A new column in a database schema changes the shape of everything built on it. In SQL, adding a column means deciding type, constraints, nullability, and defaults before execution. It’s never just one command. Done wrong, it locks rows, blocks writes, or slows every query. Done right, it is invisible to the workflow and live to users in seconds.
In PostgreSQL, ALTER TABLE ADD COLUMN is straightforward for small datasets. For large tables, you plan for concurrency. You avoid full table rewrites by adding columns without defaults, then populating them in batches. You index last, not first, to prevent long locks.
In MySQL, adding a new column can trigger a table rebuild if your engine is not using instant DDL. With InnoDB in newer versions, ALTER TABLE ... ADD COLUMN with ALGORITHM=INSTANT can avoid downtime. You verify version, storage engine, and replication before you run it.
For distributed databases, schema changes need coordination. Adding a column hits every node. You track schema agreement and replicate the change in a safe order. In systems like Cassandra, new columns are schema changes in the system tables, but you still test read and write paths against it before rollout.