Hi,
for serverless the differentiation on collected traces is indeed not that big. Still the Dynatrace AWS Lambda Layers have some big advantages over using OpenTelemetry only. For example is that we provide automatic real-user monitoring for webapps served from AWS Lambda. Due to the nature how traces/spans are processed (e.g. OpenTelemetry only sends finished spans) we can provide much better realtime service analysis for spans coming from AWS Lambda Layer. Our approach is to not compete with OpenTelemetry and use it complementary to our OneAgent based instrumentation. E.g. if you need additional tracing into e.g. AWS DynamoDB (which is currently not provided within our AWS Lambda Layer), then use additionally the client SDK instrumentation within AWS Distro for OpenTelemetry to get traceability also into this service.
This allows you to mix&match traces from OpenTelemetry with PurePath however the spans are consumed, either with OneAgent or via API. This is important as majority of environments are still very heterogenic, using all different cloud services from IaaS/PaaS/SaaS/.. . So this comes along with all our other efforts to consume telemetry from any 3rd party source, such as metrics & logs from e.g. cloudwatc which Dynatrace then puts into context to maximize the insights & automatic analysis within such an environment.
