Manu's Vault

ACID

4 min readStatus:Seedling

Core Idea

ACID is a set of properties (Atomicity, Consistency, Isolation, Durability) that guarantee reliable processing of database transactions by ensuring all operations either complete fully or fail completely.

Definition

ACID (Atomicity, Consistency, Isolation, Durability) is a set of properties that guarantee reliable processing of database transactions. ACID ensures data validity and integrity despite errors, power failures, or other system mishaps by guaranteeing that all operations within a transaction either complete fully or fail completely, leaving the database in a valid state.

The acronym captures four fundamental guarantees: transactions are indivisible units (atomic), maintain database validity rules (consistent), execute independently without interference (isolated), and persist permanently once committed (durable).

Key Characteristics

  • Atomicity: Each transaction completes fully or not at all—operations cannot partially execute

    • If any step in a transaction fails, all changes revert to the previous state
    • Prevents data corruption from incomplete operations
    • Makes the series of operations “indivisible” (atomic)

  • Consistency: Transactions move the database from one valid state to another

    • Only valid data can be written according to defined rules and constraints
    • Prevents data corruption by rejecting invalid state transitions
    • Maintains referential integrity, foreign key constraints, and business rules

  • Isolation: Concurrent transactions execute as if they were sequential

    • Incomplete transaction effects are invisible to other transactions
    • Prevents race conditions and conflicts between simultaneous operations
    • Various isolation levels (Read Uncommitted, Read Committed, Repeatable Read, Serializable) trade performance for stronger guarantees

  • Durability: Committed transactions persist even after system failures

    • Completed transactions survive power outages, crashes, and hardware failures
    • Typically achieved by writing to non-volatile storage (disk, SSD)
    • Enables recovery and maintains data integrity across restarts

Why It Matters

ACID properties are foundational for systems requiring strong consistency and data integrity—particularly financial systems, inventory management, and transactional applications where partial failures are unacceptable. Relational databases (MySQL, PostgreSQL, Oracle, SQL Server) are ACID-compliant by default, making them the standard choice for applications that cannot tolerate data inconsistency.

However, ACID comes with trade-offs. In distributed systems, maintaining ACID guarantees across multiple nodes conflicts with the CAP Theorem—strong consistency reduces availability during network partitions. This tension led to BASE (Basically Available, Soft state, Eventual consistency) as an alternative model for distributed systems prioritizing availability and scalability over immediate consistency.

Understanding ACID is essential for architects choosing between consistency models. ACID databases use vertical scaling and strong consistency, while BASE systems (many NoSQL databases) favor horizontal scaling and eventual consistency. Modern distributed databases often offer configuration options for both consistency and availability, allowing context-specific trade-offs.

Examples

  • Banking transactions: Transferring money between accounts requires atomicity (debit and credit both succeed or both fail) and consistency (account balances remain valid)
  • E-commerce checkout: Order placement, inventory deduction, and payment processing must all complete together or all fail to prevent overselling
  • Relational databases: PostgreSQL, MySQL, Oracle, and SQL Server provide full ACID compliance through transaction logs and write-ahead logging
  • NewSQL databases: Systems like Google Spanner use worldwide private fiber networks and GPS clock synchronization to maintain ACID properties in distributed environments

Sources

Note

This content was drafted with assistance from AI tools for research, organization, and initial content generation. All final content has been reviewed, fact-checked, and edited by the author to ensure accuracy and alignment with the author’s intentions and perspective.

Graph View

Backlinks