dbt enables data analysts and engineers to transform their data using the same practices that software engineers use to build applications.

dbt is the T in ELT. Organize, cleanse, denormalize, filter, rename, and pre-aggregate the raw data in your warehouse so that it's ready for analysis.

ODPS

ODPS, called MaxCompute before. This adapter is a wrapper bridged PyOdps and DBT together.

MaxCompute Features:

Feature	Status
Partition Table	❎
Cluster Table	❎
External Table	❎
Table Properties	❎

DBT features:

Name	Status
Materialization: Table	✅
Materialization: View	✅
Materialization: Incremental - Append	✅
Materialization: Incremental - Insert+Overwrite	✅
Materialization: Incremental - Merge	✅
Materialization: Ephemeral	✅
Seeds	✅
Tests	✅
Snapshots	✅1
Documentation	✅

Getting Started

Install DBT-ODPS

python setup.py install --force 
or pip install dbt-odps-winwin

How to Configure?

For Fresh Codebase

Run following command after installing dbt-odps:

dbt init

For Existing Codebase

Read more in here: https://docs.getdbt.com/docs/core/connection-profiles

Configuration options:

Property	Description	Example
Endpoint	The endpoint of odps, read more in https://help.aliyun.com/document_detail/34951.html	http://service.cn-shanghai.maxcompute.aliyun.com/api
database	The project name of odps, which you can find in https://maxcompute.console.aliyun.com/{your area}/project-list	odps-test-project
schema	Using default if you don't know what is schema.	default
access_id	access id	LTAXXXXXXXXX
secret_access_key	secret access key	bZXXXXXXXXXX
type	odps	odps

NOTES

1. When using merge statement, ODPS required that table is a transactional table. So, we have to create the snapshot table before select. Under the hook, we using the first referred table as source data structure to create table, so this data source must be a table, view is not supported.

DEVELOPER REF

DBT docs what-are-adapters