PrestaShop - problem with DB utilization

[Magicie]

First of all I would like to apologize, but I will use the Google Translator because I am kind of busy. :-/


A few days and unfortunately unsuccessfully solve the problem with all the server's CPU utilization is running DB and thus also the Apache proxy on the latest version of PrestaShop.

Have we already tried all settings in the administration and also as we have to do so will be also available on the server side, we managed to trace the internet and it looks like at this moment so that the only solution is coarser intervention into the code the application.

The PrestaShop that we have is about 25 to 30 thousand. Product and SQL query download them I find it all so maybe that number will show on the page, say the 20th, ie. Only then the PHP page.

If you click on the "+" at the categories and subcategories to this extract, which will successively select the desired one, so everything is alright and the individual products will be loaded. Just the category to "only" 2388 line.

In the 'details a lot of this query:

SELECT Mr id_product ``, `Condition` Mr., Mr. id_manufacturer ``, `quantity` Mr., Mr. `weight`,
(SELECT GROUP_CONCAT (id_category ``) FROM `ps_category_product WHERE` cp cp. id_product `= p.` `` id_product) and ids_cat,
(SELECT GROUP_CONCAT (id_feature_value ``) FROM `ps_feature_product WHERE` fp fp. id_product `= p.` `` id_product) and ids_feat,
(GROUP_CONCAT SELECT (DISTINCT (pac. id_attribute ``))
Ps_product_attribute_combination `FROM` Pac
LEFT JOIN `p` ON ps_product_attribute (Pa. id_product_attribute `=` pts. id_product_attribute ``)
WHERE AR. id_product `= p.` `` id_product) and ids_attr
FROM ps_product p
Mr. WHERE `active` = 1 AND `Mr. id_product` IN (SELECT FROM id_product ps_category_product `WHERE` cp cp. id_category `` = 678)

However, when you directly to the category with more product, perhaps the MySQL database does not manage to answer a whole, are finishing a 504 gateway timeout, eventually 502 Proxy error.

Not in some way more optimized queries, eventually adjust so that we managed to break in?