Predicting and proactively resolving service outages
Predicting or forecasting a service outage enables organizations to move from being reactive to being preventive. Being predictive means organizations can contain smaller issues before they become larger problems, prevent service outages, take timely actions, and focus on key value drivers. Predictive features help organizations to maximize the IT performance and save the time and cost required to manage IT environments.
The AI-based service-centric prediction in BMC Helix AIOpsconsumes service health indicators and metrics from multicloud and hybrid IT environments. It has the capability to predict service outages and helps in identifying and fixing issues before they happen. As a result, it helps reduce an organization's overall mean time to resolve (MTTR).
The following video (2:16) shows a high-level overview of service failure prediction in BMC Helix AIOps:
The Predictions page provides quick insights on forecasted service failure events. The prediction algorithm is capable of learning from past events to predict failures in the future. The algorithm uses the historical information to calculate a future service failure event based on the defined metric and identifies the threshold point at which the first impact will occur along with the predicted severity.
As time progresses, the algorithm continuously keeps learning and recalculates the status. Based on the recalculations, the availability status can change. This overriding of the previous status occurs at a preconfigured interval set within the algorithm.
Consider the following scenarios where you can use service predictions.
The following table lists the actions you must perform, depending on your role, to enable and start monitoring service predictions in BMC Helix AIOps:
Enable the service-centric predictions feature
As a tenant administrator, select the AIOps Service Predictions option on the Manage Product Features page.
|Enabling the AIOps features
Monitor service predictions
As an operator or SRE, monitor service predictions on: