[beta] Python Profiling support

[astuyve]

What does this PR do?

This PR enables the (beta) use of the new serverless profiling capabilities of ddtrace-py in Lambda.

Motivation

Testing Guidelines

Additional Notes

Types of Changes

• Bug fix
• New feature
• Breaking change
• Misc (docs, refactoring, dependency upgrade, etc.)

Check all that apply

• This PR's description is comprehensive
• This PR contains breaking changes that are documented in the description
• This PR introduces new APIs or parameters that are documented and unlikely to change in the foreseeable future
• This PR impacts documentation, and it has been updated (or a ticket has been logged)
• This PR's changes are covered by the automated tests
• This PR collects user input/sensitive content into Datadog
• This PR passes the integration tests (ask a Datadog member to run the tests)

[maxday]

LGTM! do we want to build and release this layer in sandbox to see the impact on cold start (when profiling is disabled) since the package size has grown a bit?

[astuyve]

LGTM! do we want to build and release this layer in sandbox to see the impact on cold start (when profiling is disabled) since the package size has grown a bit?

We've got a release on sandbox already, will test a cold start version shortly

[astuyve]

We ran three versions of the same handler code (simple hello world).

• without-profiling (version 60 of this library, no profiler code at all)
• with profiling (pre-release of 61 of this library, which imports the profiler and turns it on)
• with profiling but disabled (pre-release of 61 of this library, which has the larger filesize compared to without-profiling)

With profiling packaged and enabled, we see ~200ish ms cold start increase.
We see almost no cold start increase for including profiling, but not loading that code.