| name | nemo-relay-export-openinference |
| description | Configure and use NeMo Relay OpenInference export for OTLP backends that understand OpenInference semantics |
| author | NVIDIA Corporation and Affiliates |
| license | Apache-2.0 |
Export OpenInference Traces
Use this skill when the destination expects OpenInference semantic conventions,
for example Arize Phoenix or another OpenInference-aware OTLP backend.
Default Path
- Build the binding-specific
OpenInferenceConfig
- Set endpoint, transport, service metadata, and headers
- Construct and register the subscriber
- Run instrumented scoped work
- Deregister, flush, and shut down when done
Embedded OpenInference Semantics
- OpenInference export is for OTLP backends that understand model-centric
OpenInference semantic conventions.
- Configure
transport, endpoint, service_name, optional namespace and
version, instrumentation scope, headers, resource attributes, and timeout.
- Start with
http_binary transport and an OTLP/HTTP traces endpoint. Use
grpc only when a Tokio runtime is active.
- Scope, tool, and LLM start inputs become
input.value.
- Scope, tool, and LLM end outputs become
output.value.
- LLM usage metadata maps token counters when provider responses include usage.
- Use explicit config fields for endpoint, headers, resource attributes, and
service identity in application code.
- Validate export by checking construction logs, collector traffic, and spans
from the same
root_uuid in the tracing backend.
Important Semantics
- Spans include OpenInference semantic attributes
- LLM spans derive
input.value from request content, not request headers
- Scope types map to OpenInference span kinds
- Orphan mark events still export as zero-duration spans
Troubleshooting Focus
- No spans in the OpenInference-aware backend
- Expected semantic attributes missing
- Wrong scope types or no active scope
- Wrong OTLP transport for the chosen binding or target
Related Skills
nemo-relay-setup-observabilitynemo-relay-export-otelnemo-relay-typed-wrappers-codecs
