Metrics and the Metrics Browser

Wavefront provides observability for several different metric types including time series metrics, histograms, and traces/spans. This page looks at the anatomy of a time series metric and shows you how to explore it in the metrics browser.

Video

In the following video, Wavefront chief architect and co-founder Clement Pang explains why the concept of cardinality is so important for observability, what high cardinality means, and why Wavefront deals so well with high cardinality input.

You can also watch the following videos to learn more about metrics in Wavefront:

• Tagging your Data with Wavefront
• Time Series and Interpolation
• Getting Data Into Wavefront

Wavefront Time Series Metric Structure

A Wavefront time series has, at a minimum, the metric name, value/timestamp, and source. In many cases, the metric is ingested with additional information represented as tags.

Simple Time Series

Here’s one example that shows the minimum elements of a time series.

Each time series is a unique combination of:

• Metric name–Describes the metric. There’s often a hierarchy of metrics, each with a corresponding time series.
• Value & Timestamp–Value at the specified time.
• Source–The source of the metric. Host, VM, etc. In contrast to some other observability platforms, this dimension is always part of the metric.

Here’s a screenshot of the time series that is shown in the diagram above in a Wavefront chart.

Time Series with Tags

In most cases, the time series includes one or more tags to allow a more fine-grained analysis. The Wavefront ~sample data, for example, include point tags for environment and availability zone.

Point tags offer a powerful way of labeling data so that you can slice and dice it in almost any way you can imagine. For example, you can use point tags, to label a point’s datacenter, version, etc. and can then group by datacenter or version.

You use point tags to add extra dimensions to your data, and can then focus your exploration just on that dimension.Fine Tune Queries with Point Tags explains how to use point tags.

Here’s a screenshot of the time series that includes point tags in a Wavefront chart.

How Filtering with Tags Improves Usability

How the point tag filters are useful becomes obvious when the source= filter is removed. The result of all time series for ~sample.disk.space.used is visually confusing.

When you add filters for env and az, the information makes sense.

Metrics Browser

Select Browse > Metrics to display the Metrics Browser. Use the Metrics Browser to find metrics that sent at least one data point within the last four weeks.

To make search easier, you can

• Drill down and go up the hierarchy.
• Filter by source.
• Hide and redisplay metrics or groups of metrics.

Examine Metrics

To examine metrics Select Browse > Metrics Select folder icons to drill down to individual metrics. With a metric selectd, click Info to show sources and point tags for that metric. Click the metric name to show a chart with that metric.

Hide and Redisplay Metrics

You can manually hide metrics from the Metrics browser. Those metrics become unavailable from the auto-complete menu as well. Manually hiding metrics does not permanently delete a metric or metric namespace.

To hide one or more metrics: Select Browse > Metrics Click the Manage Hidden Metrics button In the dialog type a complete metrics name (e.g. requests.latency) or a metric prefix (e.g. requests., cpu.loadavg.). This field does not support auto-complete, so you have to type the entire metric name or metric prefix. The text is case sensitive. Wildcards are not supported. The star * character is considered part of the text string. Press Enter to add the metric(s) to the list and click Save.

To view hidden metrics: Select Browse > Metrics Click the Manage Hidden Metrics button. Click the Unhide button to the right of the metric or metric prefix to unhide and click Save. The selected metrics and metric prefixes appear again as long as they have had at least 1 reported data value in the last 4 weeks. Otherwise, these metric/metric prefixes are considered obsolete metrics and Wavefront hides them. You can show obsolete metrics for individual charts or alerts.