0Pricing
Neo4j Graph Database Fundamentals · レッスン

トランザクションとセッションの処理

堅牢で信頼性の高いデータベース操作を実現するため、アプリケーションコードでトランザクションとセッションを管理する方法を理解します。

「トランザクションとセッションの処理」はCoddyKit上の無料Neo4j Graph Database Fundamentalsレッスンです。 これはレッスン3/4です。 下記で完全なレッスンを無料で読むことができます。その後、ブラウザ内の組み込みコードエディタと24時間対応のAIチューターでハンズオン演習できます。 これはNeo4j Graph Database Fundamentals学習パスの一部であり、ウェブとCoddyKitアプリ全体で進捗が同期されます。 Neo4j Graph Database Fundamentalsコースには全4レッスンが含まれています。

このレッスンの一部はまだ翻訳されておらず、英語で表示されています。

Why Transactions Matter

When you update data in a database, especially critical information, you want to ensure the operation is reliable. This is where transactions come in.

Transactions help maintain data integrity and consistency, even if something goes wrong during an operation.

Understanding Transactions

A transaction is a single, logical unit of work. It's a sequence of operations performed as a single atomic operation.

Think of it like transferring money between bank accounts: either both the debit and credit happen, or neither does. You wouldn't want money to leave one account without arriving in another!

The ACID Test

Transactions are often described by their ACID properties:

  • Atomicity: All or nothing. Either the entire transaction succeeds, or none of it does.
  • Consistency: Ensures the database moves from one valid state to another.
  • Isolation: Concurrent transactions don't interfere with each other.
  • Durability: Once a transaction is committed, its changes are permanent.

Connecting with Sessions

In Neo4j, you interact with the database using sessions. A session manages the communication channel and provides methods to execute Cypher queries.

Sessions are lightweight and designed to be opened and closed frequently. They are your primary interface for running queries, including those within transactions.

Basic Session Read

For simple read operations, you often don't need explicit transaction management. The Neo4j driver handles this for you with a read transaction.

Here's how to fetch a node count using a session:

from neo4j import GraphDatabase

URI = "bolt://localhost:7687"
USERNAME = "neo4j"
PASSWORD = "password"

driver = GraphDatabase.driver(URI, auth=(USERNAME, PASSWORD))

def get_node_count(tx):
    result = tx.run("MATCH (n) RETURN count(n) AS count")
    return result.single()["count"]

with driver.session() as session:
    count = session.read_transaction(get_node_count)
    print(f"Total nodes in graph: {count}")

driver.close()

Explicit Write Transactions

For operations that modify the database (like CREATE, MERGE, SET, DELETE), it's best practice to use explicit write transactions. This ensures atomicity.

The write_transaction method automatically handles committing if successful, or rolling back if an error occurs.

from neo4j import GraphDatabase

URI = "bolt://localhost:7687"
USERNAME = "neo4j"
PASSWORD = "password"

driver = GraphDatabase.driver(URI, auth=(USERNAME, PASSWORD))

def create_person(tx, name):
    tx.run("CREATE (p:Person {name: $name})", name=name)
    print(f"Created Person: {name}")

with driver.session() as session:
    session.write_transaction(create_person, "Alice")
    session.write_transaction(create_person, "Bob")

driver.close()

Commit or Rollback?

When using explicit transactions, your changes are not permanently saved until you commit them. If an error occurs, the transaction is rolled back, undoing all changes.

The session.write_transaction() and session.read_transaction() methods handle this implicitly for you. If the function passed to them completes without error, it commits. If an exception is raised, it rolls back.

Robust Error Handling

It's crucial to handle errors within your transaction logic. Any unhandled exception will cause the transaction to roll back, preventing partial updates.

You can use standard Python try...except blocks within your transaction function to manage specific errors or perform cleanup.

from neo4j import GraphDatabase

URI = "bolt://localhost:7687"
USERNAME = "neo4j"
PASSWORD = "password"

driver = GraphDatabase.driver(URI, auth=(USERNAME, PASSWORD))

def create_unique_node(tx, label, name):
    try:
        # Attempt to create a node. For demo, we'll re-raise to show rollback
        tx.run(f"CREATE (n:{label} {{name: $name}})", name=name)
        print(f"Node created: {name}")
    except Exception as e:
        print(f"Error creating node {name}: {e}")
        # Re-raise to trigger rollback for the entire write_transaction
        raise 

with driver.session() as session:
    # First call will likely succeed (unless node exists)
    try:
        session.write_transaction(create_unique_node, "City", "London")
    except Exception:
        print("Transaction for 'London' rolled back due to error.")

    # Second call, intentionally designed to fail for demonstration
    # (e.g., if we had a unique constraint on City.name)
    try:
        # Forcing an error to demonstrate rollback
        def force_error(tx):
            tx.run("CREATE (x)")
            raise ValueError("Simulated error!")
        session.write_transaction(force_error)
    except Exception:
        print("Transaction rolled back due to simulated error.")

driver.close()

Pythonic Context Managers

For even finer-grained control, or when performing multiple operations within a single transaction, you can use the session as a context manager and explicitly manage the transaction object.

The session.begin_transaction() method returns a transaction object that also works as a context manager.

from neo4j import GraphDatabase

URI = "bolt://localhost:7687"
USERNAME = "neo4j"
PASSWORD = "password"

driver = GraphDatabase.driver(URI, auth=(USERNAME, PASSWORD))

with driver.session() as session:
    with session.begin_transaction() as tx:
        tx.run("CREATE (a:Item {id: 1, name: 'Laptop'})")
        tx.run("CREATE (b:Item {id: 2, name: 'Mouse'})")
        # If no error, transaction commits automatically here when 'with tx' block exits
        print("Two items created in one transaction.")

driver.close()

Transaction Challenge

Consider a scenario where you are updating two properties of a single node in Neo4j within a Python application. If the second update fails due to a network error, what should happen to the first update if it was part of the same explicit write transaction?

Recap & Next Steps

We've covered the critical role of transactions in maintaining data integrity and consistency in your Neo4j applications.

  • Transactions follow ACID properties.
  • You use sessions to interact with the database.
  • session.read_transaction() and session.write_transaction() provide convenient, atomic operations.
  • Errors within a transaction lead to an automatic rollback, ensuring no partial data is committed.
  • Context managers offer a Pythonic way to manage sessions and transactions.

Understanding transactions is key to building robust and reliable Neo4j applications!

よくある質問

「トランザクションとセッションの処理」レッスンは無料ですか？

はい。「トランザクションとセッションの処理」の完全なテキストはこのウェブで無料で読めます。インタラクティブに演習し（組み込みコードエディタと24時間対応のAIチューター）、Neo4j Graph Database Fundamentalsコースの残りをアンロックするには、CoddyKit PROにアップグレードしてください。 Neo4j Graph Database Fundamentalsコースには全4レッスンが含まれています。

「トランザクションとセッションの処理」で何を学びますか？

堅牢で信頼性の高いデータベース操作を実現するため、アプリケーションコードでトランザクションとセッションを管理する方法を理解します。 ブラウザで直接実行するハンズオンコードでNeo4j Graph Database Fundamentalsを演習し、24時間対応のAIチューターがレッスンを進める中での質問に答えます。

Neo4j Graph Database Fundamentalsを始めるのに経験は必要ですか？

事前経験は必要ありません。CoddyKitのNeo4j Graph Database Fundamentalsは初級者から上級者向けに構成されているため、ここから始めるか最初から始めて、自分のペースで進むことができます。 これはレッスン3/4です。

「トランザクションとセッションの処理」レッスンにはどのくらい時間がかかりますか？

ほとんどのCoddyKitレッスンは約5～10分かかります。各レッスンはコンパクトでインタラクティブなので、着実に進歩し、ウェブとアプリ全体で正確に前回の場所から再開できます。

このNeo4j Graph Database Fundamentalsレッスンでコードを書いて実行できますか？

はい。すべてのNeo4j Graph Database Fundamentalsレッスンに組み込みコードエディタが含まれているため、ブラウザでリアルコードを書いて実行し、即座のAIフィードバックを取得できます。ローカル設定は不要です。

このコースのすべてのレッスン

  1. Pythonドライバーで接続する
  2. プログラムによるCRUD操作
  3. トランザクションとセッションの処理
  4. コネクションプールとエラー処理
← Neo4j Graph Database Fundamentalsに戻る