Key Concepts & Self-Assessment20 Key Facts
Review key ACID Properties: Transaction Reliability & Data Integrity exam facts and rate your mastery to track revision.
Progress: 0/20 Rated 0 Mastered 0 Review Later
#1
ACID is an acronym representing Atomicity, Consistency, Isolation, and Durability, defining the standard integrity requirements for database transactions.
#2
A database transaction represents an atomic logical unit of work comprising one or more read, write, update, or delete operations.
#3
The ANSI/ISO SQL standard officially codifies four transaction isolation levels: Read Uncommitted, Read Committed, Repeatable Read, and Serializable.
#4
Database consistency requires all schema rules, foreign key constraints, column data types, and check triggers to remain satisfied before and after a transaction.
#5
Computer scientist Jim Gray formalized the core conceptual model of transaction processing, atomicity, and crash recovery in the late 1970s.
#6
Andreas Reuter and Theo Härder coined the specific acronym ACID in their seminal 1983 computer science research paper on database recovery.
#7
C. Mohan and his IBM research team developed the Algorithms for Recovery and Isolation Exploiting Semantics (ARIES) protocol in 1992, establishing modern write-ahead logging.
#8
The ANSI SQL standard adopted formalized definitions of concurrency phenomena and transaction isolation tiers in 1992.
#9
Write-ahead logging (WAL) mandates that log records describing data mutations must be written to stable storage before dirty cache pages are flushed to disk.
#10
The transaction manager coordinates transaction states, transitioning operations through active, partially committed, committed, failed, and aborted statuses.
#11
Two-phase locking (2PL) protocols maintain a growing phase where locks are acquired and a shrinking phase where locks are released to ensure serializability.
#12
Multi-Version Concurrency Control (MVCC) enables concurrent read and write operations without mutual blocking by maintaining point-in-time snapshots of data records.
#13
Dirty reads occur when a transaction reads uncommitted modifications made by another concurrent transaction, permitted only at Read Uncommitted isolation.
#14
Non-repeatable reads happen when re-reading a row returns altered data because another concurrent transaction modified and committed the record.
#15
Phantom reads emerge when a query re-executes a range search and discovers newly inserted or deleted rows committed by a concurrent transaction.
#16
Database checkpoint intervals write dirty buffer pool pages to disk, reducing crash recovery time by bounding the necessary replay window of redo logs.
#17
PostgreSQL enforces ACID compliance through Multi-Version Concurrency Control, assigning monotonically increasing 32-bit transaction IDs (XIDs) to mutations.
#18
MySQL InnoDB storage engine maintains ACID reliability using its doublewrite buffer, undo tablespaces, and redone log files.
#19
Distributed systems often relax strict ACID guarantees in favor of the BASE model (Basically Available, Soft state, Eventual consistency) to scale horizontally across partitions.
#20
SQLite implements full ACID transactions locally using either a rollback journal file or a write-ahead log (WAL) file architecture.
Subject Specialist Commentary
Analytical perspective & practical exam advice from the Master10 academic board
Think of an ACID database transaction like a bank ATM withdrawal. When you withdraw cash, the machine must deduct money from your account and dispense the banknotes simultaneously. If the power cuts out halfway, atomicity ensures you do not lose money without getting the cash; it cancels the whole operation. Durability guarantees that once you get your receipt, the bank never forgets the transaction, even if its main servers restart immediately after.
In technical and public service examinations, questions frequently test transaction isolation levels and concurrency anomalies. Examiners often set traps confusing dirty reads with phantom reads. Remember that Repeatable Read prevents non-repeatable reads but might still permit phantom inserts unless range locks are applied. To memorize the four ANSI isolation levels in ascending strictness, use the mnemonic 'Uncommitted Can Read Serial,' moving from Read Uncommitted up to Serializable.
Related Knowledge Topics to Discover
Looking for more GK practice?
Explore 52,789+ questions across 65 General Knowledge categories.