Setting resource attributes in the service telemetry pipeline breaks the pods

[ipversiongent]

SUMMARY:

We're trying to set resource attributes directly in the internal telemetry (service.telemetry), so that we can catch these signals and route (we route based on a custom product.line resource attribute).

This is the configuration we tried (version 0.117.0)

telemetry: resource: attributes: # - name: "product.line" #tried two different options # value: "corporate" - product.line: corporate attributes_list: "product.line=corporate"

However, the collector pods crash with any approach when using the resource attributes, with the error below

'resource[attributes]' expected type 'string', got unconvertible type '[]interface {}', value: '[map[name:product.line type:string_array value:corporate]]' 2025/03/27 15:46:35 collector server run finished with error: failed to get config: cannot unmarshal the configuration: decoding failed due to the following error(s): error decoding 'service.telemetry': decoding failed due to the following error(s): 'resource[attributes]' expected type 'string', got unconvertible type '[]interface {}', value: '[map[name:product.line type:string_array value:corporate]]

The helm chart suggests this is supported:

opentelemetry-configuration/examples/kitchen-sink.yaml

Line 783 in a94e70f

resource:

[jack-berg]

Can you share the formatted version of your config? Its all been condensed on onto one line and hard to interpret.

With that said, I imagine that your encountering a difference in the version of schema used by the collector in 0.117.0 and the version you're referencing in kitchen-sink.yaml. The way that resource attributes are represented in the schema changed in 0.3 in #115:

# before
resource:
  attributes:
    foo: bar
---
# after
resource:
  attributes:
    - name: foo
       value: bar