Using Pulumi & AWS CDK together If somebody wanted to, could they use AWS CDK & Pulumi in a project? I know it means two state locations but how viable would this be?

yes, https://www.pulumi.com/blog/aws-cdk-on-pulumi/

Ah ha - NICE, thank you so much @millions-furniture-75402

Wow this is awesome while at the same time sounding like a good foot gun.

The use case I see here, is that more vendors are providing examples/solutions using CDK. That requires either your organization also supports deployment via CDK, you convert to Pulumi, or encapsulate their IaC some other way. Hopefully it's not "just because" I want to mix the two. Similarly, I've deployed Tableau using more-or-less off-the shelf CloudFormation YAML that I deployed using Pulumi. https://www.pulumi.com/registry/packages/aws/api-docs/cloudformation/stack/ While not quite the same, this example also serves to demonstrate the opposite side of the problem of encapsulating a vendor's IaC. I don't see them as Pulumi managed resources.

OMG Tableau ... many years ago I wrote Cloudformation deploying Tableau server and then designed the first ever Tableau CI pipeline that allowed my client to transfer projects between disconnected servers. I mentioned that tableau is insanely slow for their usecase (IoT data streams) but it wasn't my decision. 2 years later I was replacing it with Grafana 😅 and wrote a bit of CDK.

Why CDK over Pulumi in that case?

Pulumi was too young and the client was heavily invested in CFN on the corporate level, adopting a new tool would cause a lot of friction. Even for CDK, I had to invent stuff and design processes they weren't ready for. ATM I maintain deployment of several CFN stacks via Pulumi because they came as ready-made solutions. But it is good to know I can mix and match with CDK as well.

Haha, yeah, that sounds about right. So you understand this use case well!

I also talked to Tableau developers and even they did not know how could I achieve the pipeline and yet I did it!