sommai - Fotolia
IT environments are increasingly complex, generating a plethora of data in the form of real-time network metrics, log files and more. IT environments also undergo constant change. This combination of complexity and frequent changes can make troubleshooting difficult and time consuming, and in some cases make identifying root causes nearly impossible.
At the same time, networking pros need to be able to respond quickly and effectively to performance problems and outages. And they need to predict potential issues and automate fixes before users and the business are affected. They also need an IT operations tool that automatically processes the mountain of operational data into relevant, useful information.
This is where IT operations analytics (ITOA) comes into play.
IT operations analytics explained
ITOA gathers and analyzes data from many different sources within an organization, providing insights that help management make more informed decisions regarding IT operations and business services. At its foundation, ITOA employs big data principles and technologies to automate the collection and analysis of IT operations data, thus allowing the identification of patterns and problems.
Large IT environments typically use network performance monitoring (NPM) and application performance monitoring (APM) tools. While NPM tools monitor network and server performance, along with other infrastructure components, APM tools focus more on applications and transactions. But to determine how an application might affect system performance, and whether an infrastructure issue is hamstringing users from accessing an application they need, networking pros need to know what's happening in both these areas -- and right away.
IT operations analytics removes most of the manual processes involved in correlating infrastructure changes to application performance, automatically pulling data together to provide a real-time view of systems and applications. This view, along with anomaly discovery and pattern identification, greatly simplifies troubleshooting and resolution.
With ITOA, networking pros can also predict and proactively address issues. For example, a pain point for networking pros is meeting key performance indicator (KPI) goals. Failing to detect and address critical issues before they affect users or the business can lead to bottlenecks and slowdowns, or worse. Because ITOA typically provides predictive analytics as well as baselining and trend analysis, networking pros can better manage or outright avoid resource spikes and availability crunches.
How IT operations analytics works
ITOA software can either be administered on-site, or accessed via a cloud provider. The software gathers data from running (live) systems, including operating systems, hypervisors, network devices and sensors, as well as data from various types of logs, such as applications, databases, devices and web. The data is warehoused and indexed in a data store.
From there, ITOA normalizes and transforms structured and unstructured data, turning it into usable information. The data may include different kinds of KPIs as there are active network connections and server CPU utilization, application usage patterns and user responsiveness metrics among them. Many ITOA tools "learn" what's normal for a system, network or environment, creating baselines and then identifying patterns in the data to detect anomalies, especially in log data and network captures.
Finally, IT operations analytics software sends alerts to a management console that recommends actions for administrators to take to resolve issues. A real-time alert might be issued when a KPI exceeds a threshold or strays from an expected norm. Most ITOA products prioritize issues as well, allowing network administrators to focus on critical problems and return to lower-risk problems as time allows.
IT operations analytics features to look for
ITOA is still evolving, and there is no standard model against which to evaluate vendor offerings. However, many vendors have entered the ITOA market, catering mainly to enterprise organizations as well as some midsize entities.
When evaluating ITOA vendors, look for the following features:
Cross-domain collection and reporting. A comprehensive IT operations analytics tool should be able to search for and gather logged and real-time data across an organization's infrastructure -- physical, virtual and cloud. That information should be automatically correlated with events, application groups (groups of users and computers) and so on, and should then provide alerts and actionable reports to remedy performance and availability issues.
Baseline establishment. The tool should have the capability to establish system, network or environment baselines from historical trend data.
Automated log analysis and predictive analytics. This includes automated anomaly detection and behavioral analytics based on machine learning. Predictive analytics uses self-learning behavior derived from past events to predict when problems may occur in the future or to predict future trends for use in capacity planning and resource allocation.
Scalability. If the volume of your data increases, your ITOA software must be able to handle the new load without any downtime or major upgrades.
Ease of installation and use. Look for ITOA software that can be up and running in minutes or hours rather than days or weeks. Dashboards should be clear, concise and customizable, enabling efficient troubleshooting and root cause analysis.
Be aware that one vendor's product may be an all-in-one package, whereas another vendor may offer separate modules for specific functionality, such as log analysis and predictive analytics.
The bottom line
Now that you know what IT operations analytics is and how it works, you're ready to talk to vendors about their products.
Have a current systems and application inventory for your environment on hand that you can refer to during discussions or share with vendors. Ask each vendor for a list of all data types, networks and platforms its product works with. Most vendors attempt to cover a wide variety, but they can't guarantee compatibility with every type of data or system in use, especially legacy versions.
To get the most out of the product evaluation process, create a few cases specific to your organization that the vendors must address. These cases should summarize a business or technical issue you're facing and need to resolve. Approaching product evaluations with a certain scenario in mind lets you compare feature sets much more accurately.
Also, be sure to run trials in your own environment rather than simply going through a vendor's demo. Every environment is different, which can affect the effectiveness of an ITOA product.
Apply network analytics to quickly solve service issues
Learn how operations analytics are advancing
From ITOA and on to AIOps, and the coming generation of network analytics