Prisma Cluster with AWS AppSync


Has anyone explored this using a private prisma cluster with AWS Appsync?

I came across this article:

The problem is that it integrates with dynamobd using VTL but it does seem like you could call out to other data sources via lambda:

Would this be where you would use prisma bindings to call the prisma cloud endpoint? If so how would this look/work?

It’s my understanding that most people are using graphql yoga which I guess I could just continue down that path. Unfortunately, there are lots of pieces to get working together i.e. gateway, s3, lambda etc… and I have struggled to get everything to work correctly (especially in a locked down environment) even with up and serverless.

Using appsync seems to alleviate some of this (can simply use the console). It also has the added benefit of supporting subscriptions out of the box which is nice. It would be beneficial if we had some more options as well as some boiler plates to get everything going in AWS (which is a complex beast in itself). I still find myself spending way too much time doing dev ops even with prisma cloud :frowning:


They are not restricting you to any specific DB backend, however the integration happens via AWS lamda which do add another hop to your setup. There are plenty of ways to deploy a graphql server on lamda, but there is no way around that additional hop.

But since graphql takes the pain out of querying relational data out of dynamo and other no relational stores and knowing that via graphql you will be able to get the data out in a standardized format. There is very little reason (besides legal) left to not just use Dynamo while you are running on AWS anyway and use graphql based exporting as a way to get data out if you have to. Stating this even though I have strong personal bias for prisma.

That said the offline capabilities + integration are looking very promising. Its the one piece that hasnt been solved in a standardized and open source manner. I think this is where the graphql community should focus on now after great improvements for subscriptions and tooling , now is the time to turn our focus to generalizing data sync for offline use cases