Question

Need some advice on how to optimize my articles table for read operations. I have articles table where I store articles editors write. There is requirement that editors can enter an article with a date_publish set in future. These articles can not be displayed in cover page at any time until the publish_date has actually come.

So my question here is should I have an index on date_publish field for better performance? I am using MySQL database, with InnoDB engine. I store dates as unixtimestamps in unsigned INT(11) field.

I when I make a read for list articles for cover page I do something like this:

SELECT articles.* FROM articles WHERE date_publish < $time

Answer 1

Adding an index on the column date_publish would optimize the following simple query:

SELECT * FROM articles WHERE date_publish < $time

However, if you change the query, such as add an ORDER clause to order by a column other than date_publish, you may need a compound (multi-column) index to optimize the query.

EDIT

To be able to fully utilize an index, a “covering” index must include all columns in the WHERE, JOIN and ORDER clauses, usually in that order. So, if you have a range in your WHERE clause on date_publish, and ORDER BY article_name, then you may wish to index on both columns (date_publish, article_name). That way MySQL can use the index for both selection and sorting.