Adding a new column in a live system is simple on paper but dangerous in practice. It changes schema. It alters queries. It can block writes if done poorly. Understanding the mechanics of ALTER TABLE is not enough—you need to know how storage engines apply the change, how indexes shift, and how locks behave under load.
In MySQL, ALTER TABLE ADD COLUMN can be online or blocking depending on the engine and version. In PostgreSQL, adding a nullable column without a default is fast, but adding one with a default rewrites the table. On massive datasets, that cost is real. Always measure the effect in a staging environment before committing to production.
Schema migration tools like Liquibase, Flyway, or native migration frameworks can batch column changes and run them in controlled steps. Use feature flags to gate logic until the new column is ready. Backfill data incrementally to keep write latencies stable.