Feature request: Tracer disable response capturing on a per-method basis

[misterjoshua]

Description of the feature request

I'd like tracer.captureMethod() to allow customers to disable capturing the response as metadata on a per-method basis.

Problem statement

I find when capturing a method that returns a boatload of data, the subsegment may fail to send because the subsegment exceeds X-Ray's 64KB segment document size limit.

In my project, I'm finding it useful to capture the response by default. But, in this particular case, there's just this one method with an oversized method response. I'd want to disable response capture on a per-method basis rather than globally, so the POWERTOOLS_TRACER_CAPTURE_RESPONSE env doesn't fit my need.

Summary of the feature

I think that AWS Lambda Powertools for TypeScript should implement the Python version's decorator option.

Code examples

const tracer = new Tracer();

class MyThing {
  @tracer.captureMethod({ captureResponse: false })
  getBigResponse() { ... }
}

Benefits for you and the wider AWS community

Allowing customers to disable response capturing on a per-method basis would allow them to observe these subsegments in their trace timelines where they would otherwise fail to send.

Describe alternatives you've considered

I considered POWERTOOLS_TRACER_CAPTURE_RESPONSE, but I don't want to disable response capturing across my entire app.

Additional context

None

Related issues, RFCs

None

[dreamorosi]

Thank you @misterjoshua for opening this feature request.

I can see how this behavior is desirable, especially in light of the X-Ray payload limit that you have mentioned. At the moment the behavior is all-or-nothing and you can't really opt out on a per-method basis.

Please give us some time to prioritise the request and fit it in our roadmap. As soon as we have an implementer either from the community or from the maintainers team we'll assign the issue and change the status to "In Progress".

[github-actions]

⚠️ COMMENT VISIBILITY WARNING ⚠️

Comments on closed issues are hard for our team to see.
If you need more assistance, please either tag a team member or open a new issue that references this one.
If you wish to keep having a conversation with other community members under this issue feel free to do so.