I don’t understant why I would need to define every crud operation on the server when prisma is supposed to expose all those operations through their api

so

prisma

creates a graphql api for your database(s) with all the crud operations, which you can then call in your own graphql server, which you will have to set-up and define

the point is that you probably dont want to expose every crud operation to your frontends

but that might be your use case, if so then you wont need an intermediate graphql server

some operations you might want to jsut “pass through” which you can do easily

if you look on the homepage, there’s some good examples in the with/without prisma: https://www.prisma.io/

Ok gotcha, thanks. So is it basically an ORM for graphql?

The reason I ask is I really like everything graphcool offers but the shared hosting for it is performing terribly. My requests take multiple seconds. So I thought I’d check out prisma… but maybe I should just look into self hosting for graphcool.

What I personally find super convenient with Prisma is that you can use it to prop up a new database locally or on any hosting provider based on your schema. As a service they also provide Heroku integration which is super nice in theory, but still has a few quirks still

Thanks for the help guys, makes more sense now

I noticed an issue when using the “request” method with javascript prisma-binding library. Depending on the server’s location I get different results. In some cases the response object has the property “data”, and others it excludes it. For example: My server in NYC

const result = await client.request(`
  query {
    users {
      id
    }
  }
`);

console.log(result)
>> {"data":{"users":[{"id":"SOMEUSERID"},]}}

My server in DC area:

console.log(result)
>> {"users":[{"id":"SOMEUSERID"},]}

Has anyone else experienced this issue?

Hi, what If I wanted to expose an existing graphql-api trough a gateway, but I would want to override a type field, i.e. replace

author: String!

by a locally defined type like

author: User!

on a

Post

type. Now, when I query the existing api through its binding passing the

info

object, it will complain about the

info

object, since the

author

field specified there is no longer of the scalar type

String

but of an object type

User

, which is unexpected for the exposed api. Is there a more elegant way to solve this issue then to modify the

info

object? Thanks for your input!

I commented on this issue already but hoping someone here might have some insight on the magical back relations. (https://github.com/prisma/prisma/issues/2693) Are we no longer able to use these back relations in our resolvers? We were using them and now they are no longer generated when doing a codegen to our server. They are still generated when deploying locally though. Is there a recommended way to reverse lookup through types? We have the following resolver that no longer seems to work remotely:

async getDebitTransactionsForWriteOffReport(_, args, ctx: Context, info) { return ctx.db.query.debitTransactions({ where: { _MagicalBackRelation_DebitTransactionsInClaim_every: { _MagicalBackRelation_ClaimsForPatient_every: { _MagicalBackRelation_PatientsInWriteOffReport_every: { id: args.id } } } } }, info)

Hi someone know how can I save the state of redux in the localstorage on nextjs?

Hello everyone! I had a DB generated with prisma datamodel, when I make the introspection of this database the datamodel generated looks like. And when I try to run

prisma deploy

the following error occurs:

Could not find the model table _UserHasProfile in the database

Does anyone know how to solve this?

anyone else having widespread timeouts on the simple API? (back up after about 5 minutes down, this seems to happen a lot)

I noticed when prisma deploy locally, the magical back relation functions exist but when I prisma deploy to my service, the magical back relations get removed. Is this expected behavior and if so, is there a way to use magical back relation?

can anyone point me to a working prisma-client-typescript repo/example?

How do I get the client IP address ?

Hi everyone. prisma green I am currently looking into the security of my app with Prisma. I am wondering: how do I protect the Prisma Server (not my backend) against unwanted requests ? I found documentation on how to protect the management api (with management_api_secret). But I can't find informations about how to secure the rest (services/endpoints). Do you guys have information on this ?

Is forwardTo removed in Prisma client?

Second, inputs and outputs are hashes

What about SQLite adapter?

Hey everyone need some help after not working with prisma few months and i try to make it again here my issue when

prisma deploy

User
    ✖ You are creating a required relation, but there are already nodes that would violate that constraint.
  Department
    ✖ You are creating a required relation, but there are already nodes that would violate that constraint.

here my datamodel

type Department {
  createBy: User! @relation(name: "DepartmentsByUser")
  company: Company! @relation(name: "CompanyDepartments")
}

type Company {
  id: ID! @unique
  createBy: User! @relation(name: "CompanyByUser")
  department: [Department!]! @relation(name: "CompanyDepartments", onDelete: CASCADE)
}

type User {
  id: ID! @unique
  department: [Department!]! @relation(name: "DepartmentsByUser", onDelete: CASCADE)
  company: Company! @relation(name: "CompanyByUser", onDelete: CASCADE)
}

Hey guys where can I find some good tuts on prisma version 1.17?

With the new prisma-client, how can we conveniently import the generated type to our app's

schema.graphql

like we previously did from

src/generated/prisma.graphql

? Do we need to copy and paste the type to

schema.graphql

now? Previously I think we can do

# import User from './generated/prisma.graphql'

if we want to use the

User

type in

schema.graphql

, but now in prisma-client, the schema is defined in a

.js

file and I am not sure how to import that.

Anyone have idea how i can implement mutation with nested / relation i’m still figure out how this you see all my question at hereh https://www.prisma.io/forum/t/mutations-nested/4430

hello. I would like to migrate a database and I think I'll manage to generate the normalized data format. I have one single problem: What do I have to take into account when generating the 15 bytes IDs? They aren't completely random as far as I can see.

Hi, how does

graphql-import

work? This is my situation:

type Query {
  object(where: ObjectWhereUniqueInput!): Object
}

FAILS, unknown type

ObjectWhereUniqueInput

. Adding

# import ObjectWhereUniqueInput from "./schemas/object-service.graphql"

solves the issue. However, adding

# import huhu from "./schemas/object-service.graphql"

Also solves the issue. Is this a bug or am I overlooking something? Should I open an issue?

Has anyone installed Prisma on DigitalOcean? I’m following this guide but stuck at step 10: https://www.prisma.io/docs/1.14/tutorials/deploy-prisma-servers/digital-ocean-(docker-machine)-texoo9aemu