Great question! This is a common point of confusion because the terms are often used interchangeably, but let’s break it down clearly:

In most mainstream database contexts (like MySQL, SQL Server, PostgreSQL), non-clustered indexes and secondary indexes are effectively the same concept. The difference is just in how we name them:

• "Non-clustered index" describes the index's relationship to the underlying data (it doesn't rearrange the physical storage order of the table's rows).
• "Secondary index" describes the index's role relative to the table's primary clustered index (it's an auxiliary index built on top of the primary clustered index, not the main sorted data structure).

Both types share these key traits:

• They don’t alter the physical sort order of the table’s data file.
• Each index entry points to the actual data record (the pointer might be a direct physical row address, or the value of the clustered index key—depending on the database engine).
• They’re used to speed up queries on columns that aren’t covered by the clustered index (like searching for users by email instead of user_id).

For example:

• In MySQL’s InnoDB engine, a secondary index’s leaf nodes store the primary key value of the matching row. To get the full row data, the engine uses that primary key to look up the row in the clustered index (the primary key index). This is exactly how non-clustered indexes work in SQL Server too.

In very niche or older database documentation/textbooks, you might see a tiny distinction:

• Some sources use "secondary index" as a broader umbrella term that includes all non-clustered index types (like covering indexes, filtered indexes, etc.), while "non-clustered index" refers specifically to a basic secondary index that doesn’t include extra columns beyond the indexed key and pointer.
• This is rare, though—you’ll almost never run into this in modern database work.

Don’t overcomplicate it: for 99% of your learning and day-to-day work, non-clustered indexes and secondary indexes are two names for the same thing. The only time you need to care about a difference is if a specific database’s documentation explicitly defines them separately.

内容的提问来源于stack exchange，提问作者user13242010