Trying out prisma 3, it’s not running properly in my lambda function/layer setup. I get the error

ENOENT: no such file or directory, open '/var/task/schema.prisma'

Can I at least tell it what path my

schema.prisma

file is at somehow? I couldn’t find anything in the docs I commented on https://github.com/prisma/prisma/issues/9435

Hello, @Ryan We are running 2 apps with prisma 2. in frontend side it looks like every apis response correctly, but as you can see, Some queries still remain and piles up until there is no connection slot available anymore and making the app crash. Do you have any idea?

anyone try using a lazy getter for accessing

PrismaClient

? in prisma 3 constructing the client now tries to load the schema.prisma file which I don’t have in my lambda functions that don’t use prisma (but do happen to import the file that constructs the client)

I just wanted to provide some timing info for running prisma 3.1.1 on a lambda function with 1024MB RAM and DEBUG=* I requested a simple function that does a single query to a table with 1 row. First as a cold start and then a second warm request. Looks like on cold start, initializing prisma took ~0.75s (14:55.075Z - 14:54.321Z) On a warm lambda the entire request took 923ms to run

prisma.tenant.findMany()

which seems like a while. It's a lambda function in a VPC talking to an aurora serverless pg database, I would expect it to be a bit faster than that. Just wanted to share in case this is useful information for someone.

any way to instrument prisma with xray? maybe using something like aws-xray-sdk-postgres?

Hello is it possible in Prisma schema to define (numeric)

CHECK

constraints ?

What does this error mean? I’m using prisma-appsync with ESM and my AWS lambda layer “Issue parsing prismaClientModels from auto-injected environment variable

PRISMA_APPSYNC_GENERATED_CONFIG

.” Prisma client loads just fine for all my other lambdas, only prisma-appsync seems to have the problem

is it possible to do nested connectOrCreate? for example

const album = {
        ...result,
        songs: {
          connectOrCreate: result.songs.map(s => ({
            create: {
                ...s,
                artists: {
                    connectOrCreate: s.artists.map(a => ({
                    create: a,
                    where: { spotify_id: a.spotify_id }
                }))
            },
            where: { spotify_id: s.spotify_id }
          }))
        }
      }

schema:
model Album {
   artists					Artist[]
   songs					Song[]
}

model Artist {
    albums    				Album[]
	songs					Song[]
}

model Song {
   album    				Album    	@relation(fields: [album_spotify_id], references: [spotify_id], onDelete: Cascade)
	artists					Artist[]
}

hi, ,anyone encountering "Field does not exist on enclosing type" even if the field exists?

Why is Prisma’s generated-client-code not checked into version control? Whenever a developer working in the codebase runs

git pull

, they must then run

prisma generate

to update their local environment for any schema changes others may have made. I’m surprised this is the default expectation because it is a hassle. We could create a git-hook to automatically regenerate on every

pull

, but it would be simpler to check-in the generated-client-code and add a CI check that ensures the client-code is up-to-date.

I keep getting

Already 10 Prisma Clients are actively running.

showing up in my terminal where I've ran

prisma studio

after a few minutes. I'm only instantiating the PrismaClient once when my app launches. This has been the case in the past year but I've only recently started getting this warning so I can't pinpoint the cause. Any help would be gladly appreciated

I have a schema of the following:

enum MemberRole {
  Owner
  Viewer
  Editor
  Unauthorized
}

model Member {
  id      String         @id @default(uuid())
  role    MemberRole @default(Unauthorized)
  pending Boolean?       @default(true)

  user   User   @relation(fields: [userId], references: [id])
  userId String

  createdAt DateTime @default(now())
  updatedAt DateTime @updatedAt
}

And using prisma client, I’m trying to create a new

Member

like the image attached. For some reason, the role property isn’t showing up even though I tried running generate, and

db push

multiple times. Am I doing something wrong?

Hi there! I need to check whether two dates overlap. Previously (with TypeORM) I used a raw query like so:

qb.andWhere(
            '("Appointment"."start_date", "Appointment"."end_date") OVERLAPS (:startDate, :endDate)',
            { startDate: startDate, endDate: endDate }
          );

Is there any way to make use of OVERLAPS in prisma without having to build the whole query yourself?

I encounter this error when I run

npx prisma migrate dev

Error: P3014

Prisma Migrate could not create the shadow database. Please make sure the database user has permission to create databases. Read more about the shadow database (and workarounds) at <https://pris.ly/d/migrate-shadow>

Original error: 
db error: ERROR: permission denied to create database
   0: sql_migration_connector::flavour::postgres::sql_schema_from_migration_history
             at migration-engine/connectors/sql-migration-connector/src/flavour/postgres.rs:354
   1: migration_core::api::DevDiagnostic
             at migration-engine/core/src/api.rs:108

The postgres database connects well

Is there any way I can generate typescript types to use on frontend code?

Trying to write a query that fetches users that have unread messages How do I express this in prisma? In particular the LEFT JOIN

https://www.prisma.io/docs/guides/testing/integration-testing#running-the-tests This says

"migrate:init": "npx prisma migrate dev --name init"

- why the

--name init

bit?

Having trouble with this queryRaw example from the docs. I had to edit it slightly to get it to type check properly

interface TableNameQueryResult {
  tablename: string
}
export async function truncateAllTables() {
  const qres = await prisma.$queryRaw<
    TableNameQueryResult[]
  >`SELECT tablename FROM pg_tables WHERE schemaname='public'`
  for (const { tablename } of qres) {
    if (tablename !== "_prisma_migrations") {
      // try {
      await prisma.$queryRaw`TRUNCATE TABLE "public"."${tablename}" CASCADE;`
      // } catch (error) {
      // console.log({ error })
      // }
    }
  }
}

It gives me this error:

Invalid `prisma.queryRaw()` invocation:


      Your raw query had an incorrect number of parameters. Expected: `0`, actual: `1`.

      81 |     if (tablename !== "_prisma_migrations") {
      82 |       // try {
    > 83 |       await prisma.$queryRaw`TRUNCATE TABLE "public"."${tablename}" CASCADE;`
         |       ^

hi, I have many-to-many relationship tables with User and Company where users can have many companies and vice-versa. How to create new users where client provides multiple company ids? I’m struggling to figure it out. Should I try in prisma.user.createMany? or prisma.usersCompanies.create()?

getting an error trying to add an enum type in a postgres migration:

Error: P3006

Migration `20211013101052_candidate_id_shouldnt_be_unique` failed to apply cleanly to the shadow database.
Error:
db error: ERROR: ALTER TYPE ... ADD cannot run inside a transaction block
   0: sql_migration_connector::flavour::postgres::sql_schema_from_migration_history
             at migration-engine/connectors/sql-migration-connector/src/flavour/postgres.rs:354
   1: sql_migration_connector::validate_migrations
             at migration-engine/connectors/sql-migration-connector/src/lib.rs:342
   2: migration_core::api::DevDiagnostic
             at migration-engine/core/src/api.rs:108

trying to run

ALTER TYPE "CandidateSource" ADD VALUE 'OTHER';

Why I’m eager for prisma generated clients to support ES modules:

Hello, is there any way to use the

include

API to automatically return all relational fields without specifying them?

Can you do an “optional”

include

? Like return my objects even if the relation doesn’t exist, like a

LEFT JOIN

situation?

Hello. Is it possible in Prisma to run

SELECT FOR UPDATE

statements ?

👋 Hi! I'm wondering if prisma is considering adding a new method to the prisma client to watch changes (change streams). I would love to have type-safe change streams. For mongodb it's easy with https://docs.mongodb.com/manual/changeStreams/ but it may be possible for SQL databases like PostgreSQL using triggers/notify. Any thoughts? It would allow to open prisma to real-time use cases too (graphql subscription, etc).

Hello. Basic question but can’t seem to find out anywhere in the docs. If I am using the postgres engine, is there a way to view the generated sql from a given prisma query (i.e.

prisma.model.findMany(...)

?

And also, is there any good articles on how to fine tune query performance?

Why does

prisma generate

say it is using a different version of Prisma Client than it has installed? Consequently, it wrongly warns about version mismatch. Thank you!

docs point to this closed ticket. someone said they need it for postgres but I don't know where to go +1 that. i need bulk pg upserts.

Hi, I have case where I have two entities

User

and

PhoneNumber

. Phone number can be assigned to only one User. When creating new User I need to assign a phone number, which is coming in request, for created user. The request must fail if phone number is already in use and I also need to take into account possible concurrent requests. My problem is that how I can check that phone number is not already in use? Is the Interactive transactions only way to achieve this?