Find the service behind the alert.
Read service calls, remaining error budgets and latency together. Follow one service across the layers to investigate a specific relationship.
Premium Invented demonstration data · Library 1.1.5
How to read this example
- 01
Start with a relationship.
The centre matrix reads row → column. Pick a service call and compare it with the reverse direction in the tooltip.
- 02
Check the budget.
The surrounding goal rings show the remaining 30-day error budget for each service. The linked selection keeps the same services in focus.
- 03
Place the latency in time.
The outer bars show p95 latency per service in four-hour windows. Compare their timing with the service context before investigating the cause.
Try thisHover Matching → Geo index to compare the call and its reverse, and follow both services into the surrounding rings. Arrow keys move through the matrix.
Data requirements
- A service-to-service call matrix in requests per second.
- Remaining error budget for each service, using the same service keys.
- p95 latency in milliseconds per service and four-hour window.
The example configuration shows the data shape, link keys and required modules. Inspect the setup →
Read the chart data
Open the example’s Agent view to read its actual chart.describe() model as text, a table or JSON. In multi-chart examples, choose the view you want to inspect.
The optional describe module is loaded on demand. The website does not calculate its own substitute totals.
Use this configuration in your project
Inspect the configuration, connect your data, and choose the rights that fit your application.