Chart doesn't configure frontend collectorUrl properly

[srcreigh]

Hi! Thanks for the great tool.

I think the helm chart uses the wrong env var to pass otlpExporterEndpoint to the frontend.

I deployed HyperDX with kubernetes, and the app seemed to endlessly send data to localhost:4318, no matter what helm chart values I set.

So I dug around, found that the HyperDX app/frontend uses an env var NEXT_PUBLIC_OTEL_EXPORTER_OTLP_ENDPOINT to initialize HyperDX browser trace/log sender.

• https://github.com/hyperdxio/hyperdx/blob/56fd856d7a839a28560e1a59d1d74cc52499f2f6/packages/app/src/config.ts#L19-L21
• https://github.com/hyperdxio/hyperdx/blob/56fd856d7a839a28560e1a59d1d74cc52499f2f6/packages/app/pages/api/config.ts#L12
• https://github.com/hyperdxio/hyperdx/blob/56fd856d7a839a28560e1a59d1d74cc52499f2f6/packages/app/pages/_app.tsx#L85

The problem is that the helm chart sets OTEL_EXPORTER_OTLP_ENDPOINT, not NEXT_PUBLIC_OTEL_EXPORTER_OTLP_ENDPOINT.

https://github.com/hyperdxio/helm-charts/blob/2161d8151b434c81ab6aab6c33b32f2827031099/charts/hdx-oss-v2/templates/configmaps/app-configmap.yaml

I think it's a mistake and the chart should just get updated to set NEXT_PUBLIC_OTEL_EXPORTER_OTLP_ENDPOINT. The OTEL_EXPORTER_OTLP_ENDPOINT value does not seem to be used for anything except mapping to NEXT_PUBLIC_* for docker deployments.

There is a workaround to patch the config map to set NEXT_PUBLIC_*, but it would be better to fix the chart imo.