smartbotic-database

Smartbotic Database

A standalone key-value document database with gRPC API, designed for microservice architectures.

Features

• Document Storage - JSON document store with collections
• Version History - Per-document version tracking with restore by version or date
• Health Check - gRPC HealthCheck() RPC for service status verification
• Replication - Multi-master replication via DatabaseReplication gRPC service
• Persistence - WAL (Write-Ahead Log) + snapshots with LZ4 compression
• Encryption - Field-level AES-256-GCM encryption for sensitive data
• Events - Pub/sub event subscription for real-time updates
• File Storage - Binary file management with streaming support
• Vector Storage - Embedding vector storage with SIMD-accelerated cosine similarity search
• Migrations - Declarative JSON-based schema migrations

Architecture

smartbotic-database/
├── client/           # Client library (injectable into projects)
├── service/          # Database service (standalone deployable)
├── proto/            # gRPC protocol buffer definitions
└── packaging/        # Docker build and systemd files

Component Separation

Component	Target	Description
Client Library	smartbotic-db-client	Lightweight gRPC client, injectable via submodule
Database Service	smartbotic-database	Full server with persistence, encryption, replication
Proto Library	smartbotic_db_proto	Generated protobuf/gRPC code

Building

Prerequisites

# Debian/Ubuntu
sudo apt install cmake ninja-build \
    protobuf-compiler libprotobuf-dev \
    libgrpc++-dev protobuf-compiler-grpc \
    nlohmann-json3-dev libspdlog-dev libssl-dev liblz4-dev

Standalone Build

cmake -B build -G Ninja
ninja -C build

# Run the database
./build/service/smartbotic-database --config config/database.json

As Embedded Submodule

When used as a git submodule in another project:

# Add submodule
git submodule add ssh://git@git.smartbotics.ai:10022/fszontagh/smartbotic-database.git external/smartbotic-database

Configure in your CMakeLists.txt:

# Build options
set(SMARTBOTIC_DB_BUILD_CLIENT ON CACHE BOOL "" FORCE)   # Always need client
set(SMARTBOTIC_DB_BUILD_SERVICE ON CACHE BOOL "" FORCE)  # Optional: include service

add_subdirectory(external/smartbotic-database)

# Link to client library
target_link_libraries(your-service PRIVATE smartbotic-db-client)

Client Library Usage

Basic Operations

#include <smartbotic/database/client.hpp>

// Connect to database
smartbotic::database::Client client("localhost:9004");

// Create a document
nlohmann::json user = {
    {"name", "John Doe"},
    {"email", "john@example.com"}
};
auto id = client.put("users", user);

// Get a document
auto doc = client.get("users", id);

// Query documents
smartbotic::database::QueryOptions opts;
opts.filter = {{"name", "John Doe"}};
auto results = client.query("users", opts);

// Delete a document
client.remove("users", id);

Working with Files

// Upload a file
std::vector<uint8_t> audioData = loadAudioFile("recording.wav");
auto fileId = client.uploadFile("call-recordings", "recording.wav", "audio/wav", audioData);

// Download a file
auto [data, metadata] = client.downloadFile(fileId);

// Delete a file
client.deleteFile(fileId);

Vector Storage & Similarity Search

Store embedding vectors alongside documents and perform cosine similarity search — no external dependencies required.

// Create a vector-enabled collection (dimension must be fixed at creation)
client.createCollection("memories", 0, false, 0, 384);  // 384-dim embeddings

// Insert documents with _vector field
nlohmann::json doc = {
    {"id", "mem-1"},
    {"content", "The user prefers dark mode"},
    {"_vector", embedding}  // std::vector<float> of size 384
};
client.put("memories", doc);

// Similarity search — returns top-K results sorted by cosine similarity
auto results = client.similaritySearch("memories", queryEmbedding, /*topK=*/5, /*minScore=*/0.7f);
for (const auto& r : results) {
    std::cout << r.id << " (score=" << r.score << "): "
              << r.data["content"] << std::endl;
}

Notes:

• _vector is extracted and stored separately — it is not returned in Get() or Query() responses
• Dimension is validated on every insert/update — mismatches are rejected
• Documents without _vector are stored normally but excluded from similarity search
• SIMD-accelerated (AVX2/SSE4.1) brute-force scan, suitable for up to ~100K vectors

Subscribing to Events

// Subscribe to collection events
client.subscribe("users", [](const smartbotic::database::Event& event) {
    if (event.type == smartbotic::database::EventType::Created) {
        std::cout << "New user created: " << event.documentId << std::endl;
    }
});

Version History

Every update and delete saves the previous document state to version history. You can browse history, retrieve old versions, and restore documents — including deleted ones.

// Get version history
auto history = client.getVersionHistory("users", docId);
for (const auto& ver : history.versions) {
    std::cout << "v" << ver.version << " at " << ver.timestamp
              << " by " << ver.updatedBy << std::endl;
}

// Get a specific old version
auto oldVer = client.getDocumentVersion("users", docId, 2);

// Restore to a specific version (creates new version with old data)
uint64_t newVer = client.restoreVersion("users", docId, 2, "admin");

// Restore to what the document looked like at a given time
auto [restoredFrom, newVersion] = client.restoreToDate("users", docId,
    1706000000000, "admin");  // timestamp in ms

// Restore a deleted document
client.remove("users", docId);
uint64_t restored = client.restoreVersion("users", docId, 1, "admin");

History depth is configurable per-collection via max_versions (0 = unlimited). History survives restarts through snapshots and WAL replay.

gRPC API

DatabaseService

RPC	Description
HealthCheck	Check service health and status
Get	Retrieve document by ID
Put	Create or update document
Delete	Remove document by ID
Query	Query documents with filters
Watch	Subscribe to document changes
ListCollections	List all collections
GetVersionHistory	List version history for a document
GetDocumentVersion	Get document data at a specific version
RestoreVersion	Restore document to a previous version (copy-forward)
RestoreToDate	Restore document to version active at a timestamp
UploadFile	Upload binary file (streaming)
DownloadFile	Download binary file (streaming)
DeleteFile	Delete file by ID
SimilaritySearch	Cosine similarity search over document vectors
GetFileMetadata	Get file metadata

DatabaseReplication

RPC	Description
Sync	Bidirectional sync stream
GetChanges	Get changes since sequence number
ApplyChanges	Apply changes from peer

Configuration

{
  "log_level": "info",
  "database": {
    "bind_address": "0.0.0.0",
    "rpc_port": 9004,
    "node_id": "db-node-1",
    "data_directory": "/var/lib/smartbotic/database",
    "migrations": {
      "enabled": true,
      "directory": "/etc/smartbotic/migrations",
      "auto_apply": true,
      "fail_on_error": true
    },
    "persistence": {
      "wal_sync_interval_ms": 100,
      "snapshot_interval_sec": 3600,
      "compression": "lz4"
    },
    "encryption": {
      "enabled": true,
      "key_file": "/var/lib/smartbotic/database/storage.key",
      "auto_generate_key": true
    },
    "files": {
      "max_file_size_mb": 500,
      "allowed_types": ["audio/*", "application/octet-stream"],
      "cleanup_orphans_interval_sec": 3600
    },
    "replication": {
      "enabled": false,
      "peers": ["db-node-2.example.com:9004"],
      "conflict_resolution": "last_writer_wins"
    }
  }
}

Configuration Reference

Section	Option	Default	Description
database	bind_address	0.0.0.0	Interface to bind gRPC server
database	rpc_port	9004	gRPC server port
database	node_id	db-1	Unique node identifier for replication
database	data_directory	/var/lib/smartbotic/database	Data storage path
migrations	enabled	true	Enable migration system
migrations	directory	-	Path to migration files
migrations	auto_apply	true	Auto-apply pending migrations on startup
migrations	fail_on_error	true	Stop if migration fails
persistence	wal_sync_interval_ms	100	WAL sync frequency
persistence	snapshot_interval_sec	3600	Snapshot creation interval
persistence	compression	lz4	Compression algorithm
encryption	enabled	false	Enable field-level encryption
encryption	key_file	-	Path to encryption key
encryption	auto_generate_key	false	Generate key if missing
replication	enabled	false	Enable multi-master replication
replication	peers	[]	List of peer node addresses
replication	conflict_resolution	last_writer_wins	Conflict resolution strategy

Migration System

Creating Migrations

Create JSON migration files in the migrations directory. Files are processed in alphabetical order (use numeric prefixes).

001_create_users.json:

{
  "version": "001",
  "name": "create_users",
  "description": "Create users collection with encryption",
  "operations": [
    {
      "type": "create_collection",
      "collection": "users",
      "options": {
        "encrypted": true,
        "sensitive_fields": ["password_hash", "api_key"],
        "max_versions": 50
      }
    }
  ]
}

002_create_memories.json:

{
  "version": "002",
  "name": "create_memories",
  "description": "Create vector-enabled memories collection",
  "operations": [
    {
      "type": "create_collection",
      "collection": "memories",
      "options": {
        "vector_dimension": 4096
      }
    }
  ]
}

003_seed_admin.json:

{
  "version": "003",
  "name": "seed_admin",
  "description": "Create default admin user",
  "operations": [
    {
      "type": "insert_if_not_exists",
      "collection": "users",
      "id": "admin",
      "data": {
        "username": "admin",
        "role": "administrator",
        "created_at": "{{NOW}}"
      }
    }
  ]
}

Migration Operations

Operation	Description	Required Fields
create_collection	Create collection if not exists	collection, options (optional)
drop_collection	Drop collection	collection, confirm: true
insert	Insert document (fails if exists)	collection, id, data
insert_if_not_exists	Insert only if ID doesn't exist	collection, id, data
upsert	Insert or update document	collection, id, data
update	Update existing document	collection, id, data
delete	Delete document	collection, id

Migration Tracking

Applied migrations are tracked in the _migrations collection:

{
  "id": "001_create_users",
  "version": "001",
  "name": "create_users",
  "applied_at": "2024-01-15T10:30:00Z",
  "status": "applied",
  "checksum": "sha256:abc123..."
}

Template Variables

Variable	Description
{{NOW}}	Current ISO 8601 timestamp
{{NODE_ID}}	Database node identifier
{{VERSION}}	Database version

Deployment Patterns

Pattern 1: Standalone Database

Deploy as a separate service:

┌──────────────────┐         ┌──────────────────┐
│   Your App       │──gRPC──▶│ smartbotic-db    │
│   (client lib)   │         │ (separate node)  │
└──────────────────┘         └──────────────────┘

# Build and package database separately
cd packaging
./build.sh
# Deploy smartbotic-database-x.y.z.tar.gz to database servers

Pattern 2: Bundled with Application

Include database binary in your application package:

# In your CMakeLists.txt
set(SMARTBOTIC_DB_BUILD_SERVICE ON)
add_subdirectory(external/smartbotic-database)

Both binaries deployed together:

/opt/myapp/
├── bin/
│   ├── myapp
│   └── smartbotic-database
└── etc/
    └── database.json

Pattern 3: High Availability

Multiple database nodes with replication:

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│  db-node-1  │◄───▶│  db-node-2  │◄───▶│  db-node-3  │
│  (primary)  │     │  (replica)  │     │  (replica)  │
└─────────────┘     └─────────────┘     └─────────────┘
       ▲                  ▲                   ▲
       │                  │                   │
       └──────────────────┴───────────────────┘
                    Load Balancer

Enable replication in config:

{
  "replication": {
    "enabled": true,
    "peers": ["db-node-2:9004", "db-node-3:9004"],
    "conflict_resolution": "last_writer_wins"
  }
}

Health Checking

# Using grpcurl
grpcurl -plaintext localhost:9004 smartbotic.databasepb.DatabaseService/HealthCheck

# Response
{
  "healthy": true,
  "nodeId": "db-node-1",
  "uptime": 86400,
  "documentsCount": 15234,
  "collectionsCount": 8,
  "diskUsageBytes": 52428800
}

Systemd Service

# /etc/systemd/system/smartbotic-database.service
[Unit]
Description=Smartbotic Database Service
After=network.target

[Service]
Type=simple
User=smartbotic
ExecStart=/opt/smartbotic/bin/smartbotic-database --config /etc/smartbotic/database.json
Restart=always
RestartSec=5

[Install]
WantedBy=multi-user.target

CMake Targets

Target	Description	Build Option
smartbotic_db_proto	Generated protobuf/gRPC code	Always
smartbotic-db-client	Client library	SMARTBOTIC_DB_BUILD_CLIENT=ON
smartbotic-database	Service executable	SMARTBOTIC_DB_BUILD_SERVICE=ON

Modifying from Consumer Project

When using as a submodule, you can modify the database code directly:

cd external/smartbotic-database
# Make changes
git add -A && git commit -m "Fix: description"
git push origin main

# Update submodule reference in parent project
cd ../..
git add external/smartbotic-database
git commit -m "Update smartbotic-database submodule"

License

Proprietary - Smartbotics AI