Updates to AI Genie Forecasting
Our AI Genie forecasting engine now uses hypothesis testing and neural networks for prediction. Results are more reliable and visuals are compelling.
Distributed Tracing No Longer in Beta
For several releases, we’ve fine-tuned our distributed tracing functionality based on internal and customer feedback. The feature has exited Beta and is now considered General Availability.
The new Tracing Best Practices guidelines help you get optimal results.
Improved Wavefront Monitoring
This release includes some handy tools for monitoring your Wavefront instance.
Updates to Spy API
New wftop Tool
Wavefront Top (wftop) is an open source interactive tool for examining the shape of the live metric ingestion data and for exploring which metric namespaces were used in the last X days.
The github page has a complete list of features and sample screenshots.
New and Changed Functions
This release includes new histogram functions and other improvements including a complete set of string functions.
New Histogram Manipulation Functions
This release includes two functions for making histogram visualization easier:
summary()function returns a summary of a histogram.
alignedSummary()function returns an aligned summary of a histogram.
For both functions, we return the following information by default:
max, P999, P99, P90, P75, avg, median (P50), P25, and min
You can specify a list of percentiles instead.
Forecasting, String Manipulation, and New Parameter for at()
- New forecasting function,
nnforecast(), which uses hypothesis testing and neural networks for prediction. This function is used by the AI Genie forecasting UI.
- New string functions.
- For the
at()function, new parameter
lookbackDistance. When you call
at(now), we look back by 1 hour by default. You can now specify a time period instead.