OLTP
OLTP (Online Transaction Processing) refers to database systems and workloads optimized for large numbers of short, concurrent read/write transactions, such as processing an order or updating an account balance.
Definition
OLTP (Online Transaction Processing) refers to database systems and workloads optimized for large numbers of short, concurrent read/write transactions, such as processing an order or updating an account balance.
Overview
OLTP systems are the backbone of everyday application databases — every checkout on an e-commerce site, every bank transfer, every row inserted when a user signs up runs against an OLTP-optimized database. These workloads are characterized by many small, fast transactions touching a handful of rows each, high concurrency (thousands of users hitting the database simultaneously), and a strong need for ACID properties guarantees so that concurrent transactions don't corrupt or overwrite each other's data. OLTP schemas are typically highly normalized (see database normalization) to minimize redundancy and keep individual write operations fast and simple, since a normalized schema means updating a single fact requires touching only one row rather than many duplicated copies. Relational databases such as PostgreSQL, MySQL, and Microsoft SQL Server are the traditional home for OLTP workloads, relying heavily on database indexing to keep those small transactional queries fast even as tables grow. OLTP is directly contrasted with OLAP (Online Analytical Processing), which runs complex, read-heavy queries across huge volumes of historical data for reporting and analysis — a single OLAP query might scan millions of rows to compute an aggregate, which would be far too slow and disruptive to run against a live OLTP system serving production traffic. This is why organizations typically pipe OLTP data into a separate data warehouse or data lakehouse via ETL or ELT processes, keeping analytical workloads from competing with transactional ones for the same database resources. Understanding the OLTP/OLAP split is foundational for anyone designing a data architecture and is covered in data engineering-focused courses like Apache Airflow & Orchestration.
Key Concepts
- Optimized for many short, concurrent read/write transactions
- Strong reliance on ACID guarantees for correctness under concurrency
- Typically highly normalized schemas to keep writes fast and simple
- Heavy use of indexing to keep small transactional queries responsive
- Powers everyday application workloads: orders, signups, payments
- Contrasted with OLAP, which handles large-scale analytical queries
- Source data commonly piped into a data warehouse via ETL/ELT for analytics