MongoChunkSplitter

中文 / EN

MongoChunkSplitter is a MongoDB sharding tool designed to facilitate pre-sharding and pre-chunking operations for MongoDB users. This tool offers a set of Python functions to calculate shard key split points and to automate sharding and chunk moving operations in a MongoDB cluster.

Features

• Supports multiple character sets for defining split points.
• Enables automatic sharding to any number of shards without requiring any data in the collection.
• Automatically calculates split points based on data size and chunk size.
• Supports execution of sharding operations through command line or function calls.
• Provides a retry mechanism for enhanced reliability.

Installation

Installation packages are not currently available. Clone the GitHub repository to use:

git clone https://github.com/finishy1995/MongoChunkSplitter.git
cd MongoChunkSplitter

Usage Instructions

As a Python Library

1. Import the MongoChunkSplitter module.
2. Create a MongoClient instance and connect to MongoDB.
3. Use the split_hex_range function to calculate the split points.
4. Use the perform_splitting function to execute sharding operations.

from pymongo import MongoClient
from MongoChunkSplitter import split_hex_range, perform_splitting

# MongoDB connection setup
mongo_uri = "your_mongodb_uri"
client = MongoClient(mongo_uri)

# Database and collection names
database_name = "your_database_name"
collection_name = "your_collection_name"

# Shard key field
shard_key_field = "userId"  # or "userId,relevantUserId" for compound keys

# Calculate split points
data_size_mb = 1000  # Example data size in MB
chunk_size_mb = 64   # Example chunk size in MB
char_set = '0123456789abcdef'  # Example character set for hexadecimal
split_points = split_hex_range(char_set, data_size_mb, chunk_size_mb)

# Perform splitting
perform_splitting(client, database_name, collection_name, shard_key_field, split_points)

# Close the client connection
client.close()

Through Command Line

(Planned, not yet implemented)

Contributing

Contributions in the form of Pull Requests or Issues on GitHub are welcome.

License

This project is licensed under the MIT License.