pgCompare

Data Compare

---

Data Compare Made Simple

pgCompare is a Java-based tool for validating data consistency after replication or migration between databases. It's designed for scenarios like:

• Data migration from Oracle/DB2/MariaDB/MySQL/MSSQL to Postgres: Compare data during or post-migration.

• Logical replication between same or different database platforms: Validate data across platforms while minimizing database overhead.

• Active-Active replication configuration: Regularly verify data consistency to mitigate risks.

pgCompare uses hashing to compare table data efficiently. Hash values for primary keys and remaining columns are stored in a repository, reducing storage and network demands. Comparisons are processed in parallel, improving performance.

This open-source project is maintained by Crunchy Data under the Apache 2.0 License and is made available for broader use, testing, and feedback.

Features

• Supports Oracle, PostgreSQL, DB2, MariaDB, MySQL, and MSSQL.
• Efficient parallel comparisons using hashing.
• Handles batch processing for performance tuning.
• Stores configurations for multiple comparison projects in a central repository.

Installation

Requirements

Before initiating the build and installation process, ensure the following prerequisites are met:

1. Java 21 or later.
2. Maven 3.9 or later.
3. Postgres 15 or later (for the repository).
4. Supported JDBC drivers (Snowflake, DB2, Postgres, MySQL, MSSQL and Oracle currently supported).
5. Direct Postgres connections (e.g., no pgBouncer).

Limitations

• Date/Timestamps compared only to the second (format: DDMMYYYYHH24MISS).
• Unsupported data types: blob, long, longraw, bytea.
• Cross-platform comparison limitations with boolean type.
• Low precission types (float, real) cannot be compared to high precission types (double).
• All low precission types are cast using a scale of 3 (1 for Snowflake). If a higher scale is required consider using the map-expression override option.
• Different databases cast float to different values. Use float-cast option to switch between char and notation (scientific notation) if there are compare problems with float data types.

Getting Started

1. Fork the repository

2. Clone and Build

git clone --depth 1 git@github.com:<your-github-username>/pgCompare.git
cd pgCompare
mvn clean install

3. Configure Properties

Copy pgcompare.properties.sample to pgcompare.properties and update the connection parameters for your repository, source, and target databases. By default, the application looks for the properties file in the execution directory. Use PGCOMPARE_CONFIG environment variable to specify a custom properties file location.

At a minimal the repo-xxxxx parameters are required in the properties file (or specified by environment parameters). Besides the properties file and environment variables, another alternative is to store the property settings in the dc_project table. Settings can be stored in the project_config column in JSON format ({"parameter": "value"}). Certain system parameters like log-destination can only be specified via the properties file or environment variables.

4. Initialize Repository

Run the script or use the command below to set up the PostgreSQL repository:

java -jar pgcompare.jar init

5. Discover Tables

Discover and map tables in specified schemas:

java -jar pgcompare.jar discover

Usage

Command Line

java -jar pgcompare.jar <action> <options>

Actions:

• check: Recompare the out of sync rows from previous compare
• compare: Perform database compare
• copy-table: Copy pgCompare metadata for table. Must specify table alias to copy using --table option
• discover: Disocver tables and columns
• init: Initialize the repository database

Options:

-b|--batch {batch nbr}

-p|--project Project ID

-r|--report {file} Create html report of compare

-t|--table {target table}

--help

Define Table Mapping

1. Automatic Discovery

   Discover and map tables in specified schemas:

   java -jar pgcompare.jar discover

2. Manual Registration

   Insert mappings into dc_table, dc_table_map, dc_table_column, and dc_table_column_map tables in the repository.

Run Data Comparison

java -jar pgcompare.jar compare --batch 0

Batch 0 processes all data. Use PGCOMPARE-BATCH or specify the batch number using the --batch argument to specify a batch number.

Recheck Discrepancies

Revalidate flagged rows:

java -jar pgcompare.jar check --batch 0

Upgrading

Version 0.5.0 Enhacements

• Snowflake Support - Full integration for Snowflake as source/target
• SQL Fix Generation - Automatic generation of INSERT/UPDATE/DELETE statements (Preview, limited ability)
• Web UI - Modern Next.js-based interface (preview)
• Performance Improvements
• Bug Fixes

Note: Drop and recreate the repository to upgrade to 0.5.0.

For more details review the v0.5.0 Release Noes

Version 0.4.0 Enhancements

• Improved casting of low precision data types.
• Added html report generation.
• Refactored code for efficiency.
• Modified arguments and added 'verb' clause to command line.

Note: Drop and recreate the repository to upgrade to 0.4.0.

Version 0.3.0 Enhancements

• DB2 support.
• Case-sensitive table/column name handling.
• New project configurations for easier management.

Note: Drop and recreate the repository to upgrade to 0.3.0.

Advanced Configuration

Properties

Define properties via a file, environment variables, or the dc_project table. Environment variables override file settings and must be prefixed with PGCOMPARE_.

Examples:

• File: batch-fetch-size=2000
• Env: PGCOMPARE_BATCH_FETCH_SIZE=2000

Tuning Performance

• Batch size: Adjust batch-fetch-size and batch-commit-size for memory efficiency.
• Threads: Use loader-threads (default: 4) for parallel processing.
• Observer throttle: Enable to prevent overloading temporary tables (observer-throttle=true).
• Java Heap Size: For larger datasets, there may be a need to increase the Java Heap size. Use the options -Xms and -Xmx when executing pgCompare (java -Xms512m -Xmx2g -jar pgcompare.jar).

Repository Recommendations

• Minimal requirements: 2 vCPUs, 8 GB RAM.
• PostgreSQL settings:
  • shared_buffers=2048MB
  • work_mem=256MB
  • max_parallel_workers=16

Projects

Projects allow for the repository to maintain different mappings for different compare objectives. This allows a central pgCompare repository to be used for multiple compare projects. Each table has a pid column which is the project id. If no project is specified, the default project (pid = 1) is used.

Viewing Results

Summary from Last Run

WITH mr AS (SELECT max(rid) rid FROM dc_result)
SELECT compare_start, table_name, status, equal_cnt+not_equal_cnt+missing_source_cnt+missing_target_cnt  AS total_cnt,
       equal_cnt, not_equal_cnt, missing_source_cnt + missing_target_cnt AS missing_cnt
FROM dc_result r
         JOIN mr ON (mr.rid = r.rid)
ORDER BY table_name;

Out-of-Sync Rows

SELECT COALESCE(s.table_name, t.table_name) AS table_name,
       CASE
           WHEN s.compare_result = 'n' THEN 'out-of-sync'
           WHEN s.compare_result = 'm' THEN 'missing target'
           WHEN t.compare_result = 'm' THEN 'missing source'
           END AS compare_result,
       COALESCE(s.pk, t.pk) AS primary_key
FROM dc_source s
         FULL OUTER JOIN dc_target t ON s.pk = t.pk and s.tid=t.tid;

Reference

Properties

Properties are categorized into four sections: system, repository, source, and target. Each section has specific properties, as described in detail in the documentation. The properties can be specified via a configuration file, environment variables or a combination of both. To use environment variables, the environment variable will be the name of the property in upper case with dashes '-' converted to underscore '' and prefixed with PGCOMPARE. For example, batch-fetch-size can be set by using the environment variable PGCOMPARE_BATCH_FETCH_SIZE.

System

batch-fetch-size

Sets the fetch size for retrieving rows from the source or target database.

Default: 2000

batch-commit-size

The commit size controls the array size and number of rows concurrently inserted into the dc_source/dc_target staging tables.

Default: 2000

batch-progress-report-size

Defines the number of rows used in mod to report progress.

Default: 1000000

column-hash-method

Determines how the hash is performed. Valid values are database and hybrid. When set to database the column value hash is performed on the source/target database. For hybrid the hash is performed by the pgCompare thread.

Default: database

database-sort

Determines if the sorting of the rows based on primary key occurs on the source/target database. If set to true, the default, the rows will be sorted before being compared. If set to false, the sorting will take place in the repository database.

Default: true

float-scale

Set the preferred scale used to cast low precision numbers.

Default: 3

loader-threads

Sets the number of threads to load data into the temporary tables. Set to 0 to disable loader threads.

Default: 0

log-level

Level to determine the amount of log messages written to the log destination.

Default: INFO

log-destination

Location where log messages will be written.

Default: stdout

message-queue-size

Size of message queue used by loader threads (nbr messages).

Default: 100

number-cast

Defines how numbers are cast for hash function (notation|standard). Valid values are notation for scientific notation and standard for standard number casting.

Default: notation

observer-throttle

Set to true or false, instructs the loader threads to pause and wait for the observer thread to catch up before continuing to load more data into the staging tables.

Default: true

observer-throttle-size

Number of rows loaded before the loader thread will sleep and wait for clearance from the observer thread.

Default: 2000000

observer-vacuum

Set to true or false, instructs the observer whether to perform a vacuum on the staging tables during checkpoints.

Default: true

stage-table-parallel

Default parallel degree to set on staging table.

Default: 0

standard-number-format

Format used to cast numbers

Default: 0000000000000000000000.0000000000000000000000

Repository

repo-dbname

Repository database name.

repo-host

Host name of server hosting the Postgres repository database.

repo-password

Postgres database user password.

repo-port

Repository Postgres instance port.

repo-schema

Name of schema that owns the repository tables.

repo-sslmode

Set the SSL mode to use for the database connection (disable|prefer|require)

repo-user

Postgres database username.

Source|Target

source|target-dbname

Database or service name.

source|target-host

Database server name.

source|target-password

Database password.

source|target-port

Database port.

source|target-schema

Name of schema that owns the tables.

source|target-sslmode

Set the SSL mode to use for the database connection (disable|prefer|require)

source|target-type

Database type: oracle, postgres, mariadb, mssql, mysql, snowflake, db2

source|target-user

Database username.

source|target-warehouse

Used only for Snowflake, sets the virtual warehouse to be used for the compare operations.

Property Precedence

The system contains default values for every parameter. These can be over-ridden using environment variables, properties file, or values saved in the dc_project table. The following is the order of precedence used:

• Default values
• Properties file
• Environment variables
• Settings stored in dc_project table

License

pgCompare is licensed under the Apache 2.0 license.