I was sitting at my desk this afternoon, diligently working on this blog, when suddenly the ringing started and that familiar voice came over the intercom: “Please evacuate the building via the nearest exit. This is not a drill.”

In fact, it was not a drill. (Here in Canada you know it’s serious when the please comes before the request.) But luckily there were no flames either. The fire department came and checked everything out and determined that a pair of facilities contractors accidently tripped the alarm while working. No harm, no foul. In fact, we were treated to some sunny, brisk weather during our wait outside—both rarities for Vancouver this time of year.

During the orderly ingress, I was thinking about how many events that fire alarm kicked off. When the alarm sounds, there’s a protocol in place to get everyone safely out of the building and sensors to detect where the alarm originated. The fire department is quickly notified, which in turn deploys the people and equipment to fight the fire. Thankfully, we had a false alarm, but I’m confident that the existing systems would ensure nothing tragic happens to the building’s occupants.

Like our building’s fire protection plan, IT departments have many systems and procedures in place to deal with issues both big and small. Akin to the fire policies, these practiced responses have been tested and evolved over time to the point they’re at now—very good and very fast.

But what if IT organizations didn’t have to react to issues? What if the front-line response wasn’t the only option? What if we could use the data that IT is already gathering to peer into the future and make predictions?

This is the central thesis for a new Database Trends and Applications article written by Mike Flannigan, senior vice president of SAP Analytics and SAP Leonardo. The article outlines the principles that SAP IT Operations Analytics (SAP ITOA) brings to IT organizations, especially around predictive functionality. It’s time for IT to move away from being reactive and start becoming proactive.

Out-of-the-Box Predictive Functionality

SAP IT Operations Analytics was designed with predictive use cases in mind, which is why time-series forecasting is available right out-of-the-box. This style of forecasting, while lower on the predictive hierarchy of math, is attractive due to its ease of use, requiring little to no preparation of the data. For example, time-series forecasting can aid in capacity planning and management, which in turn provides justification for those always-shrinking IT budgets.

However, the real power of time-series forecasting is to eliminate static thresholding, transform alerting, and introduce automation into your IT organization. In the image below you can see the red dot at the top of a peak of alerts. The red color denotes an outlier based on the forecasted model and when an outlier is detected, an alert can be sent. These alerts could trigger the traditional (and infinitely inconvenient) email/SMS combination, be collected in the alerting dashboard, open a ticket in an external ticketing system, or execute a script which could take an action to self-remedy the issue. Since the predictive model is based on historical data, the outliers are personalized to your IT ecosystem rather than some suggested industry baseline.

Integration with SAP Predictive Analytics

For those who feel like getting a little more sporty with their data, other software can be used in conjunction with SAP ITOA to unearth further predictive use cases. Because SAP IT Operations Analytics is an SAP HANA native application, in addition to getting real-time insights into your IT landscape, the SAP HANA Predictive Analysis Library is at your disposal.

The IT data stored in SAP ITOA is also accessible to other SAP applications, such as SAP Predictive Analytics. In this case, the data can be brought into a model using SAP Predictive Analytics with the results then exported and implemented in SAP ITOA. This opens up a whole new realm of predictive possibilities, including predictive maintenance.

Bringing in Master Data

The IT organization may be the underlying driver of the digital transformation, but at the end of the day the business is still a business. How do we measure the effect that IT has on the business? With SAP IT Operations Analytics, master data can be brought into the system so that specific business functions can be related to clusters or individual pieces of infrastructure. This allows you to see just how much revenue that faulty switch is costing your e-commerce platform. But SAP ITOA is not limited to transactional data—everything from configuration management database (CMDB) to human resources data can be brought in to support your use case.

From our predictive maintenance demo below, you can see information regarding the probability of failure for a swath of servers. Besides the expected data about CPU, memory, and alerts, SAP ITOA is also showing the service level agreement priority of the stack supported by those servers. Having this information at-a-glance helps IT team prioritize which pieces of equipment gets serviced in what order. It also doesn’t hurt to know the key stakeholders, business users, and support team assigned so that everyone is informed when changes are made.

As Mike said in his article, predictive analytics is the key to shifting IT organizations from reactive to proactive behavior. With the time-series forecasting available out-of-the-box and tight integration with SAP Predictive Analytics, SAP IT Operations Analytics is the right tool to anticipate, predict, and avert IT issues. Start leveraging predictive functionality in your IT organization and leave the firefighting to those with the big red trucks. Speaking of, there’s that ringing again…

Learn More