Ingestion Policies and Per-Policy Usage Monitoring
This release supports ingestion policies.
Note: Wavefront is performing a staggered rollout of this feature. Even if your Wavefront instance is updated to 2019.38.x, it’s possible that ingestion policies are not yet available.
Ingestion policies let you find out which accounts (user or service accounts) are responsible for which part of total ingestion.
- Admins with Account Management permissions can create one or more ingestion policies and add accounts to each policies. To avoid duplication, each account can belong to only one policy.
- All users can view the ingestion breakdown dashboard, that we added to the Wavefront Usage integration. This preconfigured dashboard:
- Includes Usage by Ingestion Policy and Usage by Account and helps you explore usage patterns.
- Allows you to track usage history.
- Supports export to CSV.
Drag-and-Drop Dashboard Layout
Our v2 UI supports drag-and-drop dashboard layout. If your cluster has v2 enabled, you can switch UI versions from the gear icon.
|An improved drag-and-drop UI allows you to optimize dashboard layout. You can drag up to 12 charts into 1 row.|
Chart Improvements (v2)
Our v2 UI has several dashboard improvements in this release.
Faster Chart Loading with Sampling
To support faster loading of charts, this release supports a Sampling preference for each user that you can override for each chart.
- When Sampling is on for a chart, we display a maximum of 100 time series.
- You can toggle Sampling with a global preference or from the Sampling pulldown menu on charts in edit mode.
Drilldown for Single Stat Charts
Single stat charts now include a Drilldown Link tab that allows you to specify a target dashboard. You can optionally specify settings for the target dashboard. For example, you map the value of a point tag in the source dashboard to a variable value in the target dashboard.
This release makes several histogram functions publicly available. Click the links for details and examples.
- The cdf() function is a cumulative distribution function for histograms. You can use it to find out how likely it is that a histogram is less than or equal to a specified value.
- The cumulativeHisto() function allows you to visualize cumulative histograms, such as those coming from Prometheus or Telegraf, in Wavefront charts. Use other query language functions, such as
percentile()to manipulate the output of
- The histo() function returns a histogram distribution for a time series. Note: Send you histograms directly to the Wavefront proxy if possible for best accuracy. Use this function otherwise.
Other Query Language Changes
timespan()event management function creates a single synthetic event with the specified start and end timestamps. It now supports specifying a display name and tags.
Intelligent Trace Sampling
Starting with this release, Wavefront automatically performs intelligent sampling to reduce the volume of ingested traces. Using proprietary algorithms, we retain traces that are likely to be informative, and discard traces that are redundant or otherwise not worth inspecting. You can use Wavefront intelligent sampling in conjunction with explicit sampling strategies.