Database integration with nestrs

nestrs supports SQL via SqlxDatabaseModule for direct AnyPool queries or nestrs-prisma for Prisma-style schema-driven access. MongoDB support is also available.

nestrs gives you two SQL integration paths depending on how much structure you want. SqlxDatabaseModule (feature database-sqlx) wraps an SQLx AnyPool and is the right choice when you want direct control over SQL queries with minimal abstraction. nestrs-prisma adds a PrismaModule, a PrismaService, and the prisma_model! macro for a Prisma-style developer experience without requiring a separate codegen step. MongoDB lives under a third path via MongoModule.

Path 1: SqlxDatabaseModule (direct SQLx)

SqlxDatabaseModule manages a single AnyPool shared across all injected consumers. Call for_root before NestFactory::create to register the URL, then import the module.

Cargo.toml

[dependencies]
nestrs = { version = "0.3.8", features = ["database-sqlx"] }
sqlx = { version = "0.8", default-features = false, features = ["runtime-tokio", "macros", "postgres"] }
tokio = { version = "1", features = ["macros", "rt-multi-thread"] }

Environment variable

export DATABASE_URL="postgresql://user:password@127.0.0.1:5432/myapp"
# or for SQLite:
# export DATABASE_URL="sqlite:./dev.db"

Module registration

use nestrs::prelude::*;

#[tokio::main]
async fn main() {
    let db_url = std::env::var("DATABASE_URL").expect("DATABASE_URL must be set");
    let _ = SqlxDatabaseModule::for_root(db_url);

    NestFactory::create::<AppModule>()
        .listen(3000)
        .await;
}

#[module(
    imports = [SqlxDatabaseModule],
    providers = [UserService],
    controllers = [UserController],
)]
struct AppModule;

Injecting SqlxDatabaseService

SqlxDatabaseService exposes the pool via pool() and a ping() health check. Use pool() to run any SQLx query directly.

use nestrs::prelude::*;
use std::sync::Arc;

#[injectable]
pub struct UserService {
    db: Arc<SqlxDatabaseService>,
}

impl UserService {
    pub async fn health(&self) -> &'static str {
        match self.db.ping().await {
            Ok(_) => "up",
            Err(_) => "degraded",
        }
    }

    pub async fn user_count(&self) -> Result<i64, String> {
        let pool = self.db.pool().await?;
        let row: (i64,) = sqlx::query_as("SELECT COUNT(*) FROM users")
            .fetch_one(pool)
            .await
            .map_err(|e| e.to_string())?;
        Ok(row.0)
    }
}

---

Path 2: nestrs-prisma

nestrs-prisma ships PrismaModule and PrismaService with higher-level helpers: query_all_as for typed row mapping, execute for DDL and DML, and prisma_model! for declarative repositories that generate find_unique, find_many, create, update, delete, and more.

Cargo.toml

[dependencies]
nestrs = "0.3.8"
nestrs-prisma = { version = "0.3.8", features = ["sqlx", "sqlx-postgres"] }
async-trait = "0.1"
sqlx = { version = "0.8", default-features = false, features = ["runtime-tokio", "macros", "postgres"] }
serde = { version = "1", features = ["derive"] }
tokio = { version = "1", features = ["macros", "rt-multi-thread"] }

Enable exactly one SQLx backend feature: sqlx-postgres, sqlx-mysql, or sqlx-sqlite. async-trait must be a direct dependency of any crate that uses prisma_model!.

PostgreSQL

nestrs-prisma = { version = "0.3.8", features = ["sqlx", "sqlx-postgres"] }

export DATABASE_URL="postgresql://user:password@127.0.0.1:5432/myapp"

MySQL

nestrs-prisma = { version = "0.3.8", features = ["sqlx", "sqlx-mysql"] }

export DATABASE_URL="mysql://user:password@127.0.0.1:3306/myapp"

SQLite

nestrs-prisma = { version = "0.3.8", features = ["sqlx", "sqlx-sqlite"] }
sqlx = { version = "0.8", default-features = false, features = ["runtime-tokio", "macros", "sqlite"] }

export DATABASE_URL="sqlite:./dev.db"
# or for tests:
# DATABASE_URL="sqlite::memory:?cache=shared"

Schema

Create a Prisma schema file at prisma/schema.prisma. nestrs-prisma reads the schema for optional sync but does not require the Prisma CLI at runtime.

datasource db {
  provider = "postgresql"
  url      = env("DATABASE_URL")
}

model User {
  id    Int    @id @default(autoincrement())
  email String @unique
  name  String
}

Apply the schema to your database:

npx prisma migrate dev
# or for quick local work:
npx prisma db push

Bootstrap PrismaModule

Call PrismaModule::for_root_with_options before NestFactory::create so the pool is ready before DI builds providers.

use nestrs::prelude::*;
use nestrs_prisma::{PrismaModule, PrismaOptions};

#[tokio::main]
async fn main() {
    let db_url = std::env::var("DATABASE_URL").expect("DATABASE_URL must be set");

    let _ = PrismaModule::for_root_with_options(
        PrismaOptions::from_url(db_url)
            .pool_min(1)
            .pool_max(10),
    );

    NestFactory::create::<AppModule>()
        .listen(3000)
        .await;
}

#[module(
    imports = [PrismaModule],
    re_exports = [PrismaModule],
)]
pub struct DataModule;

#[module(
    imports = [DataModule],
    controllers = [UserController],
    providers = [UserService],
)]
pub struct AppModule;

Raw queries with PrismaService

query_all_as maps rows to any type that implements sqlx::FromRow. execute runs DDL or parameterless DML. query_scalar is useful for health checks.

use nestrs::prelude::*;
use nestrs_prisma::PrismaService;
use std::sync::Arc;

#[derive(Debug, serde::Serialize, sqlx::FromRow)]
pub struct UserRow {
    pub id: i64,
    pub email: String,
    pub name: String,
}

#[injectable]
pub struct UserService {
    prisma: Arc<PrismaService>,
}

impl UserService {
    pub async fn health(&self) -> &'static str {
        match self.prisma.query_scalar("SELECT 1").await {
            Ok(_) => "up",
            Err(_) => "degraded",
        }
    }

    pub async fn list_users(&self) -> Result<Vec<UserRow>, String> {
        self.prisma
            .query_all_as(r#"SELECT "id", "email", "name" FROM "User" ORDER BY "id""#)
            .await
    }
}

Declarative repositories with prisma_model!

prisma_model! generates a full repository from a table declaration. The macro expands a struct, CreateInput, Where, Update, OrderBy, and a PrismaUserRepository trait — all accessible via prisma.user().

nestrs_prisma::prisma_model!(User => "users", {
    id: i64,
    email: String,
    name: String,
});

After the macro expands, you can use the full repository API:

impl UserService {
    pub async fn create_user(&self, email: &str, name: &str) -> Result<UserRow, HttpException> {
        let row = self
            .prisma
            .user()
            .create(UserCreateInput {
                email: email.into(),
                name: name.into(),
            })
            .await
            .map_err(HttpException::from)?;
        Ok(UserRow { id: row.id, email: row.email, name: row.name })
    }

    pub async fn find_by_id(&self, id: i64) -> Result<UserRow, HttpException> {
        let u = self
            .prisma
            .user()
            .find_unique(user::id::equals(id))
            .await
            .map_err(HttpException::from)?
            .ok_or_else(|| NotFoundException::new("user"))?;
        Ok(UserRow { id: u.id, email: u.email, name: u.name })
    }
}

PrismaError implements Into<HttpException>. Map errors with .map_err(HttpException::from) or ? when the return type is Result<_, HttpException>.

HTTP controller

#[controller(prefix = "/users", version = "v1")]
pub struct UserController;

#[routes(state = UserService)]
impl UserController {
    #[get("/")]
    pub async fn list(State(s): State<Arc<UserService>>) -> Result<Json<Vec<UserRow>>, HttpException> {
        s.list_users().await.map(Json).map_err(InternalServerErrorException::new)
    }

    #[get("/health/db")]
    pub async fn db_health(State(s): State<Arc<UserService>>) -> &'static str {
        s.health().await
    }
}

Run the quickstart example

The nestrs-prisma crate ships a full end-to-end example with two related models, CRUD operations, and schema sync:

cargo run -p nestrs-prisma --example quickstart --features "sqlx,sqlx-sqlite"

---

Path 3: MongoModule

MongoModule wraps the official mongodb driver. Call for_root with a connection URI before NestFactory::create, then inject MongoService to access databases and collections.

Cargo.toml

[dependencies]
nestrs = { version = "0.3.8", features = ["mongo"] }
mongodb = "3"
bson = "3"
futures-util = { version = "0.3", features = ["sink"] }
serde = { version = "1", features = ["derive"] }
tokio = { version = "1", features = ["macros", "rt-multi-thread"] }

Bootstrap and inject

#[tokio::main]
async fn main() {
    let _ = nestrs::MongoModule::for_root(
        std::env::var("MONGODB_URI").unwrap_or_else(|_| "mongodb://127.0.0.1:27017".into()),
    );

    NestFactory::create::<AppModule>()
        .listen(3000)
        .await
        .expect("server");
}

#[module(imports = [MongoModule], controllers = [ProfileController], providers = [ProfileStore])]
pub struct AppModule;

Service with typed collections

use bson::doc;
use futures_util::TryStreamExt;
use nestrs::prelude::*;
use serde::{Deserialize, Serialize};
use std::sync::Arc;

#[derive(Debug, Serialize, Deserialize)]
pub struct ProfileDoc {
    pub email: String,
    pub display_name: String,
}

#[injectable]
pub struct ProfileStore {
    mongo: nestrs::MongoService,
}

impl ProfileStore {
    pub async fn find_by_email(&self, email: &str) -> Result<Option<ProfileDoc>, String> {
        let db = self.mongo.database("app").await?;
        let col = db.collection::<ProfileDoc>("profiles");
        col.find_one(doc! { "email": email })
            .await
            .map_err(|e| e.to_string())
    }
}

Troubleshooting

Symptom	What to check
sqlx pool / TLS errors	DATABASE_URL scheme matches the sqlx-* feature; Postgres often needs sslmode in the URL.
”feature not enabled”	Exactly one of sqlx-postgres, sqlx-mysql, or sqlx-sqlite on nestrs-prisma.
In-memory SQLite flaky	Set pool_max(1) — in-memory SQLite works most reliably on a single connection.
Macro / trait errors on prisma_model!	Add async-trait as a direct dependency in your Cargo.toml.
”MongoModule must be called…”	MongoModule::for_root must run before NestFactory::create in the same process.