Sliding window aggregation: Refined control over alert monitoring
Blog post from New Relic
New Relic has introduced sliding window aggregation, a new method designed to enhance the management of alerts by smoothing out visualizations and detecting infrequent or erratic signals, such as latency and errors. Users can set alert conditions utilizing NRQL queries and advanced signal settings, which offer more precise control over data signals that may be intermittently missing. The sliding window aggregation allows data to be grouped into overlapping time windows, facilitating trend detection by smoothing the chart lines and reducing noise. Users can specify time periods from 1 to 120 minutes, and adjust aggregation settings through an intuitive mechanism, allowing for alerts to focus on significant trends. To implement this feature, users can adjust advanced signal settings and set slide-by intervals in the New Relic platform, although it is important to build an aggregation buffer to ensure the alerts function correctly. The blog post encourages users to explore this feature, offering a video demonstration and additional resources for those interested in leveraging sliding window aggregation for more effective alert management.