Home/ Questions/Q 7680481
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-31T18:07:26+00:00

I appreciate that LIKE queries are slow as they cannot be indexed. However, I

  • 0

I appreciate that LIKE queries are slow as they cannot be indexed. However, I am curious about the performance hit in a situation like this:

Say I have a table like:

user_id  |  message 
-------------------
   1     |  foo bar baz
   1     |  bar buz qux
   .     .      .
   .     .      .
   2     |  bux bar foo
   2     |  bar

where I have say 1 million rows, but 10,000 users, so each user has about 100 messages.

Clearly a search like:

SELECT * FROM table WHERE message like '%ar%';

is going to be very slow. However in my application I would only ever search a user’s messages:

SELECT * FROM table WHERE message like '%ar%' AND user_id = 2;

where the user_id column would be indexed.

Am I right in thinking that in a scenario like this, Postgres would only ever perform the slow LIKE query on the users ~100 rows, after using the indexed user_id column, rather than the full table – thus limiting my performance hit?

And also that a query like this wouldn’t get significantly slower with 10 or 100 million users, as long as any one user only had ~100 messages?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    The optimiser determines many things when compiling SQL into a plan.

    One of them is how to filter data (with index seeks, etc) before applying other conditions on a row by row basis.

    In your case, provided you have a suitable index the LIKE will only be applied to the records after that filtering is done.

    To understand a bit more about it, get the plan that is created by your query. You should be able to see where indexes are used to sub-set/filter the data, and then a separate step applying the LIKE condition.

    • 0