Sankey

A Sankey chart visualizes how a numeric metric flows across ordered stages. Use it to understand movement between categories and spot imbalances.

A Sankey chart shows how a numeric metric flows across ordered stages. It highlights sources → stages → targets. Link width is proportional to the metric value.

It’s ideal for:

• Log or event flow tracking

• Data pipeline visualization

• Request routing

• Cost or traffic movement

• Namespace → app → destination mapping

What a Sankey represents

Each column represents one stage in the flow. Each node is a category inside a stage. Each link shows the amount moving between two categories.

• Node: a category at a stage

• Link: flow between two categories

• Link width: magnitude of the metric

• Color: grouping (depends on the field used by your data)

Example Sankey chart

Configure a Sankey chart

Set Chart type to Sankey. Use the Plot panel to define flow stages and the metric.

Sankey configuration (Plot panel)

Sankey chart options

Use these options to control how the Sankey chart renders.

Option	Purpose	Typical values / notes
Flow columns (ordered)	Defines the path data follows (left → right).	Examples: cluster → namespace → app, namespace → app → pod, service → endpoint → status_code. Order matters. Minimum 2.
Value column	Numeric metric that sets link thickness.	Examples: value, request_count, cost, bytes. Duplicate paths are aggregated.
Apply	Applies changes to the widget.	Changes do not render until you click Apply.

Key rules:

• Use at least 2 flow columns.

• Column order defines flow direction.

• Value column must be numeric.

When to use a Sankey chart

Use Sankey charts when you want to:

• Understand movement between systems or states

• Identify bottlenecks or dominant paths

• Detect imbalances (skewed flows)

• Explain pipelines and routing

They’re best for composition and routing analysis, not trends over time.