Question

I have a MySQL table with an auto-incremented integer primary key. I want to get a bunch of rows from the table based on an array of integers I have in memory in my program. The array ranges from a handful to about 1000 items.

What’s the most efficient query syntax to get the rows?

I can think of a few:

1. “SELECT * FROM thetable WHERE id IN (1, 2, 3, 4, 5)” (this is what I do now)
2. “SELECT * FROM thetable where id = 1 OR id = 2 OR id = 3”
3. Multiple queries of the form “SELECT * FROM thetable WHERE id = 1”. Probably the most friendly to the query cache, but expensive due to having lots of query parsing.
4. A union, like “SELECT * FROM thetable WHERE id = 1 UNION SELECT * FROM thetable WHERE id = 2 …” I’m not sure if MySQL caches the results of each query; it’s also the most verbose format.

I think using the NoSQL interface in MySQL 5.6+ would be the most efficient way to do this, but I’m not yet up to MySQL 5.6.

Answer 1

No.1 is the fastest and easiest to maintain. So you’re already where you want to be.