A
A
Alexander2020-06-01 07:34:06
SQL
Alexander, 2020-06-01 07:34:06

Is a UNIQUE INDEX lookup faster than a non-UNIQUE lookup?

One experienced person here told me that there is no difference. The understanding of the world is shaken.

I started looking and reading about it wherever possible. I found several articles explaining that there is a performance gain when searching for unique ones . But then I remembered that by index, the search usually takes place in a binary tree, and it has several values ​​\u200b\u200bthat lie in one node, which means the speed should not differ. So what is true and what is not?

I will not conduct tests, it is not possible to create conditions for adequate testing

Answer the question

In order to leave comments, you need to log in

3 answer(s)
S
Sergey Gornostaev, 2020-06-01
@AleksandrB

The central indicator of the effectiveness of the index is selectivity. For unique values, it is maximum.

D
Dr. Bacon, 2020-06-01
@bacon

UNIQUE is primarily about restrictions, it is strange to compare them in this way, if you do not have such a condition, then you will not be able to apply UNIQUE, and if there is, then why it was not immediately applied (well, except for rare reasons)

R
Roman Mirilaczvili, 2020-06-01
@2ord

If the index data is unique, then theoretically they are the same. In practice, this depends on optimizations in the DBMS.
UNIQUE INDEX guarantees that the data will be unique, while a simple index is not.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question