Slow SQL Query with Sub Query with Order By

Question

Here's the query:

SELECT product.catalog_product_id AS catalog_id, listing.id AS listing_id, 
    product.size AS size, 0 AS amount, 
    listing.list_price AS price, 
    listing.created_at AS created_date 
FROM product
INNER JOIN listing ON listing.product_id = product.id 
WHERE product.catalog_product_id = XXXX
    AND listing.id = (
        SELECT l.id
        FROM listing l
        INNER JOIN product i ON l.product_id = i.id
        WHERE i.size = product.size AND i.catalog_product_id = XXXX
        ORDER BY l.list_price ASC, l.created_at ASC
        LIMIT 1
    ) 

Sub Query is this get listing ID for lowest price. Sub Query is slowing down with Order by. I have already created indexes and it still takes 5-6 seconds.

Table Structure:

catalog: This is master products catalog table

Catalog
-------
id
sku
name
description

products : Stores product variant(size options)

Products
--------
id
catalog_id
size 

listing table:: stores product listing, One Product size option can have multiple product listings with different prices.

listing
---------
id
product_id
list_price
created_at

Output: http://prntscr.com/kqh7fg

It shows lowest price for each listing and size.

Answer 1

I was able to sort out this.

SELECT product.catalog_product_id AS catalog_id, listing.id AS listing_id, 
product.size AS size,
listing.list_price AS price, 
listing.created_at AS created_date
FROM product
INNER JOIN listing ON listing.product_id = product.id
AND product.catalog_product_id = XXXX 
AND listing.id IN 
(
SELECT MIN(l.id)
FROM listing l
INNER JOIN product i ON l.product_id = i.id
WHERE i.catalog_product_id = XXXX 
AND list_price = (
SELECT MIN(list_price)
FROM listing l
INNER JOIN product it ON l.product_id = it.id
WHERE it.catalog_product_id = XXXX AND it.size = i.size
)
GROUP BY i.size
)

Instead of using:

ORDER BY l.list_price ASC, l.created_at ASC  LIMIT 1

I have used Min(Price) and Group By Size to get min price listing.

Query Performance improved from 5 sec to .5 seconds

Thanks!