Very slow nested writes when connecting with remote DB Postrgsql on AWS

I am currently testing different ORMs to find a good fit for our project. I write a very basic query with a nested write in Prisma which takes an extremly long time.

See my schema:

model Context {
  id          Int         @id @default(autoincrement())
  contextName String
  createdAt   DateTime    @default(now())
  statements  Statement[]
}

model Statement {
  id        Int      @id @default(autoincrement())
  content   String
  context   Context  @relation(fields: [contextId], references: [id], onDelete: Cascade)
  contextId Int
  createdAt DateTime @default(now())
}

I create a new context with around 600 statements inside. This the code sample:

const context = await this.prisma.context.create({
      data: {
        contextName: contextData.name,
        statements: {
          create: statements,
        },
      },
      include: {
        statements: true,
      },
    });

I’ve tried to debug the request but haven’t noticed anything suspicious from the Prisma output.

Check out the repo to see the full code and the request.

If I write to the local DB with prisma it takes around 4s vs to remote DB is more than 100s

The same operation with typeorm takes around 2s to the remote DB.

Any hints on how can improve the query?

_Originally posted by @monapasan in https://github.com/prisma/prisma/discussions/9552_

[TasinIshmam]

It is not. See here. Moreover you make a point on your website when comparing typeorm and prisma about nested writes. I hope this has not been left out intentionally. Many thanks for your reply and your help! And let me know if I should create a PR to the docs with this information. Happy to contribute!

I understand your point. Thanks for agreeing to help, I have already created an issue about this in the docs repository. Please feel free to follow it or drop a comment if you have any suggestions!

And you can be assured this wasn’t an intentional omission 😅

[TasinIshmam]

This solution worked very well. It’s fast enough. This also has fixed my issue with MongoDB. Thank you so much!

Awesome. Glad I could help 😄

This is such a crucial point. I am wondering why do you not mention this in the API docs? Specifically, that you create a separate query for every nested write?

This is a very good point. We need to make this information more discoverable if it isn’t already. I will make sure to bring this up internally 👌

Btw I’m closing this issue as I’m assuming your problem has been resolved.