Database Management¶

Comprehensive guide to database lifecycle, configuration, and resource management in ArcadeDB Python bindings.

Overview¶

ArcadeDB databases are embedded, file-based databases stored on your local filesystem. The database lifecycle includes creation, opening, operations, and cleanup.

Database Modes:

Embedded: Database runs in your Python process
In-Process: Direct Java API access via JPype
File-Based: Data persisted to disk

Quick Start¶

Create New Database¶

import arcadedb_embedded as arcadedb

# Simple creation
with arcadedb.create_database("./mydb") as db:
    # Use the database
    db.command("sql", "CREATE VERTEX TYPE User")
    with db.transaction():
        db.command("sql", "INSERT INTO User SET name = ?", "Alice")

Open Existing Database¶

# Open existing database
with arcadedb.open_database("./mydb") as db:
    # Query data
    result = db.query("sql", "SELECT FROM User")
    for user in result:
        print(user.get("name"))

Check if Database Exists¶

if arcadedb.database_exists("./mydb"):
    with arcadedb.open_database("./mydb") as db:
        pass
else:
    with arcadedb.create_database("./mydb") as db:
        pass

Database Lifecycle¶

1. Creation¶

import arcadedb_embedded as arcadedb

# Create new database
with arcadedb.create_database("./mydb") as db:
    print(f"Database created at: {db.get_name()}")

What happens during creation:

Directory ./mydb created on filesystem
Schema files initialized
System files created (dictionary, configuration)
Database opened in READ_WRITE mode
Ready for operations

2. Opening¶

# Open existing database
with arcadedb.open_database("./mydb") as db:
    print(f"Database opened: {db.get_name()}")

3. Using¶

with arcadedb.open_database("./mydb") as db:
    # Create schema
    db.command("sql", "CREATE VERTEX TYPE Person")
    db.command("sql", "CREATE PROPERTY Person.name STRING")

    # Insert data
    with db.transaction():
        db.command("sql", "INSERT INTO Person SET name = ?", "Bob")

    # Query data
    result = db.query("sql", "SELECT FROM Person")
    for person in result:
        print(person.get("name"))

4. Closing¶

# Explicit close
db.close()

# Or use context manager (recommended)
with arcadedb.open_database("./mydb") as db:
    # Use database
    pass
# Automatically closed

What happens during close:

Active transactions rolled back
Buffers flushed to disk
Files closed
Resources released
Database removed from active instances

5. Dropping¶

with arcadedb.open_database("./mydb") as db:
    # Drop database (deletes all files)
    db.drop()

# Database and all files are permanently deleted

⚠️ Warning: drop() is irreversible and deletes all data!

DatabaseFactory Class¶

For advanced use cases, use DatabaseFactory directly:

import arcadedb_embedded as arcadedb

# Create factory
factory = arcadedb.DatabaseFactory("./mydb")

# Open or create via factory
if factory.exists():
    print("Database exists")
    with factory.open() as db:
        with db.transaction():
            # Operations
            pass
else:
    print("Creating new database")
    with factory.create() as db:
        with db.transaction():
            # Operations
            pass

Factory Pattern Benefits¶

Explicit control: Clear separation of creation/opening logic
Reusability: One factory for multiple operations
Configuration: Set options before creating/opening

Context Managers¶

Database Context Manager¶

# Automatic resource cleanup
with arcadedb.open_database("./mydb") as db:
    # Use database
    result = db.query("sql", "SELECT FROM User")
    for row in result:
        print(row.get("name"))
# Database automatically closed on exit

Benefits:

✅ Automatic close on normal exit
✅ Automatic close on exception
✅ Guaranteed resource cleanup
✅ Pythonic style

Nested Context Managers¶

# Open database and use transaction
with arcadedb.open_database("./mydb") as db:
    with db.transaction():
        db.command("sql", "INSERT INTO User SET name = ?", "Charlie")
# Transaction committed, database closed

Configuration¶

Database Directory Structure¶

./mydb/
├── configuration.json      # Database configuration
├── schema.json             # Schema definition
├── schema.prev.json        # Schema backup
├── dictionary.*.dict       # String dictionary
├── statistics.json         # Database statistics
├── User_0.*.bucket         # User type data files
├── HasFriend_0.*.bucket    # Edge type data files
└── .lock                   # Lock file (when open)

Database Location¶

import os

# Relative path
with arcadedb.create_database("./mydb"):
    pass

# Absolute path
with arcadedb.create_database("/var/data/mydb"):
    pass

Resource Management¶

Proper Cleanup¶

# ✓ Recommended: Context manager
with arcadedb.open_database("./mydb") as db:
    # Use database
    pass

# ✓ Acceptable: Explicit close
db = arcadedb.open_database("./mydb")
try:
    # Use database
    pass
finally:
    db.close()

# ✗ Bad: No cleanup
db = arcadedb.open_database("./mydb")
# Operations...
# Database never closed!

Multiple Databases¶

with arcadedb.open_database("./database1") as db1, \
     arcadedb.open_database("./database2") as db2:
    # Use both databases
    result1 = db1.query("sql", "SELECT FROM User")
    result2 = db2.query("sql", "SELECT FROM Product")

# Or with context managers
with arcadedb.open_database("./database1") as db1, \
     arcadedb.open_database("./database2") as db2:
    # Use both databases
    pass

Database Locking¶

Lock File:

Created when database opens: .lock
Prevents concurrent access from same/different processes
Automatically removed on clean close
Manual removal only if process crashed

# If database locked by another process
try:
    with arcadedb.open_database("./mydb"):
        pass
except Exception as e:
    print(f"Database locked: {e}")

    # Check if process is still running
    # If not, remove lock file
    import os
    lock_file = "./mydb/.lock"
    if os.path.exists(lock_file):
        os.remove(lock_file)

Common Patterns¶

Database Initialization¶

def init_database(path: str):
    """Initialize database with schema."""
    # Create if doesn't exist
    if not arcadedb.database_exists(path):
        with arcadedb.create_database(path) as db:
            db.command("sql", "CREATE VERTEX TYPE User")
            db.command("sql", "CREATE PROPERTY User.email STRING")
            db.command("sql", "CREATE INDEX ON User (email) UNIQUE")

            db.command("sql", "CREATE VERTEX TYPE Post")
            db.command("sql", "CREATE PROPERTY Post.title STRING")

            db.command("sql", "CREATE EDGE TYPE Authored UNIDIRECTIONAL")

            print(f"Database initialized at {path}")

    return arcadedb.open_database(path)

# Usage
with init_database("./myapp") as db:
    pass

Database Reset¶

def reset_database(path: str):
    """Drop and recreate database."""
    if arcadedb.database_exists(path):
        with arcadedb.open_database(path) as db:
            db.drop()
            print(f"Database dropped: {path}")

    with arcadedb.create_database(path) as db:
        print(f"Database created: {path}")

    return arcadedb.open_database(path)

# Usage
with reset_database("./mydb") as db:
    pass

Database Backup Pattern¶

import shutil
import datetime

def backup_database(db_path: str, backup_dir: str):
    """Backup database files."""
    # Close database first
    if arcadedb.database_exists(db_path):
        with arcadedb.open_database(db_path):
            pass  # Ensure clean state

    # Create backup with timestamp
    timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
    backup_path = f"{backup_dir}/backup_{timestamp}"

    # Copy database directory
    shutil.copytree(db_path, backup_path)
    print(f"Backup created: {backup_path}")

    return backup_path

# Usage
backup_path = backup_database("./mydb", "./backups")

Database Migration¶

def migrate_database(old_path: str, new_path: str):
    """Migrate data from old to new database."""
    with arcadedb.open_database(old_path) as old_db, \
         arcadedb.create_database(new_path) as new_db:
        # Copy schema
        new_db.command("sql", "CREATE VERTEX TYPE User")
        new_db.command("sql", "CREATE VERTEX TYPE Post")
        new_db.command("sql", "CREATE EDGE TYPE Authored UNIDIRECTIONAL")

        # Copy data
        batch_size = 1000

        # Migrate users
        result = old_db.query("sql", "SELECT FROM User")
        users_batch = []
        for user in result:
            users_batch.append(user)

            if len(users_batch) >= batch_size:
                with new_db.transaction():
                    for u in users_batch:
                        new_db.command(
                            "sql",
                            "INSERT INTO User SET name = ?, email = ?",
                            u.get("name"),
                            u.get("email"),
                        )
                users_batch = []

        # Flush remaining
        if users_batch:
            with new_db.transaction():
                for u in users_batch:
                    new_db.command(
                        "sql",
                        "INSERT INTO User SET name = ?, email = ?",
                        u.get("name"),
                        u.get("email"),
                    )

        print("Migration complete")

# Usage
migrate_database("./old_db", "./new_db")

Database Information¶

Check Transaction Status¶

with arcadedb.open_database("./mydb") as db:
    print(db.is_transaction_active())  # False

    with db.transaction():
        print(db.is_transaction_active())  # True

    print(db.is_transaction_active())  # False

Error Handling¶

Database Already Exists¶

try:
    with arcadedb.create_database("./mydb") as db:
        pass
except Exception as e:
    print(f"Error: {e}")
    # Database already exists
    with arcadedb.open_database("./mydb") as db:
        pass

Database Not Found¶

try:
    with arcadedb.open_database("./nonexistent") as db:
        pass
except Exception as e:
    print(f"Database not found: {e}")
    # Create it
    with arcadedb.create_database("./nonexistent") as db:
        pass

Database Locked¶

import time

def open_with_retry(path, max_retries=3):
    """Open database with retry logic."""
    for attempt in range(max_retries):
        try:
            return arcadedb.open_database(path)
        except Exception as e:
            if "locked" in str(e).lower():
                if attempt < max_retries - 1:
                    print(f"Database locked, retry {attempt + 1}/{max_retries}")
                    time.sleep(1)
                    continue
            raise
    raise Exception("Failed to open database after retries")

# Usage
with open_with_retry("./mydb") as db:
    pass

Advanced Topics¶

JVM Lifecycle and Databases¶

# JVM starts on first database/importer creation
import arcadedb_embedded as arcadedb

# Optional: configure JVM before first database/importer
arcadedb.start_jvm(heap_size="8g")

# JVM runs for entire Python process
with arcadedb.create_database("./db1") as db1:
    pass

with arcadedb.open_database("./db1") as db2:  # Same JVM
    pass

# JVM shuts down when Python exits

Multiple Databases, One JVM¶

# All databases share the same JVM
with arcadedb.open_database("./database1") as db1, \
     arcadedb.open_database("./database2") as db2, \
     arcadedb.open_database("./database3") as db3:
    # Efficient: shared JVM resources
    pass