With a combination of physical, virtual, and cloud infrastructures, and legacy and modern technology components, today’s enterprise IT environments are complex and dynamic. This technology mix changes rapidly with the transition to agile processes and automation of infrastructure setup, application deployments, and workload distribution. There are lots of changes taking place across too many touch points.
Traditional approaches have focused on establishing defined processes and taking on issues as they happen. But modern approaches have replaced manual processes, with automation addressing scalability and resilience issues through elastic resources. However, these approaches rely on all activities being executed through the automation platform.
To deliver stability, performance, and secure IT operations, staffs often take shortcuts, circumventing manual and automated processes. This results in issues caused by application defects and infrastructure failures. Adding to this complexity, users generate workloads that are not accounted for by the business systems.
Nevertheless, IT departments still need to be in control of their environments with tools in hand that allow them to diagnose and pre-empt incidents and problems. Unfortunately, most of the existing IT tools do not collect enough data and do a lousy job of handling the data that they do collect. Event dashboards are cluttered, configuration management databases (CMDBs) are frequently out-of-date, and alerts are ignored because there are simply too many of them.
Using IT Big Data Analytics: Make it Actionable
A typical organization’s IT environment produces terabytes of data, comprising everything from system metrics, change records, log events, and a variety of other operational data types. It is possible to get extremely extensive raw data describing the current state and history of an IT environment. The challenge is being able to use this data and make it actionable.
Critical to success is doing this rapidly, with no overhead to IT staff and IT environments. Because IT is expected to drive more activities with fewer resources, IT specialists do not have months to invest in rollout, training, and use of a typical enterprise solution. On top of this, the same specialists also are busy with fighting fires on a daily basis, leaving them with even less time for a proactive approach to operations management.
IT Operations Analytics
A solution comes with the rise of IT Operations Analytics (ITOA). The only way to make IT’s big data usable is by blending and correlating this data, automatically generating actionable operational insights. This means starting with as much data as possible to avoid missing critical details, then narrowing down the data to conclusions that matter. Business has already taken this approach, effectively managing business operations in real time. Ironically, IT is supplying businesses with the necessary big data analytics technologies for this, yet, as usual, IT itself is lagging behind the business that it serves.
In the past, with increased pressure to make use of growing piles of unstructured data, the business side of the organization found itself confronting a similar state as has been described for IT. Current business trends center on tackling big data with technologies for managing and analyzing large, diverse data sets. These Business Analytics tools process any amount of data in any format from anywhere, and have the ability to correlate the data to provide business managers with new insights to help run their businesses.
Yet while Business Analytics gives a platform to slice and dice the data as users want, ITOA must rely on domain understanding of data to provide meaningful insights, giving IT operations teams visibility and insight into the behavior of business systems, automatically identifying and isolating critical events (such as changes) that have the potential for disruptions. Using mathematical algorithms and domain-specific heuristics, ITOA provides immediate awareness for potential issues, facilitating a rapid understanding of these issues from the sea of raw data collected by management and monitoring technologies, helping IT operations to determine the best course of action to restore or meet performance and availability expectations.
IT must focus on the changes themselves, which still remains a blind spot for IT operations. Each time a change happens in an application, infrastructure, data, or workload, business systems are exposed to risk. Changes remain a significant source of operational issues.
With change serving as a focus for analysis, IT specialists can effectively correlate data on performance, availability, security, and other types of data, identifying actual causes and potential issues. Applying such techniques as complex-event processing, statistical pattern discovery, behavior learning engines, topology mapping and analysis, and multidimensional database analysis, ITOA can make use of the terabytes of operations data in real time, spotting and presenting issues in an understandable context.
By applying an ITOA approach, IT operations teams can stay on top of how changes are distributed and what risks they introduce. They can review the frequent daily changes, including unplanned and unauthorized activities.
Sasha Gilenson is founder and CEO of Evolven Software, an expert leader in IT Operations Analytics (ITOA). Prior to Evolven, Sasha spent 13 years at Mercury Interactive, participating in establishing Mercury’s SaaS and BTO strategies. Sasha studied at the London Business School and has 15+ years of experience in IT operations.
Subscribe to Data Informed for the latest information and news on big data analytics for the enterprise.