See what depends on what
Services come from your traces, logs and metrics. Connections come from calls between services in your traces.
Platform
Draw your services and the calls between them.
Topology shows each service as a node and each call between services as an arrow. Nodes carry request rate, error share and average latency, and the unhealthy ones stand out.
What you get
Find the broken connection
Each arrow shows how many calls it carried and how many failed. Open a connection to jump to its most recent failed trace.
Add what traces cannot see
Admins can declare a relationship: calls, depends on, reads from or writes to. Databases and queues appear on the map once you declare them.
How it works
Traces link services
When a span in one service has a parent in another, that is a call between them.
The map reads recent calls
Topology looks back at most six hours, and the page says so. Declared relationships ignore the time range.
Narrow the map
Search, filter by tag, health or resource type, group by an attribute, or show only the services that need attention.
Limits worth knowing
- No zoom. The map scrolls sideways instead.
- Declared relationships need services. Both services must have sent traces in the last six hours or be stored in the workspace.
- Also called NexusMap. The same map is listed as NexusMap under Solution Areas in the app.
See how your services connect.
Send traces from two services and the connection appears.