Transactions for Database Management Systemss
Transactions group several operations into one unit that is atomic, consistent, isolated and durable. Concurrency control is the part that provides isolation when many transactions run together, using two-phase locking, timestamp ordering or multi-version concurrency control. It aims at serialisable behaviour while allowing as much parallelism as possible.
Stage: Transactions and Distribution · Part of the Database Management Systems roadmap · Last reviewed 19 November 2026
What to learn in Transactions
- Define ACID and explain which component implements each property
- Check whether a schedule is conflict serialisable using a precedence graph
- Describe strict two-phase locking and how deadlocks are detected
- Explain how MVCC lets readers avoid blocking writers
- Compare read committed, snapshot isolation and serialisable
- Recognise anomalies such as dirty reads and write skew
Transactions interview questions
What is the difference between two-phase locking and MVCC?
Two-phase locking makes transactions acquire locks, so readers and writers can block each other. MVCC keeps several versions of each row with timestamps, so readers see a consistent snapshot without blocking writers, though write conflicts must still be resolved.
What is write skew?
Write skew is an anomaly allowed under snapshot isolation where two transactions read overlapping data and write different rows, each valid alone but together breaking an invariant, such as two doctors both going off call. Serialisable isolation prevents it.
Where Transactions fits in the Database Management Systems roadmap
Transactions belongs to the Transactions and Distribution stage (advanced level). Correctness at scale.
Other units in this stage
Back to the full Database Management Systems roadmap