Today, on irc, someone asked interesting question.

Basically she ran a query like:

SELECT a, b, c, d, e, f FROM TABLE ORDER BY a

then, she processed the query to get, for each a array of unique values of b, c, d, e, and f, and then he inserted it back to database, to some other table.

It was a problem, because the table had many rows (millions I would assume), and the whole process was slow.

So, how to make it faster?

Continue reading Getting list of unique elements in table, per group