每次我需要在 Lambda 中使用它时如何避免打包请求？

Question

This is in reference to the change to removal of the vendored version of requests from botocore as seen here: https://aws.amazon.com/blogs/developer/removing-the-vendored-version-of-requests-from-botocore/

I've been changing existing lambdas for this change. For simple get requests I've just been using urllib3 since it is part of std lib but have ran into issues attempting to use it for more complex calls such as setting many headers, authing and posting content. I've found various posts, including on SO, that recommend just using requests, which I must say I do prefer.

That said, the only way I know of to do that successfully in lambda going forward is by performing "pip install requests -t ./" and then zipping up a deployment package(documentation here: https://aws.amazon.com/premiumsupport/knowledge-center/build-python-lambda-deployment-package/ ) for each and every one of my lambdas that requires the use of requests.

Am I being daft here? Is there an easier or better way to do this or is it my only option? How wasteful would it be to have a lambda function behind an api gateway that imported requests and all of my other lambdas called(using urllib3) and passed parameters that made the call and replied back?

I feel like I must be overlooking a better way of doing this, please enlighten me!

Answer 1

将requests放入 Lambda 层并在您的各种 Lamdba 函数中使用该层