Question

I have a question on indexing a column and here it is

I have a table which have 3 columns of data type as below.

VARCHAR(255) --- Have many duplicates (say 10% unique records)
VARCHAR(64) --- Have many duplicates (say 10% unique records)
VARCHAR(MAX) --- Surely can’t choose this COL as key col for index

If I want to create a index on this table, How should I do it in this situation?

OR

Shall I create a auto increment column?

Can you please help me in this.

Thanks,
Rahul

Answer 1

You can create index on every column if you like, it doesn’t matter if they have unique data or not.

Indexes are used for faster retrieving info, searching and filtering on the indexed column.

An index can also be applied to a combination of columns (fields).

So, you can add an index on field1, another for field2, another for field3, another for (field1, field3) combination, etc. What is best, depends on what use you plan to do the table.

If, one the other hand, you want to create a Primary Key and not an index:

Choosing a column for primary key:

1. the column must uniquely identify every row (so, you can’t use field1 or field2 that have duplicates).

2. It’s better if it is as small as possible, so when used as foreign key in other tables, it doesn’t take much space (and for other reasons as well). SO, you better not use field3 – although you can do that in a test database or a small one or if you don’t care about performance.

So, yes, my advice would be to add another auto incremented column and use it as primary key.

---

As for unique indexes:

When creating a primary key, a unique index on the field is created as well.

You can also (or without a primary key) create a unique index on a field (or combination of). In your case where fields 1 and 2 have duplicates, you either make a unique index on field3 or add an auto incremented field and a unique index on that.