🚨 Have you already secured your free ticker for Next.JS Conf? Prisma is one of the sponsors and our DevRel rockstars Daniel Norman and Mahmoud Abdelwahab will be speaking about Prisma 🤩 👉 Claim your free ticker here: https://nextjs.org/conf 👉 And you can learn more about Prisma + Next.js (and order free ✨ stickers ✨ ) here: https://www.prisma.io/nextjs

Is there any way to update a relation? The following is the explicit many-to-many example from the docs:

model Post {
  id         Int                 @id @default(autoincrement())
  title      String
  categories CategoriesOnPosts[]
}

model Category {
  id    Int                 @id @default(autoincrement())
  name  String
  posts CategoriesOnPosts[]
}

model CategoriesOnPosts {
  post       Post     @relation(fields: [postId], references: [id])
  postId     Int // relation scalar field (used in the `@relation` attribute above)
  category   Category @relation(fields: [categoryId], references: [id])
  categoryId Int // relation scalar field (used in the `@relation` attribute above)
  assignedAt DateTime @default(now())
  assignedBy String

  @@id([postId, categoryId])
}

Suppose I want to do the following:

await prisma.post.update({
  where: {
    title: 'Foo'
  },
  data: {
    categories: {
      category: {
        name: 'Updated category name'
      }
    }
  }
});

All I get is `"error": "Unknown arg

categories

in where.categories for type PostWhereUniqueInput. Did you mean

set

?\n"` Is there any way for this to be accomplished?

how do i update a table and set one column to another columns value in the same row? ie update foo set x = y

In the second instalment of this tutorial (https://www.prisma.io/blog/fullstack-nextjs-graphql-prisma-2-fwpc6ds155#pagination), the final step has an instruction to update the resolver to add a results const. I don’t know where to put that const (it doesn’t fit with the resolver file as it is currently setup) and then the code provided for the index.tsx file does not define links. Has anyone figured out how to complete this step of the tutorial? How do I update the resolver and how do I define links in the index.tsx? Thank you.

Hi, it's known that concurrent interactive transactions on the same rows can cause timeouts – however, I'm seeing timeouts even with only one interactive transaction at a time. Has this been noticed by anybody else?

Hi 🙂

Is it also possible to get help here?

what’s the best way to create a repository pattern that is not dependent on prisma’s data structures? For instance, I want to create a repository with a “create” method:

async create(data: Prisma.AutoCreateConfigurationCreateInput) {
  return await prismaClient.autoCreateConfiguration.create({
    data,
  });
}

The problem with the above code is that the method signature has a direct dependency on “prisma” which seems like an anti-pattern since the whole point of repositories are to abstract away the data access implementation details. Anyone have any good patterns they use here?

Hi guys, I am trying to add follower/following data to the db. What is the best approach to do it? It is throwing errors when I try adding relationship with the User model twice in the table.

Hello, does anyone know what could cause this issue? I have a server side rendered page in Next.js and a standalone Node.js server which only creates an instance of PrismaClient once (I am not using Next.js's server), I get this error right away when I visit the home page and Next.js makes a call to the standalone API

Timed out fetching a new connection from the connection pool. (More info: <http://pris.ly/d/connection-pool>, Current connection limit: 3)

This is a question regarding Graphql + Next.js , in the recent video series why was the graphql API made using Next.js API routes and not a server, what benefits does it have over a server (apart from cost)

Are there any ways to properly use prisma client in multytenancy mode? E.g. overriding database connector in runutime? Currently I generate multiple clients for this and switch between them, but this is obviously memory-hungry way. Inspecting the generated client code but can't see how db parameters are even provided to query engine

Hi there 👋 I have an issue regarding the combination of

distinct

&

cursor-pagination

. TLDR: How can i specify that distinct is applied before the cursor pagination steps (cursorId, skip, take) and not afterwards. Example: Having the following Model:

model Comments {
  id       String  @id @default(uuid())
  author   User?   @relation(fields: [authorId], references: [id])
  authorId String?
  cursorId String  @unique
  text     String
}

and having that model filled with the following data:

{
      authorId: "user1",
      cursorId: "0001",
      text:     "text1",
    },
    {
      authorId: "user2",
      cursorId: "0002",
      text:     "text2",
    },
    {
      authorId: "user3",
      cursorId: "0003",
      text:     "text3",
    },
    {
      authorId: "user4",
      cursorId: "0004",
      text:     "text4",
    },
    {
      authorId: "user1", // sidenote: this is the 2nd comment linked to user1
      cursorId: "0005",
      text:     "text5",
    }

The objective is to fetch the comments in a cursor-paginated way and recieving items that are distinct on the field authorId. Considering 3 items per page, leads to the following query for the 2nd page:

const res = await prismaClient.blogPost.findMany({
      distinct: ["authorId"],
      orderBy: { cursorId: "asc" },
      cursor: { cursorId: "0003" },
      skip: 1,
      take: 3,
    });

This leads to the following res where the 2nd item is fetched although unfortunately in the first cursor-pagination fetch a item with that authorId "user1" has been fetched already before.

{
      authorId: "user4",
      cursorId: "0004",
      text:     "text4",
    }
    {
      authorId: "user1", 
      cursorId: "0005",
      text:     "text5",
    }

What i would like to have on that 2nd fetch is just the object with cursorId "0004". I assume this could be solved by the possibility of distinct being applied before the cursor pagination. Is there the possibility to do so? Thanks in advice.

Has anyone received the following error message?

Query createOneActivity is required to return data, but found no record(s).

I just created a model and trying to create a record and receiving this error.. This is my model

model Activity {
  createdAt      DateTime @default(now()) @map(name: "created_at") @db.Timestamptz(0)
  actionUserId   String   @map(name: "action_user_id") @db.Uuid
  affectedUserId String   @map(name: "affected_user_id") @db.Uuid
  event          String   @db.VarChar(64)
  entity         Json

  @@id([createdAt, actionUserId, affectedUserId, event])
  @@index(fields: [affectedUserId, createdAt], name: "actvt_fltr_idx")
  @@index(fields: [affectedUserId, createdAt, event], name: "actvt_fltr_evt_idx")
  @@map(name: "activity")
}

and this is the code that I’m using

import { PrismaClient } from '@prisma/client';
const prisma = new PrismaClient();

const handler = async () => {
  const data = await prisma.activity.create({
    data: {
      event: 'test',
      actionUserId: '85a42af9-bb28-47b3-9ef2-699858c8bbea',
      affectedUserId: '85a42af9-bb28-47b3-9ef2-699858c8bbea',
      entity: { key: 'value' },
    },
  });

  console.log(data);
};

handler();

Hi – is it possible to use RAISE EXCEPTION in executeRawUnsafe? I am receiving a syntax error when using it, but it's the only way that I can think to make Prisma's transaction functionality work for our use case. We need to look up rows, perform service logic on those rows, and subsequently create or update multiple tables. We need the create/updates to be executed as a transaction, and we also need to use optimistic concurrency control to guard against race conditions between the find+create. The best way I can think to do this is to add raw SQL to the transaction that throws if the expected version has been incremented. Is this possible with Prisma's executeRawUnsafe API?

Hi all - playing around with the golang client. I'm trying to grab a list of users with a few related records. A user has many databases. I've queried them with

db.User.Database.Some()

A user has many alerts. I've queried them with

db.User.Alert.Some()

An alert has one notification. I've tried querying it with

db.Alert.Notification.Some()

, but

Some

isn't available here. If it's a "has one" relationship, should I be using another function instead?

users, err := prisma.User.FindMany(
		db.User.Alert.Some(
			db.Alert.Notification.Some(),
		),
		db.User.Database.Some(),
	).Exec(ctx)

I'm curious when setting

referentialIntegrity = "prisma"

how does Prisma Client actually enforce referential integrity? I'm guessing it must make queries to validate operations before running them?

Hi everyone, I am having some trouble creating game instances with 6 users in. The models are as follow

model User {
  id String @id
  avatar String 
  username String
  discriminator String
  games Game[]
}

model Game {
  id      Int   @id @default(autoincrement())
  players User[] 
  team1   String[]
  team2   String[]
  result  Int?

  createdAt DateTime @default(now()) @db.Timestamp(6)
}

When I try to create the game instance it wont let me create the array of Users... How do I connect multiple players to the game?

const playerQueue = queue.map((user) => ({
			id: user.id,
			username: user.username,
			discriminator: user.discriminator,
			avatar: user.avatarURL()!
		}));

		const test = await this.container.prisma.game.create({
			data: {
				players: {
					create: playerQueue
				},
				team1: playerQueue.splice(0, 2).map((usr) => usr.id),
				team2: playerQueue.splice(2, 5).map((usr) => usr.id)
			}
		});

i have some code that loops through a list of emails and creates users from it if conditions are met. there can be several hundred users. it looks like this:

responses.map(r => {

            if (emailSchema.isValidSync(creatorEmail))
              await db.user.upsert({
                where: {
                  email: r.creatorEmail,
                },
                update: {},
                create: {
                  email: r.creatorEmail,
                },
              })

            if (emailSchema.isValidSync(approverEmail))
              await db.user.upsert({
                where: {
                  email: r.approverEmail,
                },
                update: {},
                create: {
                  email: r.approverEmail,
                },
              })
      })

the user model has a unique constraint on the email column:

model User {
  id            String    @id @default(cuid())
  email         String?   @unique
  // etc...
}

i intermittently get `Unique constraint failed on the fields: (

email

)` errors 😞 is this a race condition? if so, how can i fix it?

running into issues with: `PrismaClientRustPanicError: PANIC: called

Option::unwrap()

on a

None

value in query-engine/core/src/query_document/parser.rs25087` I see there’s one mention of it on GH issues and nothing beyond that. hard to reproduce

anyway seen these strange PrismaClientRustPanicError errors before?

Hey, I know it's been asked before but I couldn't understand from previous responses how to solve this thing: I've a monorepo, I've two packages that both should represent Prisma client, each one with a different database (different RDS instance for that matter) and different schema. I need to use these two packages, call them

@me/db1

and

@me/db2

from every other package (

@me/*

). This is Typescript so generating to

src

or the root of the package is not ideal, generating straight into

dist

will cause compilation to potentially fail, or make

tsserver

unhappy. Creating the packages in another repo, outside of the monorepo is kinda losing the entire point of monorepo. Did anyone solve something similar to that?

Hey!

Hello, I’m willing to invest my free time to implement a Cassandra provider for Prisma. I love this database but there are not many good implementations for NodeJS/TypeScript, that’s why I think it’s a good fit to have it on Prisma. CQL is not that different from SQL, I’ve implemented a CQL builder a couple of years ago and I could start with the basic stuff without Materialized Views, Allow Filtering or Static Columns on this implementation. I commented recently on this ticket and I created a feature request for this more than a year ago on May 10, 2020. I did fork Prisma recently and I’m trying to get familiar with the codebase, but it doesn’t seem to have any kind of adaptor or specific files for each provider. It would be good if I could get a bit of guideness on where to look at.

Hello, I'm considering using Prisma on some future projects and my team relies quite a bit on PostGIS for spatial features. I'm glad to see that support for them is on the roadmap but I understand that the team probably has a lot of complex features to tackle before then and so far it looks like there are some reasonable escape hatches with

@Unsupported

and

queryRaw()

. My question is has there been any open discussion as to what the Prisma Client API could look like for spatial features? I think it could be extremely powerful to have equivalents of PostGIS functions like

ST_Contains

and

ST_Intersects

in the Prisma client for doing spatial joins.

how do i delete an entry from a many to many relation. Supposing i have User, Classroom, and userClassrooms

Trying something like

await this.prisma.userClassrooms.delete({
      where: {
        user_id: user.id,
        classroom_id: classroomId,
      },
    });

Hi Is it correct that I took prisma in libs/prisma/src/lib/schema.prisma with nestjs and nx How do I make migrations and run sids? I did this https://github.com/TemaOvchinnikov/olympus-social-network/blob/main/package.json

"prisma": {

"seed": "ts-node libs/prisma/src/lib/seed.ts",

"schema": "./libs/prisma/src/lib/schema.prisma"

`}`the syds don't work(

"seed": "node --loader ts-node/esm libs/prisma/src/lib/seed.ts",

works like this)🚲