predastore

PredaStore

PredaStore is a lightweight, high-performance S3-compatible object storage server designed for on-premise, edge, and hybrid deployments. It implements the most commonly used S3 API operations, making it a practical alternative to Amazon S3 for local and private cloud environments.

PredaStore serves as the S3 storage backend for Hive, but can also be run as a standalone S3-compatible server.

Storage Backends

Filesystem (Default)

The default storage backend uses the local filesystem with a simple, efficient layout. This mode is production-ready and suitable for single-node deployments.

make build
./bin/s3d

Distributed (Under Development)

A distributed storage backend is currently under active development, featuring:

• Reed-Solomon Erasure Coding: Data redundancy and fault tolerance through configurable data/parity shards
• Consistent Hash Ring: Deterministic shard placement across nodes
• Multi-Node Architecture: Data distributed across multiple storage nodes via QUIC transport
• Automatic Reconstruction: Recovery of corrupted or missing shards from parity data

Note: The distributed backend is experimental and not yet recommended for production use.

S3 API Compatibility

PredaStore implements key S3 API operations compatible with AWS CLI, SDKs, and existing S3 tools:

Quick Start

Build

make build

Run

./bin/s3d

Configuration

Configure via environment variables or config.yaml:

server:
  host: "0.0.0.0"
  port: 8443
  tls:
    cert_file: "config/server.pem"
    key_file: "config/server.key"

storage:
  backend: "file"
  data_dir: "/var/lib/predastore"

auth:
  access_key: "your-access-key"
  secret_key: "your-secret-key"

Usage

AWS CLI Examples

# Create a bucket
aws --no-verify-ssl --endpoint-url https://localhost:8443/ s3 mb s3://my-bucket

# List buckets
aws --no-verify-ssl --endpoint-url https://localhost:8443/ s3 ls

# Upload a file
aws --no-verify-ssl --endpoint-url https://localhost:8443/ s3 cp ./file.txt s3://my-bucket/

# Download a file
aws --no-verify-ssl --endpoint-url https://localhost:8443/ s3 cp s3://my-bucket/file.txt ./downloaded.txt

# List bucket contents
aws --no-verify-ssl --endpoint-url https://localhost:8443/ s3 ls s3://my-bucket/

Integration with Hive

PredaStore is the default S3 storage provider for the Hive platform, providing object storage for:

• EC2 AMI images
• EBS volume snapshots
• User data and cloud-init configurations
• System artifacts

When running as part of Hive, PredaStore integrates via NATS messaging for S3 operations.

Viperblock

Viperblock is a high-performance block storage system that uses PredaStore as its S3-compatible backend for edge to on-premise deployments.

Key features:

• Log-structured writes for optimized performance
• Block-to-object mapping with configurable chunk sizes
• Checksumming and intelligent caching
• Designed for VM and container storage needs

Together, PredaStore and Viperblock provide the complete storage stack for Hive's EBS-compatible block storage service.

Development

make build          # Build s3d binary
make test           # Run tests
make bench          # Run benchmarks
make dev            # Run with hot reloading (requires air)
make clean          # Clean build artifacts

Docker

make docker_s3d             # Build Docker image
make docker_compose_up      # Start with docker-compose
make docker_compose_down    # Stop services

Performance Tuning

PredaStore uses QUIC for inter-node communication in distributed mode. For optimal performance, increase the system socket buffer sizes:

sudo sysctl -w net.core.rmem_max=7500000
sudo sysctl -w net.core.wmem_max=7500000

Use Cases

• Edge AI and IoT data collection
• Private cloud object storage
• Media and video archiving
• Backup and disaster recovery
• Development and testing (local S3 replacement)

Roadmap

• S3 API core support
• AWS Signature V4 authentication
• Multipart uploads
• Distributed storage with erasure coding
• Built-in IAM
• Real-time metrics and dashboard
• Bucket versioning
• Lifecycle policies

License

Apache 2.0 License. See LICENSE for details.

Contributing

Contributions are welcome. See CONTRIBUTING.md to get started.