Hello everybody, I wanted to ask you for your opinion about my current setup. I'm especially interested in your opinion, if I could use this setup for a first production state, before going further. Here is my current setup. DB: AWS RDS -> Connection details like password and username are under the PRISMA_CONFIG in the docker-compose.yml. Prisma server: Runs "locally" (on the server) in a docker container. Has a secret defined in the prisma.yml file (which is also used in the

new Prisma

statement in the index.js of the graphql server). The playground is disabled (it should not be visible to anyone). I have a managementApiSecret in the PRISMA_CONFIG, which is also in the .env to protect the deployment process (which is probably not necessary, as the container is running locally as mentioned earlier). GraphQl server: Standard graphql server using yoga. The only thing of my api, that is actually exposed (accessible from outside), runs on the same machine, as the prisma server and accesses the service via localhost. Currently only accessible via

hostip.port

after being started with

yarn start

, but will be set up as node.js server (w/ daemon) and managed via nginx. I will also setup a ssl cert for this and direct a subdomain (like api.example.com) of my maindomain to this server and route that trough to the port of the running node.js server. What do you think? Did I forget anything? Should I pay attention to something? Whats your opinion? Thanks in advance! 😉

hello, I want to add an array like this [{OneMonth: 0, ThreeMonth: 0, SixMonth: 0, OneYear: 0}] to a field. How can I do that

Hi I'm new to Prisma and GraphQL in general, I am a React developer. I know that 1.7 is out just confused. I run

prisma init

in my folder, then chose the docker option with mysql. Then i run

graphql create server

and it created a boilerplate for me. just confused why are there two config for these?

Hey guys, Nice to meet you here.

A file was commented on

Is there a smart way to control the server console logging? When im writing custom resolvers its handy to log variables to see whats going on. However my server terminal is really busy. Is there a way of temporarily disabling the default logging or easily distinguishing the console.logs that ive added? Thanks

A file was commented on

Ive added a content type to datamodel.graphql:

type Location {
	id: ID! @unique
	name: String!
	machineName: String
}

My query in graphiql works:

query {
  locations(
    where: {
      machineName: "paris"
    }
  ) {
    id
    name
    machineName
  }
}

Ive added a new resolver:

type Query {
  location: Location
  feed: [Post!]!
  drafts: [Post!]!
  post(id: ID!): Post
  locations: [Location!]!
  me: User
}

And created the resolver:

location(parent, { machineName }, ctx, info) {
    return ctx.db.query.location({ where: { machineName } }, info);
  },

But I try to call it from React I get an error:

Network error: Response not successful: Received status code 400

export const LOCATION_QUERY = gql`
  query LocationQuery {
    location(where: { machineName: "paris" }) {
      id
      name
      machineName
    }
  }
`;

I already completed a project with UI and backend [ no DB part ] with AngularJS and Node + Apollo Graphql server. Now I created a prisma DB and got auto generated queries and mutation. So I want to connect both projects [ Apollo server project and Prisma project ]. From your link above i understood that they put actual business schema in schema.graphql and resolvers are created manually by connecting DB. So do i need to include all my resolvers and business parts of my Apollo project to the /SRC folder of Prisma project for integration. Or vice versa.

Hello! Anyone has a functioning example of a Prisma project where two separate GraphQL APIs are created out of the same Prisma Server? I’m toying with this to create one “public” API and another “Admin” one. Thoughts? Thanks!

Hey guys, any better ideas as to how to handle product variants ? I replicated Shopify datamodel, but it’s quite a pain in the ass to generate those variants, and to convert them as option selectors 😐

hey guys, what’s the best way to handle authorisation with graphql directives?

any packages I can use?

Id like to let users sign up with either email, facebook or twitter. Is it advisable to write custom code for this or use a service like Auth0 or Firebase?

Is it possible / advisable to convert an array to a single item in a resolver? My resolver returns an array, but I know that there will always be a single result. Is it possible to modifiy the resolver to reutrn a single result? This is my resolver:

location(parent, args, ctx, info) {
   const { machineName } = args;
   return ctx.db.query.locations({ where: { machineName } }, info);
 }

This query from the front-end:

{
  location(machineName: "london") {
    id
    name
    
  }
}

Will return:

{
  "data": {
    "location": [
      {
        "id": "cjgrqosbt5eum0b51hz4yqrou",
        "name": "London"
      }
    ]
  }
}

But id like something like:

{
  "data": {
    "location" {
      "id": "cjgrqosbt5eum0b51hz4yqrou",
      "name": "London"
    }
  }
}

If I serially make Prisma calls, it takes forever to save the data. Forever as in days. It ran overnight and couldn't save all the data. If I allow parallel calls, I get errors. Anyone else experiencing this?

Id like to try Prisma Cloud. How do I go about choosing between RDS vs AWS Aurora databases?

Does anyone know how to properly set the fetchPolicy on the newer Apollo Query Component? The docs don’t give an example… ie:

<Query query={MY_QUERY} fetchPolicy={'network-only'}>

or

<Query query={MY_QUERY} fetchPolicy="network-only">

Has anyone else been having issues with the

exists

operation after upgrading to Prisma 1.7? Was the syntax somehow changed? The following code, which used to work fine, now throws a TypeError: "cannot read property 'call' of undefined":

const exists = await ctx.db.exists.Media({ localSource })

Anyone that can tell me how to use cors properly? 😕 My windows computer can't send graphql requests to my mac, but I can visit the graphql endpoint and use playground.

So, I just tried to upgrade a legacy project to the new framework and gvot the following message:

The project could not be ejected because it has enabled integrations. Please migrate all integrations to resolvers first.

How do I 'migrate all integrations to resolvers'?

Excited for your Next.js book!

Hi. Are there differences between the instances on Prisma Cloud and the ones deployed locally? I’m experiencing differences between an instance on prisma cloud and one deployed locally (tested with versions 1.7 and 1.8)

I get the error above when trying to create something with connected entities, like this:

mutation {
  createAttachment(data:{
    name: "teste",
    description:"desc",
    type: ARTICLE,
    kind:WEB,
    imageUrl:"<http://www.google.com>",
    url:"<http://www.google.com>"
    organizations:{
      connect: [{
        id: "cjgsibg858lms0b06p4a6lgj4"
      }]
    }
  }) {
    id
    name
  }
}

When testing this same query against a cluster deployed on prisma cloud, it works

Ok, turns out that I was using postgres on the local cluster, so this is a problem with the postgres connector I guess

using the mysql connector on the local cluster it worked

hello, I want to add an array like this [{OneMonth: 0, ThreeMonth: 0, SixMonth: 0, OneYear: 0}] to a field. How can I do that

I need to alter my layout dependin on weather a user is logged in or not. Is this a good pattern?

const ME = gql`
  query me {
    me {
      id
    }
  }
`;

 <Query query={ME}>
          {({ loading, error, data }) => {
            if (loading) return <Loading />;
            const userisLoggedIn = error ? false : true;

            // Then I use userisLoggedIn to conditionally render stuff

It seems to work but Graphcool BaaS retunred null rather than an error when you queried for a logged in user. For some reason that felt more proper to me.