify operations issues quickly and identify the root cause of failures. Rob Markovich, CMO, Moogsoft, says that while many organizations capture application logs, systems logs, cloud stats, database stats, network stats, storage stats, and other forms of data, they must sift through it to find what is really important and actionable in real time. “ITOA tools can help discover trends and relationships between IT systems and components to support identification and resolution of these issues more quickly, but are focused on a historical approach.”
This information can be used to better understand normal patterns in an organization’s network, and by the same token, what activities do not fit into normal patterns and need to be looked at more closely. “Instead of only looking for typical issues or commonly repeated problems that a programmer thought to make the system look for, the use of advanced analytics enables organizations to find new problems or issues that haven’t been seen before,” says Conklin.
Using performance metrics, IT can also hold partners accountable to their API SLAs, as they now have concrete data on which individual API invocations violated SLAs,” says Joshi.
The potential for ITOA/AOA is promising and the role of ITOM continues to evolve.
“Every company, business, and organization depends upon the ability to connect critical data and information without interruptions or downtime. Cybercriminal and hacker activity has increased at the same time, and there is always someone looking to get into a network. ITOA can change the game for IT executives, giving them the ability to provide stronger networks for their employees and customers,.” says Conklin.
As businesses become software driven, application intelligence is going to be a mandatory requirement for survival. Joshi recalls the earlier example of the online retailer, “you could quite easily apply those customer visibility use cases to retail banking, wealth management, and healthcare. You think of an industry and you’ll find a value in application intelligence.”
“It’s time to break the vicious cycle that IT operations has been stuck in, we need to separate from the notorious, old war room practice of pulling together all of your IT silos for guesswork and blame,” says Gilenson. “A blended analytics approach that correlates and analyzes cross silo data sources can provide the necessary insights to prevent and rapidly investigate incidents.”
Gilenson emphasizes that although change is a critical component and one of the primary sources of operational issues, it has been surprisingly overlooked. Blending and analyzing multiple sources of data within a focus on change will give quick access to actual and potential root causes and accelerate resolution.
“Given the way IT environments are evolving into dynamic, software-defined infrastructures and more layers of virtualization are added to obfuscate what is the root cause of problems, an automated, analytical approach is the only way to manage and support these environments going forward,” says Markovich.
Kulkarni suggests that when ITOA/AOA are viewed primarily as new tools for optimizing IT efficiency, the market is reasonably sized at a couple of billion dollars. “But, when businesses realize that ITOA/AOA can be leveraged to improve customer engagement and expertise to grow revenue, the investments and number of use cases will sky rocket. Emerging market segments like IoT and digital business—where the lines between technology and consumers are obliterated—will lead the way.”
Miecielica agrees. “Once the business value is established, we believe the C-Suite and business leaders throughout the enterprise will mandate the use of ITOA/AOA. There has been a lot written about the pitfalls of letting the data lake you create swiftly become a data swamp. We advocate that you measure and monitor what matters most on business and IT sides, and not try to ‘boil the ocean.’”
“ITOA as a defined market, at least in the standalone sense, may have actually decreased slightly, just as SDN, SDS, and some other technologies have. But like SDN, ITOA’s opportunity is actually accelerating through embedded or reseller adoption,” says Hubbard. “Many established IT vendors are adopting ITOA technologies and methodologies and bundling them into products. ITOA enhances existing product capabilities, provides market advantage, and enables new products focused on IT intelligence or breadth of control.”
“In an ideal world, ITOA becomes such a ubiquitous component of network and systems management that enterprises demand comprehensive solutions with mature analytics abilities truly tailored to IT,” adds Hubbard.
He explains that it won’t be enough to simply be a data collection platform, or even a metrics dashboard solution. With continued convergence and increased complexity, systems need to analyze more data, transforming it into concise, useful information.
In addition to the overall operations environment, subsets within an enterprise benefit from ITOA/AOA.
Modern websites integrate a growing number of externally hosted Web services, including analytics, advertisements, security, video, social integration, and content delivery networks.
Sulijoti sees Web performance monitoring as a subset of the larger ITOA market, and has a large growth potential. “In the past few years, the market evolved from a focus on traditional desktop websites and Web applications to also include mobile sites and applications. But everything is moving towards online systems, to what’s known as the IoT. So in a few years, we can expect to see even more of our daily lives—from our computers and phones to the very homes we live in—dependent on online services. All of those things will have to be monitored smartly and effectively in order to deliver on the services they promise.”
Modern websites have drastically evolved in recent years, transforming from static HTML sites to rich composites comprising numerous third-party tags and content delivery services that transform the user experience. Sulijoti points out that as websites and applications continue to develop, any poorly performing component of a Web infrastructure can drag down performance. “Today’s website management teams need to closely monitor all the components and infrastructure compromising their sites. But sometimes, these comprehensive monitoring programs produce too much of a good thing, yielding tons of data but no insights.”
He says this often results in a wasted effort for IT staffs. “They know there’s a problem, but now they have to play detective and track down the cause, which may not even be rooted in their domain. This is like trying to find a needle in a haystack—a needle that may not even be there.”
Boubez sees particular opportunity in security analytics. “Enterprises are starting to realize the need for genuine machine learning capabilities in their products. They demand these capabilities because they recognize that they’re missing the ‘unknown unknowns,’ issues never seen before and can’t afford to continue missing.”
He says the most common solution that enterprise have turned to is a band-aid approach—eyeballs on screens. “Humans are pretty good anomaly detectors, but they don’t scale well and they’re typically not that thrilled about staring at lines on a screen all day. As a result, that sometimes mind-numbing role of monitoring network and security operations by watching a wall of charts tend to break down at scale, whether due to missed anomalies or even employee turnover issues.”
He suggests a better solution is layering some automated, real-time machine learning intelligence on top of an analyst team. This approach is why Boubez sees many traditional APM/log management vendors and enterprises working or collaborating to develop effective machine learning algorithms.
David Rizzo, director of product development, Compuware, suggests that ITOA can also enable better IT infrastructure utilization and optimization, more specifically for mainframe systems. “Inefficient, faulty mainframe application code can negatively impact organizations in two primary ways. First, given the sheer number of transactions affected, slight tweeks in mainframe application code can yield improved application processing time for millions of end users, driving customer satisfaction and revenue gains,” he continues.
Rizzo cites a Rubin Research analysis that shows mainframes are often more cost effective than commodity servers handling massive increases in computing loads. “For example, Rubin compared two banks and found that one bank that relied on commodity servers had 30 percent higher infrastructure costs than its mainframe-heavy peer. This is because the mainframe has decreased in unit cost enough to offset changes in computing volume—more so than commodity servers.”
He says mainframes are a tried-and-true platform for handling these huge transaction loads, securely and reliability. “According to IBM research, the new z13 mainframe is capable of handling 100 cyber Monday-size transactions everyday, 365 days a year. However, mainframe licensing costs often consume up to 30 percent or more of mainframe budgets. One key to keeping these costs in check—and enabling mainframe users to truly achieve the price/performance benefits of their systems—is ITOA solutions,” says Rizzo.
Success Through ITOA
As businesses continue to rely more heavily on automated processes and customer-facing applications, it is important that IT react appropriately and effectively. As more data is generated and more systems are incorporated into the network, emerging uses for ITOA and AOA simplify complex environments in a way that IT can gather meaningful insight. SW
Jul2015, Software Magazine