any

graphql-yoga

examples that show how to add routes and controllers?

Should the generated/prisma.graphql be generated after you run

prisma deploy

?

is there an example of password recovery by mail?

Prisma doesn't have

ON DELETE RESTRICT

?? Why? There is no action, set null and cascade, but no restrict?

Not sure if this question has been asked before. I see there is a

prismagraphlq/prisma-prod

docker image , is this what I am supposed to use for production deployment?

I setup a Fargate instace of prisma with an Aurora database per this tutorial:https://www.prisma.io/docs/tutorials/deploy-prisma-servers/aws-fargate-joofei3ahd/. I then went into Prisma Cloud to add the server per this video: . When I add the endpoint and the managementApiSecret, I get the following error:

Variable '$_input' cannot be non input type 'SetCloudSecretInput!'. (line 1, column 20): mutation ($_input: SetCloudSecretInput!) { ^ Unknown type 'SetCloudSecretInput'. (line 1, column 20): mutation ($_input: SetCloudSecretInput!) { ^ Cannot query field 'setCloudSecret' on type 'Mutation'. (line 2, column 3): setCloudSecret(input: $_input) { ^

I can access the Prisma instance and create a service from

prisma init

without a hitch but I can't add the server to Prisma Cloud. Anyone have some insight on this?

I have such mutation - https://gist.github.com/SilencerWeb/9384ffc2ef12cd3fdf718f1834808493 You can see the mutation and call example on this gist. I cannot understand why I get

[ { side: 'left' }, { side: 'left' } ]

here - https://gist.github.com/SilencerWeb/9384ffc2ef12cd3fdf718f1834808493#file-index-js-L46 instead of something like

[ { person: 'cjiwj0g5kztyu0a277xe3tt2j', side: 'left' }, { person: 'cjiwj0g5kztyu0a277xe3tt2j', side: 'left' } ]

. Strange part is that these created `ActionMember`s are saved in the database correctly -

[ { person: 'cjiwj0g5kztyu0a277xe3tt2j', side: 'left' }, { person: 'cjiwj0g5kztyu0a277xe3tt2j', side: 'left' } ]

, but I get another result in my

console.log

. What can be the reason of such behavior?

Hello there! What’s the proper way of renaming an enum value in Prisma when we have existing data?

Hey @picosam have you tried this https://www.prisma.io/docs/reference/service-configuration/data-model/data-modelling-(sdl)-eiroozae8u#temporary-directives ?

Hey @matic! So that would work for an enum value too?

@picosam: How do you save your data? If your data is made up of events and projections, it would be fairly easy to accomplish this.

@picosam I am not 100% percent sure. It might as well throw an error during Service Deployment. Maybe you could try creating a artificial model and test it out.

I’ll try on my staging server then first

Thanks guys!

@Arnab I’m not sure I understand to be honest lol

Please use threads 🙂

I want to duplicate a complex node, is there any way to do that easily without a long query to catch all fields I need to copy? something like

select *

?

Im able to have a resolver that returns a user:

type Mutation {
	loginOld: User!
}

async loginOld(parent, args, ctx, info) {
	const authId = await getUserAuthId(ctx);
	const existingUser = await ctx.db.query.user({ where: { authId } }, info);
	return existingUser;
},

However what I actually want to do is look for a user in the database and return a string:

type Mutation {
	login: LoginFeedback
}

type LoginFeedback {
	value: String!
}

async login(parent, args, ctx, info) {
	const authId = await getUserAuthId(ctx);
	const userExists = await ctx.db.query.user({ where: { authId } }, info);
	if (userExists) return {value: "A user exists"}
	return {value: "No user exists"}
}

Even If I don’t do anything with userExists, as soon as I query the database I get an error message:

Field 'user' of type 'User' must have a sub selection. (line 2, column 3):\n  user(where: $_v0_where)\n  ^

I think this is the schema getting upset as Ive said the type is LoginFeedback not Login, but is there not a way to query the Prisma database in mutation?

How can I use endpoint from .env in prisma.yml?

Hey guys, I have a question about prisma with multi tenancy. I want to create a new web view for some of our data. We have one main database and several account specific databases. How would i connect or setup prisma?

Hi there, there appears to be an issue with

prisma reset

at the moment. I am resetting one of my endpoints, but the data presist. I double checked that I am resetting the right endpoint. Has anyone experienced similar behavior?

I'm trying to run the prisma for first time by following this tutorial: https://www.prisma.io/docs/tutorials/setup-prisma/connect-empty-db/mysql-dusee0nore Executed this command: docker-compose up -d Starting my_project_prisma_1 ... done Finally when executed the 'prisma deploy' command, getting following error: Could not connect to server at http://localhost:4466. Please check if your server is running. Any help?

Is there any way to specify shorter data types for specific string data points? We have noticed that Prisma creates mediumtext types for all of the strong fields and in many cases this is way larger than what we need.

Hi, after a

prisma deploy

and a post deploy hook to

graphql get-schema

how can i import the whole generated schema into my own (as i have some rest API stuff as well), rather than

# import OneThing from "./generated/schema.graphql"

Let's say i have a file

datamodel.graphql

that I deployed on Prisma, is there a way to regenerate that file by querying Prisma ? I know that with GraphQL we can introspect the API, but this will give me the schema of the generated API, not the source. It may be possible to infer the source from the generated API but it seems complex. I need this to be able to generate a diff between a

datamodel.graphql

I have locally and the actual deployed schema.

Is it possible to do something like live queries currently in Prisma/Apollo? I have a third party REST endpoint that I need to poll for pretty much the entire usage cycle of my application. And I've got it wrapped in GraphQL on the server side, as our API is built with Prisma/Apollo... Right now, this is just a simple query that passes the request along to the REST endpoint, then cherry picks some things out of the response, along with adding in some extra data...and pushes a response back to the client application. It'd be hell of slick if our client application didn't have to repeatedly hit our endpoint... My gut says that is not really possible, though.

can anyone shed light on the proper ordering when adding express routes to a

graphql-yoga

server?

My relationships don't seem to work. I've setup

type User{
  id: ID! @unique
  name: String!
  rating: Int! @default(value: "0")
  createdAt: DateTime!
  updatedAt: DateTime!
  listings: [Listing!]! @relation(name: "ListingsByUser" onDelete: CASCADE)
}

type Listing {
  id: ID! @unique
  title: String!
  content: String!
  published: Boolean! @default(value: "true")
  createdAt: DateTime!
  updatedAt: DateTime!
  author: User! @relation(name: "ListingsByUser")
}

My seed:

mutation {

    user: createUser(
        data: {
            name: "Ian Jamieson"
            rating: 0
            listings: {
                create: {
                    title: "Test title"
                    content: "lorem ipsum"
                }
            }
        }
    ) {
        id
    }

}

Listing returns the user ok, but my user listings returns null

{
  "data": {
    "users": [
      {
        "id": "cjiw9mopw000407197os3fu9d",
        "name": "Ian Jamieson",
        "rating": 0,
        "listings": null
      }
    ]
  }
}

Is there something I'm missing? Found out it's my resolvers index.js

import * as path from "path";
import { mergeResolvers, fileLoader } from "merge-graphql-schemas";

const resolversArray = fileLoader(path.join(__dirname, "./**/*.resolvers.*"));

const resolvers = mergeResolvers(resolversArray);

export default resolvers;

So how do i merge them properly?

Evening .