I see db-connection details but I don't see any parameter for schema . how will prisma get to know the schema ?

Hi guys not sure if its the same problem i am having now with prisma not recognising schema on my fargate as i see this error on the playground (FYI i am using rds, fargate and ecs with the template from the original doc)

@faithful were you able to create the stack in cloud formation ?

yeah i was

Is your db MySQL or Postgres ?

Aurora mysql on rds. Connected using the instance endpoint

I am using postgres and I gave all details and I get an error with "AmazonECS" role does not exist and it roll backs .

Did you face similar issue ?

no didnt face the role issue.

Since it's not common to join the endpoints of the Prisma and application-layer APIs, is it more common to either a) communicate with either API independently on the front end, based on the needs of the request or b) replicate all the needed CRUD operations from Prisma in the Application?

@shashank you might need to create the service linked role for your account manually

@faithful Still getting same error . That didn't help

I am trying to create 1 more role for ec2ecs . https://aws.amazon.com/premiumsupport/knowledge-center/assume-role-validate-listeners/

Please use threads 🙂 If you want to discuss a bigger topic, I recommend using the Forum instead.

Is there any documentation on how to use subscriptions with prisma-binding?

@nilan started thread for the same . I am kind of stuck in deploying in AWS . I would be glad if you could help.

can i have a link to the thread?

@faithful https://prisma.slack.com/archives/CA491RJH0/p1539192677000100 . @nilan hasn't replied yet

Is Metrics not live yet, or did I set it up wrong?

Any thoughts on using Now for Prisma? I’m seeing fairly high request times, around 1-1.5 seconds on average according to Apollo Engine, which is why I was looking for performance metrics for Prisma. Are Zeit Now’s instances fast/large enough to be a suitable Prisma server host, or should I be looking for alternatives, i.e. Heroku or AWS where I can choose the size of the host instances?

My setup is: App server on Now (bruxelles), Prisma server on Now (bruxelles), and database on AWS RDS (Paris). Even resolvers of pretty simples queries run around 800ms

I’m just looking in making a prisma service that we are trying to make available across multiple regions. We have an Android app backed by AWS Lambda (running graphql yoga), which is talking to prisma, backed by an aws RDS database. Does prisma have any resources for helping us scale to different regions?

Hi @patrickdevivo, I was wondering if you have experienced autocompletion issues like me as describe there https://github.com/prisma/prisma/issues/3294 ?

Hi, why i can't deploy this schema? (only when add roles field as array of string)

deploy hangs on this

Why was resolver-codegen removed from the typescript-graphql-schema-delegation repository? (https://github.com/prisma/prisma-examples/commit/773c80079ab4475f5935eec4d003354149a0786b) Is this module no longer used when creating Prisma-typescript applications?

So I have in my application a mutation that is meant to be able to create an Account with the option to connect or create Assets that are bound to that account. My mutation looks like:

createAccount(email: String!, company: String!, accountType: AccountType, assets: AssetCreateManyInput): Account

. This would be implemented in the same way as the

newUserWithPosts

mutation in the Prisma-Client blog post. But for some reason, the generated interface for AssetCreateManyInput looks like this:

export interface AssetCreateManyInput {
    create: string;
    connect: string;
  }

Am I implementing this mutation wrong in my schema.graphql? Should I have a different type besides AssetCreateManyInput? I'm assuming that it would infer that type from the imported generated prisma.graphql