Federator.ai

Multi-layer workload prediction

Using machine learning and math-based algorithms, Federator.ai predicts containerized application and cluster node resource usage as the basis for resource recommendations at application level as well as at cluster node level. Federator.ai supports prediction for both physical/virtual CPUs and memories.

Auto-scaling via resource recommendation

Federator.ai utilizes the predicted resource usage to recommend the right number and size of pods for applications. Integrated with Datadog's WPA, applications are automatically scaled to meet the predicted resource usage.

Application-aware recommendation execution

Optimizing the resource usage and performance goals, Federator.ai uses application specific metrics for workload prediction and pod capacity estimation to auto-scale the right number of pods for best performance without overprovisioning.

Multi-cloud Cost Analysis

With resource usage prediction, Federator.ai analyzes potential cost of a cluster on different public cloud providers. It also recommend appropriate cluster nodes and instance types based on resource usage.

Application Correlation and Impact Analysis

Federator.ai provides analysis and system recommendations based on the correlation between microservices of an application, providing insights about how individual microservices are impacted by external factors that affect the primary workload. This correlation is crucial for preventing over- or under-provisioning of resources.

Custom Datadog/Sysdig Dashboards

Predefined custom Datadog/Sysdig Dashboards for workload prediction/recommendation visualization for cluster nodes and applications.