在where子句中使用多个条件的缺点

Question

As we all know that the where clause of a sql query executes before select clause. And we put some conditions in where clause to filter out the result according to our requirement.

While writing some queries I encountered a question in my mind that, is/are there any drawback(s) of using multiple conditions in the where clause and in what order they are applied to filter the result from the selected table.

For example: we have a table

    Building(name,height,owner,builder_name,age)

and we have a query:

    select * from Building
    where height between X and Y and age between A and B

Now, how this query will execute. And what about the order of the conditions ie

X<=height<=Y and A<=age<=B

Will it be something like, first the whole record will be searched for the height within the given range and then same thing will be done for age well. ???

Answer 1

The query planner will try to find the optimal way to search the table and test the WHERE clause. It will start by trying to use an index if possible, which will narrow down the rows that it needs to search. If there are multiple potential indexes, it will try to use the one that it estimates will narrow it down best.

Then it will scan all these rows, and test each of them against all of the remaining conditions. It should never need to make multiple passes over the entire table.

If you want to see how a particular query will be executed, use the EXPLAIN command.

Answer 2

the Database server has multiple options to solve that query. It will choose the option the server "thinks" is faster.

The options I see are:

• Scan the whole table and filter out rows that don't satisfy the where clause
• Seek a height range on an index on height column, then filter out rows using age between A and B predicate.
• Seek an age range on an index on age column, then filter out rows using height between X and Y predicate
• Seek both indexes, then perform an index intersection

The database server not always use an index that might be applicable, it considers some things before using it, such as:

• The index selectivity.
• The index coverage.

High selectivity indexes are more likely to be used. Covering indexes are likely used.