🗄️ AVL Storage

The Storage Fortress. The Object Engine.

AVL (fortress) + STORAGE (engine) = AVL Storage

Where objects find permanent refuge and engines deliver at speed

🇧🇷 Latency 3-8ms in Brazil | 💰 50% cheaper than S3 | 🔄 S3-compatible API

🎯 What is AVL Storage?

AVL Storage is the S3-compatible object storage for the AVL Cloud Platform - built as a fortress for your files and an engine for your data.

Like Arxis provides the mathematical citadel, AVL Storage provides the object citadel:

• 🏛️ Solid Storage: Durable, replicated, versioned object storage
• ⚙️ Transfer Engine: High-speed uploads/downloads optimized for Brazil
• 🛡️ Protection: Encryption, immutability, lifecycle policies
• 🚀 Performance: 3-8ms latency in Brazil, multipart uploads

🌟 Key Features

🇧🇷 Optimized for Brazil & LATAM

• 3-8ms latency in São Paulo, Rio, Brasília
• Sub-30ms throughout Latin America
• 50% cheaper than AWS S3 for Brazilian traffic
• No egress fees within Brazil

📦 S3-Compatible API

• Drop-in replacement for AWS S3
• SDK compatibility: aws-sdk-rust, boto3, s3cmd
• Standard operations: PUT, GET, DELETE, LIST
• Multipart uploads for large files (>5 GB)

🗜️ Intelligent Compression

• Automatic compression via avila-compress
• LZ4 for hot data (fast access)
• Zstandard for cold data (high ratio)
• Transparent decompression on read

🔒 Security & Compliance

• Encryption at rest (AES-256)
• Encryption in transit (TLS 1.3)
• Versioning and immutability
• Access logs and audit trails

💰 Pricing (Brazil)

• R$ 0,15 per GB/month storage
• R$ 0,05 per GB transfer (within Brazil)
• FREE transfers between AVL services
• No minimum storage duration

🏗️ Architecture

AVL Storage follows the Arxis philosophy - solid as a fortress, fast as an engine:

┌─────────────────────────────────────────┐
│       AVL Storage - Object Citadel      │
├─────────────────────────────────────────┤
│  🏛️ Object Layer                        │
│     - S3-compatible API                  │
│     - Multipart uploads                  │
│     - Versioning & metadata             │
├─────────────────────────────────────────┤
│  ⚙️ Compression Engine                  │
│     - avila-compress (LZ4/Zstd)         │
│     - Content-type detection            │
│     - Smart tier selection              │
├─────────────────────────────────────────┤
│  🛡️ Storage Backend                     │
│     - Local filesystem (dev)            │
│     - Distributed storage (prod)        │
│     - Replication (3 copies)            │
├─────────────────────────────────────────┤
│  🚀 Transfer Engine                     │
│     - Parallel uploads/downloads        │
│     - Resumable transfers               │
│     - CDN integration                   │
└─────────────────────────────────────────┘

📦 Installation

Add to your Cargo.toml:

[dependencies]
avl-storage = "0.1"
tokio = { version = "1", features = ["full"] }

Using the CLI

# Install AVL CLI
curl -sSL https://avila.cloud/install.sh | sh

# Configure credentials
avl storage configure

# Upload file
avl storage put my-bucket/file.txt local-file.txt

# Download file
avl storage get my-bucket/file.txt downloaded.txt

# List objects
avl storage ls my-bucket/

S3 Compatibility

# Use with s3cmd
s3cmd --host=storage.avila.cloud --host-bucket='%(bucket)s.storage.avila.cloud' \
      put file.txt s3://my-bucket/

# Use with AWS CLI
aws s3 cp file.txt s3://my-bucket/ --endpoint-url=https://storage.avila.cloud

🚀 Quick Start

Basic Operations

use avl_storage::{StorageClient, PutObjectRequest};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Connect to AVL Storage
    let client = StorageClient::connect("https://storage.avila.cloud").await?;

    // Create bucket
    client.create_bucket("my-bucket").await?;

    // Upload object
    let data = b"Hello from AVL Storage!";
    client.put_object(PutObjectRequest {
        bucket: "my-bucket".to_string(),
        key: "hello.txt".to_string(),
        body: data.to_vec(),
        content_type: Some("text/plain".to_string()),
        metadata: Default::default(),
    }).await?;

    // Download object
    let obj = client.get_object("my-bucket", "hello.txt").await?;
    println!("Content: {}", String::from_utf8(obj.body)?);

    // List objects
    let objects = client.list_objects("my-bucket", None).await?;
    for obj in objects {
        println!("- {} ({} bytes)", obj.key, obj.size);
    }

    // Delete object
    client.delete_object("my-bucket", "hello.txt").await?;

    Ok(())
}

Multipart Uploads (Large Files)

use avl_storage::{StorageClient, MultipartUpload};
use tokio::fs::File;
use tokio::io::AsyncReadExt;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = StorageClient::connect("https://storage.avila.cloud").await?;

    // Initiate multipart upload
    let upload = client.create_multipart_upload(
        "my-bucket",
        "large-file.bin"
    ).await?;

    // Upload parts (5 MB chunks)
    let mut file = File::open("large-local-file.bin").await?;
    let chunk_size = 5 * 1024 * 1024; // 5 MB
    let mut part_number = 1;
    let mut parts = Vec::new();

    loop {
        let mut buffer = vec![0u8; chunk_size];
        let n = file.read(&mut buffer).await?;
        if n == 0 {
            break;
        }
        buffer.truncate(n);

        let etag = client.upload_part(
            "my-bucket",
            "large-file.bin",
            &upload.upload_id,
            part_number,
            buffer,
        ).await?;

        parts.push((part_number, etag));
        part_number += 1;
    }

    // Complete upload
    client.complete_multipart_upload(
        "my-bucket",
        "large-file.bin",
        &upload.upload_id,
        parts,
    ).await?;

    println!("Upload complete!");
    Ok(())
}

Automatic Compression

use avl_storage::{StorageClient, PutObjectRequest, StorageClass};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = StorageClient::connect("https://storage.avila.cloud").await?;

    // Hot data (frequent access) - LZ4 compression
    client.put_object(PutObjectRequest {
        bucket: "my-bucket".to_string(),
        key: "hot-data.json".to_string(),
        body: large_json_data,
        storage_class: Some(StorageClass::Standard), // LZ4
        ..Default::default()
    }).await?;

    // Cold data (archival) - Zstandard compression
    client.put_object(PutObjectRequest {
        bucket: "my-bucket".to_string(),
        key: "archive.tar".to_string(),
        body: archive_data,
        storage_class: Some(StorageClass::Archive), // Zstd
        ..Default::default()
    }).await?;

    // Transparent decompression on GET
    let obj = client.get_object("my-bucket", "hot-data.json").await?;
    // obj.body is automatically decompressed!

    Ok(())
}

📊 Use Cases

🎮 Game Assets & Media

• Game textures, models, sounds
• Video streaming (HLS/DASH)
• Player-generated content
• Patch distribution
• 3-8ms latency for Brazilian players

🤖 ML Model Storage

• Trained models and checkpoints
• Training datasets
• Inference artifacts
• Model versioning
• Automatic compression saves 50-70% space

📸 Image & Video Processing

• User uploads (photos, videos)
• Thumbnails and transcoded versions
• CDN origin storage
• Backup and archival
• Multipart uploads for large files

📊 Data Lake & Analytics

• Raw data ingestion
• Parquet/ORC files
• Log aggregation
• Time-series data
• S3-compatible with Spark/Presto

🔬 Scientific Data

• LIGO/LISA gravitational wave data
• Telescope observations
• Simulation results
• Dataset sharing
• High-throughput parallel uploads

🆚 Comparison with Competitors

AVL Storage is 50% cheaper for Brazilian workloads! 🇧🇷

🛠️ Best Practices

Naming Buckets

// ✅ GOOD: Descriptive, DNS-compatible
"my-app-uploads"
"prod-ml-models"
"user-avatars-2024"

// ❌ BAD: Ambiguous, special characters
"bucket1"
"my_bucket" // underscores not recommended
"UPPERCASE" // use lowercase

Object Keys

// ✅ GOOD: Hierarchical, organized
"users/user123/profile.jpg"
"models/v2/checkpoint-1000.pt"
"logs/2024/11/23/app.log"

// ❌ BAD: Flat, no structure
"file1.jpg"
"data.bin"

Compression

// ✅ GOOD: Let AVL Storage compress
client.put_object(PutObjectRequest {
    body: uncompressed_data,
    // AVL Storage compresses automatically
    ..Default::default()
}).await?;

// ❌ BAD: Pre-compress yourself
// let compressed = manual_compress(data); // Redundant!

Large Files

// ✅ GOOD: Use multipart for files > 100 MB
if file_size > 100 * 1024 * 1024 {
    upload_multipart(&client, bucket, key, file_path).await?;
} else {
    upload_single(&client, bucket, key, file_path).await?;
}

🧪 Development Tools

CLI

# Configure
avl storage configure
# Access Key: your-key
# Secret Key: your-secret
# Region: sa-east-1 (São Paulo)

# Create bucket
avl storage mb s3://my-bucket

# Upload
avl storage put my-bucket/file.txt local-file.txt

# Upload directory (recursive)
avl storage sync ./local-dir/ s3://my-bucket/remote-dir/

# Download
avl storage get my-bucket/file.txt downloaded.txt

# List
avl storage ls s3://my-bucket/

# Delete
avl storage rm s3://my-bucket/file.txt

# Get object metadata
avl storage info s3://my-bucket/file.txt

Local Emulator (Development)

# Run locally (no cloud costs!)
docker run -p 9000:9000 avilacloud/avl-storage-emulator:latest

# Update endpoint
export AVL_STORAGE_ENDPOINT=http://localhost:9000

📚 Documentation

• Official Docs - Complete guide
• API Reference - Rust API docs
• S3 Compatibility - S3 API mapping
• Examples - Code samples

🏛️ Philosophy - The Storage Citadel

AVL Storage embodies the Arxis philosophy:

🏛️ ARX - Fortress

• Durability: 11 nines (99.999999999%) with 3-copy replication
• Protection: Encryption, versioning, immutability
• Reliability: Geographic redundancy, automatic healing

⚙️ AXIS - Engine

• Speed: 3-8ms latency in Brazil, parallel transfers
• Efficiency: Automatic compression saves 50-70% space
• Scalability: Unlimited storage, elastic throughput

🇧🇷 Built for Brazil

• Local presence: Data centers in São Paulo, Rio
• Low latency: 3-8ms for Brazilian users
• Fair pricing: R$ instead of USD, 50% cheaper
• Portuguese-first: Docs, support, community

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/awesome-feature)
3. Commit your changes (git commit -m 'Add awesome feature')
4. Push to the branch (git push origin feature/awesome-feature)
5. Open a Pull Request

📞 Support

Email: nicolas@avila.inc WhatsApp: +55 17 99781-1471 GitHub: https://github.com/avilaops/arxis Docs: https://docs.avila.cloud/storage

📜 License

Dual-licensed under MIT OR Apache-2.0 - See LICENSE-MIT and LICENSE-APACHE for details.

🏛️ Built by Avila

AVL Storage - The Storage Fortress Part of the AVL Cloud Platform

🏛️ Durable as a fortress ⚙️ Fast as an engine 🇧🇷 Built for Brazil

Built with ❤️ in Rust for the Brazilian and LATAM tech community.