Description

Siglens is not able to convert OTLP format back to prometheus format using /otlp endpoint.
Using otlphttp exporter, we are exporting metrics to siglens in /otlp endpoint.
We see some changes there in the metric name and metric labels.
Some prometheus specific labels like job/instance are not present.
https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/compatibility/prometheus_and_openmetrics.md#resource-attributes

Also "_count" is not getting added to the counter metrics ingested.

Generally the metrics are staying as it is in OTLP format and being displayed in siglens.
But if the same data is sent to prometheus using otlphttp exporter , we see regular behavior.
And if we use prometheusremotewrite exporter and send it to siglens on /promql endpoint, it is
able to retain the labels or the naming conventions.

Steps to Reproduce

Use otlphttp exporter and send metrics to /otlp endpoint of siglens

Expected Behavior

Conversion of otlp to prometheus should happen at siglens side to retain prometheus standards
just as /promql endpoint.

Environment

• Commit Hash: [e.g. 9ed5f0a]
• OS: [e.g. MacOS 14.1.1 on apple silicon]
• Browser and Version (if applicable): [e.g. chrome 119.0.6045.159 arm64, safari 19616.2.9.11.7]

(Optional) Additional Context

Add anything else we should know about this problem.

Slack Channel Query:
https://siglens-community.slack.com/archives/C06412TCVT7/p1750700685404029