Skip to content

Latest commit

 

History

History
159 lines (112 loc) · 3.66 KB

dbt.md

File metadata and controls

159 lines (112 loc) · 3.66 KB
Raw

DBT - Data Build Tool - Setup Guide

  • Setup Dremio Locally on our Laptop (only needed if don't already have Dremio environment)
  • Setup Python Environment
  • Configure your DBT Profile

Setup Dremio (For Setting Up an Evaluation Environment to Practice in)

  • Create a docker-compose.yml
version: "3"

services:
  # Nessie Catalog Server Using In-Memory Store
  catalog:
    image: projectnessie/nessie:0.76.0
    container_name: catalog
    networks:
      dremio-laptop-lakehouse:
    ports:
      - 19120:19120
  # Minio Storage Server
  storage:
    image: minio/minio:RELEASE.2024-01-01T16-36-33Z
    container_name: storage
    environment:
      - MINIO_ROOT_USER=admin
      - MINIO_ROOT_PASSWORD=password
      - MINIO_DOMAIN=storage
      - MINIO_REGION_NAME=us-east-1
      - MINIO_REGION=us-east-1
    networks:
      dremio-laptop-lakehouse:
    ports:
      - 9001:9001
      - 9000:9000
    command: ["server", "/data", "--console-address", ":9001"]
  # Dremio
  dremio:
    platform: linux/x86_64
    image: dremio/dremio-oss:latest
    ports:
      - 9047:9047
      - 31010:31010
      - 32010:32010
    container_name: dremio
    networks:
      dremio-laptop-lakehouse:
networks:
  dremio-laptop-lakehouse:

Setup Python Environment, install dbt

  • python -m venv venv

  • source ./venv/bin/activate

  • pip install dbt-dremio

Create a dbt project

The end result will create a profile which can be found in ~/.dbt/profiles.yml

  • dbt init <projectname>

  • select dremio

  • select dremio with software username/password

  • put 127.0.0.1 as host

  • use 9047 as port

  • put username and password (or username/PAT if using Dremio cloud or choose software with PAT)

  • use the name of nessie/metastore/object storage source for "object_storage_soure" (can also be as arctic catalog for cloud)

  • write a path to a subfolder in that source for "object_storage_path"

  • write the name of a space for "dremio_space" (for Dremio Cloud, this should be the name of an Arctic catalog)

  • write the path to a subfolder in your space for "dremio_space_path"

  • select 1 thread

dbt function

{{ config() }}

Configures the behavior for the following model.

Example Arguments:

  • materialized: view to create a sql view or table to create a table

  • database: The dremio space (view) or source (table) to create the result in

  • schema: the path to a subfolder in the source to out the results.

{{ ref() }}

Reference to a source model. This ensure that the referenced model will be run before this model.

dbt commands

  • dbt init <projectname> - create a new project
  • dbt run run dbt models

Using the project.yml to configure groups of models

You can specify different locations for different models by folder or tag like below.

models:
  my_project:
    # Apply to all models
    +materialized: view

    # Configuration for models in a specific folder
    marketing:
      +schema: marketing_schema
      +database: marketing_database

    # Configuration for models with a specific naming pattern
    tag:bi:
      +materialized: table

You can can tag different models like so...

In a single model:

{{ config(tags=["daily", "analytics"]) }}

Groups of models from project.yml

models:
  my_project:
    # Apply tags to all models in a specific directory
    marketing:
      +tags: ["marketing", "weekly"]
    # Apply tags to a specific model
    my_model:
      +tags: ["core_model", "daily"]

You can even only run models with certain tags using:

dbt run --model tag:daily