ability to configure the tracer http timeout or at least add documentation on how to do it if it already exists

[acsecureworks]

Expected Behavior

there should be an ENV variable to configure the http timeout.

DD_TRACE_HTTP_TIMEOUT

Actual Behavior

I don't see a clear/obvious way to the set the http timeout. I might very well have not looked deep enough and it already exists so if it does, please provide documentation and an example would be great

[purple4reina]

Hi @acsecureworks thanks for reaching out! Help me understand, are you looking to add the WithAgentTimeout option to your tracer?

I'm also curious to learn more about what problem you are hoping to solve.

[acsecureworks]

yes,

I am seeing traces for my app but I am also seeing errors in my datadog logs like

2025/02/14 19:18:44 Datadog Tracer v1.65.1 ERROR: lost 2 traces: Post "http://foo.datadoghq.com:8126/v0.4/traces": context deadline exceeded (Client.Timeout exceeded while awaiting headers), 11 additional messages skipped (first occurrence: 14 Feb 25 19:17 UTC)

I'm thinking if I increase that timeout, I can reduce these errors

[purple4reina]

Hmm, that's really curious. The tracer should not be attempting to post traces directly to intake, they should always be going through the Datadog Lambda Extension.

Can you share a bit more about your setup and configuration? Which layers do you have installed? What environment variables do you have set?

[acsecureworks]

I believe a figure this out

lambda.Start(ddlambda.WrapFunction(HandleRequest, &ddlambda.Config{
	TracerOptions: []tracer.StartOption{
		tracer.WithAgentTimeout(30),
	},
}))