0

When I use serverless, I don’t want to deeply integrate the SDK of cloud vendors, which will cause cloud vendors to lock in.

So on Java I can use abstraction layer like Spring Cloud Function to write code, adapt to different cloud vendor SDKs at the final deployment.

When using Python to write serverless code, is there a library like Spring Cloud Function?

jonrsharpe
  • 115,751
  • 26
  • 228
  • 437
Ten Wang
  • 51
  • 1
  • 1
  • 6
  • You can try https://www.jython.org/, this way you still use s-c-function with the `python`. Also, I am not really following the _lock in_ comment as you are simply relying on language, so in this case java or python is kind of the same thing – Oleg Zhurakousky Jun 02 '21 at 07:43
  • You always will have lock in. The Lambda that you write will most likely use a service of your cloud provider. For example S3. So your code will have S3 code in it. You won't be able to lift that code and just magically use the Azure S3 equivalent. And then there is your deployment. You probably will use Infrastructure as Code tools like Terraform, which will also have a "tight coupling" to your public cloud. So in the end there is AFAIK no way around the "lock in". – Jens Jun 02 '21 at 12:51
  • But we can use an abstraction layer like _dapr_ to avoid the integration of S3 SDK into our code, and use _spring cloud function_ to avoid integration of lambda SDK. In the end, we will encapsulate an '-aws' layer on top of the '-core'. This layer only has adapters such as _springcloudfunction-aws_ and _dapr-aws_. – Ten Wang Jun 04 '21 at 09:18

0 Answers0