Resources

Three alarms fire at once.
Three identical errors.
But where did they come from?

error metric

request_duration: 920ms

http.route = "/checkout"

http.response.status_code = 500

resource

service.name = "orders"

deployment.environment = "production"

host.name = "pod-7f8b9"

error metric

request_duration: 920ms

http.route = "/checkout"

http.response.status_code = 500

resource

service.name = "orders"

deployment.environment = "staging"

host.name = "pod-2a3b4"

error metric

request_duration: 920ms

http.route = "/checkout"

http.response.status_code = 500

resource

service.name = "payment"

deployment.environment = "production"

host.name = "pod-9c8d7"

Three identical errors. But are they from the same service? Different services?
Different environments? You can't tell.

What information is missing?

Now you can see the resources. How many services are affected?

Is this happening in production only?

Resources attach to all signals — not just metrics

span

checkout

duration = 142ms

resource

service.name = "orders"

deployment.environment = "production"

host.name = "pod-7f8b9"

log

Payment timeout after 5000ms

severity = ERROR

resource

service.name = "orders"

deployment.environment = "production"

host.name = "pod-7f8b9"

A resource is the identity card of a telemetry source.

Attributes describe what happened. Resources tell you who reported it.