Creates Package `@kbn/ml-anomaly-utils`.
- This moves some of the utility functions, constants and types related
to Anomaly Detection we previously exposed from the `ml` plugin itself
to a package `@kbn/ml-anomaly-utils`, resulting in ~5.9KB reduction of
the plugin's page load bundle size.
- To reduce increases in async bundle size for consuming plugins the
utils have been refactored into individual files to allow deep imports
and get some optimization through that. Some previously duplicated code
in consuming plugins has been deleted and replaced with deep imports of
the corresponding original code in the package.
- Types have been prefixed with `Ml`.
- Constants have been prefixed with `ML_`.
Created package via `node scripts/generate package @kbn/ml-anomaly-utils
--web --dir ./x-pack/packages/ml/anomaly_utils`.
- Extends the analysis API to return histogram data.
- Adds a column with MiniHistogram components to the analysis results table.
- Moves/consolidates fetchHistogramsForFields to @kbn/ml-agg-utils (also used in Data Visualizer and Data Grid Mini Histograms).
- So far fetchHistogramsForFields auto-identified the necessary interval and min/max. To be able to generate histogram data for the log rate spikes charts an options was added to use that information up front for the data to be fetched. This allows the buckets for the chart data for the overall (green bars) and the field/value-filtered (orange bars) histogram to have the exact same buckets.
- Introduces dedicated Kibana API endpoints as part of ML and transform plugin API endpoints and moves the logic to query and transform the required data from client to server.
- Adds support for sampling to retrieve the data for the field histograms. For now this is not configurable by the end user and is hard coded to 5000. This is to have a first iteration of this functionality in for 7.9 and protect users when querying large clusters. The button to enable the histogram charts now includes a tooltip that mentions the sampler.
* Closes#68485 by:
- use the ML plugin to query for all APM jobs
- inspect the ml job groups to find all jobs related to a particular service
- use the mlAnomalySearch client to get ml job buckets with the max anomaly score
- query for the model_plot buckets to obtain actual/median values for the ML description
- return the relevant ML job with the max anomaly score for a service
- indicate to the user that no anomalies were found for a service with an ml job
* - Use the anomalyDetectorsProvider jobs API rather than the search endpoint directly
- Defines a specific return type for the ml jobs api
- Update the empty anomaly data message
* Code and types cleanup
* Return to using record result type on anomaly queries. These are the
same values used in the anomaly explorer and it includes actual &
typical values which greatly improve performance of the previous query.
* - If anomaly data is missing show a gray border around node
* - moved AnomalyDetection out of service map Contents into own component
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
