VTS

Overview

VTS (short for Vector Transport Service) is an open-source tool for moving vectors and unstructured data. It is developed by Zilliz based on Apache Seatunnel.

Why do you need a vector and unstructured data moving tool?

1. Meeting the Growing Data Migration Needs: VTS evolves from our Milvus Migration Service, which has successfully helped over 100 organizations migrate data between Milvus clusters. User demands have grown to include migrations from different vector databases, traditional search engines like Elasticsearch and Solr, relational databases, data warehouses, document databases, and even S3 and data lakes to Milvus.
2. Supporting Real-time Data Streaming and Offline Import: As vector database capabilities expand, users require both real-time data streaming and offline batch import options.
3. Simplifying Unstructured Data Transformation: Unlike traditional ETL, transforming unstructured data requires AI and model capabilities. VTS, in conjunction with the Zilliz Cloud Pipelines, enables vector embedding, tagging, and complex transformations, significantly reducing data cleaning costs and operational complexity.
4. Ensuring End-to-End Data Quality: Data integration and synchronization processes are prone to data loss and inconsistencies. VTS addresses these critical data quality concerns with robust monitoring and alerting mechanisms.

Core Capabilities of VTS

Built on top of Apache Seatunnel, Vector-Transport-Service offers:

1. Rich, extensible connectors
2. Unified stream and batch processing for real-time synchronization and offline batch imports
3. Distributed snapshot support for data consistency
4. High performance, low latency, and scalability
5. Real-time monitoring and visual management

Additionally, Vector-Transport-Service introduces vector-specific capabilities such as multiple data source support, schema matching, and basic data validation.

Roadmap

Future roadmaps include incremental sync, combined one-time migration and change data capture, and more advanced data transformation capabilities.

To learn more details about VTS used in action, read our blog:

• Introducing Migration Services: Efficiently Move Unstructured Data Across Platforms.

Get Started

To get started with VTS, follow the QuickStart Guide.

QuickStart Guide

This guide will help you get started with how to use vts to transport vector data into milvus, currently, we support the following source connectors:

• milvus
• postgres vector
• elastic search
• pinecone
• qdrant
• tencent vectordb

Step 1: Download and Extract VTS

• Navigate to the VTS releases page: https://github.com/zilliztech/vts/releases
• Download the latest binary package
• Extract the package using:

tar -xzvf "vector-transport-service-bin-xxx.tar.gz"

Step 2: Configure the Migration

Create a configuration file named milvus_to_milvus.conf with the following structure:

env {
  parallelism = 1
  job.mode = "BATCH"
}

  source {
  Milvus {
  url="https://in01-***.aws-us-west-2.vectordb.zillizcloud.com:19530"
  token="***"
  database="default"
  collection="medium_articles"
  batch_size=100
  }
}

  sink {
  Milvus {
  url="https://in01-***.aws-us-west-2.vectordb.zillizcloud.com:19542"
  token="***"
  database="default"
  batch_size=10
  }
}

Configuration Notes:

Replace placeholder values (marked with your-*) with your actual credentials Adjust batch_size based on your data volume and system resources The parallelism value can be increased for better performance on larger datasets

Step 3: Run the Migration

Execute the migration using the SeaTunnel shell script:

Cluster Mode (Recommended)

• Start the SeaTunnel cluster:

mkdir -p ./logs
./bin/seatunnel-cluster.sh -d

• Submit the migration job:

./bin/seatunnel.sh --config ./milvus_to_milvus.conf

Local Mode

./bin/seatunnel.sh --config ./milvus_to_milvus.conf -m local

Monitoring and Troubleshooting

Check the console output for progress and any error messages

Next Steps

Verify data integrity in Milvus after migration. If Need help with specific configuration options or running into issues? Feel free to ask for clarification!

Docker Deployment

VTS can also be deployed using Docker. To deploy VTS using Docker, follow the instructions in the Docker Deployment Guide.

Tutorial

In addition to the quick start guide, vts has much more powerful features like

• lots of transformer to support TabelPathMapper, FieldMapper, Embedding etc.
• cluster mode ready for production use with restful api to manage the job
• docker deploy, etc.

For detailed information, please refer to Tutorial.md

Connectors

VTS supports a variety of connectors to move data between different systems.

Find Detailed documentation for each connector:

• Milvus
• Elasitc Search
• Pinecone
• Qdrant
• Postgres Vector
• Tencent VectorDB

Development

see Development.md for building from source code and running examples.

Support

If you require any assistance or have questions regarding VTS, please feel free to reach out to our support team: Email: [email protected]

About Apache Seatunnel

SeaTunnel is a next-generation, high-performance, distributed data integration tool, capable of synchronizing vast amounts of data daily. It's trusted by numerous companies for its efficiency and stability. It's released under Apache 2 License.

SeaTunnel is a top-level project of the Apache Software Foundation (ASF). For more information, visit the Apache Seatunnel website.