Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

https://github.com/datafold/data-diff

Compare tables within or across databases
https://github.com/datafold/data-diff

data data-diffing data-engineering data-quality data-quality-monitoring data-science database databricks-sql dataengineering dataquality dbt mysql oracle-database postgres postgresql python rdbms snowflake sql trino

Last synced: about 1 month ago
JSON representation

Compare tables within or across databases

Lists

README 

        


    Datafold






data-diff: Compare datasets fast, within or across SQL databases



![data-diff-logo](docs/data-diff-logo.png)







> [Join our live virtual lab series to learn how to set it up!](https://www.datafold.com/virtual-hands-on-lab)



# What's a Data Diff?

A data diff is the value-level comparison between two tables—used to identify critical changes to your data and guarantee data quality.



There is a lot you can do with data-diff: you can test SQL code by comparing development or staging environment data to production, or compare source and target data to identify discrepancies when moving data between databases.



# data-diff OSS & Datafold Cloud

data-diff is an open source utility for running stateless diffs as a great single player experience.



Scale up with [Datafold Cloud](https://www.datafold.com/) to make data diffing a company-wide experience to both supercharge your data diffing CLI experience (ex: data-diff --dbt --cloud) and run diffs manually in your CI process and within the Datafold UI. This includes [column-level lineage](https://www.datafold.com/column-level-lineage) with BI tool integrations, [CI testing](https://docs.datafold.com/deployment_testing/how_it_works/), faster cross-database diffing, and diff history.



# Use Cases



### Data Development Testing

When developing SQL code, data-diff helps you validate and preview changes by comparing data between development/staging environments and production. Here's how it works:

1. Make a change to your SQL code

2. Run the SQL code to create a new dataset

3. Compare this dataset with its production version or other iterations



### Data Migration & Replication Testing

data-diff is a powerful tool for comparing data when you're moving it between systems. Use it to ensure data accuracy and identify discrepancies during tasks like:

- **Migrating** to a new data warehouse (e.g., Oracle -> Snowflake)

- **Validating SQL transformations** from legacy solutions (e.g., stored procedures) to new transformation frameworks (e.g., dbt)

- Continuously **replicating data** from an OLTP database to OLAP data warehouse (e.g., MySQL -> Redshift)



# dbt Integration

  


  dbt

  



data-diff integrates with [dbt Core](https://github.com/dbt-labs/dbt-core) to seamlessly compare local development to production datasets.



Learn more about how data-diff works with dbt:

* Read our docs to get started with [data-diff & dbt](https://docs.datafold.com/development_testing/cli) or :eyes: **watch the [4-min demo video](https://www.loom.com/share/ad3df969ba6b4298939efb2fbcc14cde)**

* dbt Cloud users should check out [Datafold's out-of-the-box deployment testing integration](https://www.datafold.com/data-deployment-testing)

* Get support from the dbt Community Slack in [#tools-datafold](https://getdbt.slack.com/archives/C03D25A92UU)



# Getting Started



### ⚡ Validating dbt model changes between dev and prod

Looking to use data-diff in dbt development?



Development testing with Datafold enables you to see the impact of dbt code changes on data as you write the code, whether in your IDE or CLI.



 Head over to [our `data-diff` + `dbt` documentation](https://docs.datafold.com/development_testing/cli) to get started with a development testing workflow!



### 🔀 Compare data tables between databases

1. Install `data-diff` with adapters



To compare data between databases, install `data-diff` with specific database adapters. For example, install it for PostgreSQL and Snowflake like this:



```

pip install data-diff 'data-diff[postgresql,snowflake]' -U

```



Additionally, you can install all open source supported database adapters as follows.

```

pip install data-diff 'data-diff[all-dbs]' -U

```



2. Run `data-diff` with connection URIs



Then, we compare tables between PostgreSQL and Snowflake using the hashdiff algorithm:



```bash

data-diff \

  postgresql://:''@localhost:5432/ \

   \

  "snowflake://:@//?warehouse=&role=" \

   \

  -k  \

  -c  \

  -w 

```

3. Set up your configuration



You can use a `toml` configuration file to run your `data-diff` job. In this example, we compare tables between MotherDuck (hosted DuckDB) and Snowflake using the hashdiff algorithm:



```toml

## DATABASE CONNECTION ##

[database.duckdb_connection]

  driver = "duckdb"

  # filepath = "datafold_demo.duckdb" # local duckdb file example

  # filepath = "md:" # default motherduck connection example

  filepath = "md:datafold_demo?motherduck_token=${motherduck_token}" # API token recommended for motherduck connection



[database.snowflake_connection]

  driver = "snowflake"

  database = "DEV"

  user = "sung"

  password = "${SNOWFLAKE_PASSWORD}" # or ""

  # the info below is only required for snowflake

  account = "${ACCOUNT}" # by33919

  schema = "DEVELOPMENT"

  warehouse = "DEMO"

  role = "DEMO_ROLE"



## RUN PARAMETERS ##

[run.default]

  verbose = true



## EXAMPLE DATA DIFF JOB ##

[run.demo_xdb_diff]

  # Source 1 ("left")

  1.database = "duckdb_connection"

  1.table = "development.raw_orders"



  # Source 2 ("right")

  2.database = "snowflake_connection"

  2.table = "RAW_ORDERS" # note that snowflake table names are case-sensitive



  verbose = false

```

4. Run your `data-diff` job



Make sure to export relevant environment variables as needed. For example, we compare data based on the earlier configuration:



```bash



# export relevant environment variables, example below

export motherduck_token=



# run the configured data-diff job

data-diff --conf datadiff.toml \

  --run demo_xdb_diff \

  -k "id" \

  -c status



# output example

- 1, completed

+ 1, returned

```



5. Review the output



After running your data-diff job, review the output to identify and analyze differences in your data.



Check out [documentation](https://docs.datafold.com/reference/open_source/cli) for the full command reference.



# Supported databases



| Database      | Status | Connection string                                                                                                                   |

|---------------|-------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------|

| PostgreSQL >=10 |  🟢   | `postgresql://:@:5432/`                                                                             |

| MySQL         |  🟢     | `mysql://:@:5432/`                                                                              |

| Snowflake     |  🟢     | `"snowflake://[:]@//?warehouse=&role=[&authenticator=externalbrowser]"` |

| BigQuery      |  🟢     | `bigquery:///`                                                                                                    |

| Redshift      |  🟢     | `redshift://:@:5439/`                                                                       |

| DuckDB        |  🟢   | `duckdb://`                                                                                          |

| MotherDuck        |  🟢   | `duckdb://`                                                                                                   |

| Microsoft SQL Server*        |  🟢   | `mssql://:@//`                                                                                                   |

| Oracle        |  🟡   | `oracle://:@/servive_or_sid`                                                                          |

| Presto        |  🟡   | `presto://:@:8080/`                                                                         |

| Databricks    |  🟡   | `databricks://:@//`                                                      |

| Trino         |  🟡   | `trino://:@:8080/`                                                                          |

| Clickhouse    |  🟡   | `clickhouse://:@:9000/`                                                                     |

| Vertica       |  🟡   | `vertica://:@:5433/`                                                                        |



*MS SQL Server support is limited, with known performance issues that are addressed in Datafold Cloud.



* 🟢: Implemented and thoroughly tested.

* 🟡: Implemented, but not thoroughly tested yet.



Your database not listed here?



- Contribute a [new database adapter](https://github.com/datafold/data-diff/blob/master/docs/new-database-driver-guide.rst) – we accept pull requests!

- [Get in touch](https://www.datafold.com/demo) about enterprise support and adding new adapters and features







# How it works



`data-diff` efficiently compares data using two modes:



**joindiff**: Ideal for comparing data within the same database, utilizing outer joins for efficient row comparisons. It relies on the database engine for computation and has consistent performance.



**hashdiff**: Recommended for comparing datasets across different databases or large tables with minimal differences. It uses hashing and binary search, capable of diffing data across distinct database engines.



Click here to learn more about joindiff and hashdiff



### `joindiff`

* Recommended for comparing data within the same database

* Uses the outer join operation to diff the rows as efficiently as possible within the same database

* Fully relies on the underlying database engine for computation

* Requires both datasets to be queryable with a single SQL query

* Time complexity approximates JOIN operation and is largely independent of the number of differences in the dataset



### `hashdiff`:

* Recommended for comparing datasets across different databases

* Can also be helpful in diffing very large tables with few expected differences within the same database

* Employs a divide-and-conquer algorithm based on hashing and binary search

* Can diff data across distinct database engines, e.g., PostgreSQL <> Snowflake

* Time complexity approximates COUNT(*) operation when there are few differences

* Performance degrades when datasets have a large number of differences







For detailed algorithm and performance insights, explore [here](https://github.com/datafold/data-diff/blob/master/docs/technical-explanation.md), or head to our docs to [learn more about how Datafold diffs data](https://docs.datafold.com/data_diff/how-datafold-diffs-data).



## Contributors



We thank everyone who contributed so far!



We'd love to see your face here: [Contributing Instructions](CONTRIBUTING.md)





  








## Analytics



* [Usage Analytics & Data Privacy](https://github.com/datafold/data-diff/blob/master/docs/usage_analytics.md)







## License



This project is licensed under the terms of the [MIT License](https://github.com/datafold/data-diff/blob/master/LICENSE).