1 Million products!
-
Hi, guys!
Trying to import 1M products, and have got some speed degradation (metrics on 100 and 200 000 products).
- in "categories" section: infinite loading, blank screen, 100% cpu, only fix is rebooting
- "product form" page: loading time from less then 1s goes to 3-5s
- importing, add 1 product: 3x time to add
- admin panel, products: loading time from 1s goes to 7s
Could you help me please,
- how can i get a query monitoring for optimization?
- is there the best product fields organization? all fields to flat table or simpler?
thank you
-
Forget to ask, will it work on 6M products? Do you try it already?
-
There is the query that executes 2 seconds on opening "product form":
select * from
product_flat
whereurl_key
= '7637325' andlocale
= 'en' andchannel
= 'default'There are 330 000 products already.
-
just added indexes to fields 'url_key', 'locale', 'channel',
now query lasts only 0.0053 sec!
Note that to my 300mb base added 30mb of indexes. -
In admin panel, products section page (/admin/catalog/products)
- 5 seconds! /packages/Webkul/Ui/src/DataGrid/DataGrid.php:262
select count(*) as aggregate fromproduct_flat
left joinproducts
onproduct_flat
.product_id
=products
.id
left joinattribute_families
onproducts
.attribute_family_id
=attribute_families
.id
left joinproduct_inventories
onproduct_flat
.product_id
=product_inventories
.product_id
wherechannel
= 'default' andlocale
= 'en' group byproduct_flat
.product_id
plus
- 7 seconds! /packages/Webkul/Ui/src/DataGrid/DataGrid.php:262
selectproduct_flat
.product_id
asproduct_id
,product_flat
.sku
asproduct_sku
,product_flat
.name
asproduct_name
,products
.type
asproduct_type
,product_flat
.status
,product_flat
.price
,attribute_families
.name
asattribute_family
, SUM(product_inventories.qty) as quantity fromproduct_flat
left joinproducts
onproduct_flat
.product_id
=products
.id
left joinattribute_families
onproducts
.attribute_family_id
=attribute_families
.id
left joinproduct_inventories
onproduct_flat
.product_id
=product_inventories
.product_id
wherechannel
= 'default' andlocale
= 'en' group byproduct_flat
.product_id
order byproduct_id
desc limit 20 offset 0
please, help me to optimize this for big catalog.
- 5 seconds! /packages/Webkul/Ui/src/DataGrid/DataGrid.php:262
-
In file /packages/Webkul/Ui/src/DataGrid/DataGrid.php:
abstract class DataGrid -> public function getCollection()
this function is slow on big amount of products
$this->collection = $this->queryBuilder->orderBy($this->index, $this->sortOrder)->paginate($this->itemsPerPage)->appends(request()->except('page'))
it's generate aggregation query:
select count(*) as aggregate from product_flat left join products on product_flat.product_id = products.id left join attribute_families on products.attribute_family_id = attribute_families.id left join product_inventories on product_flat.product_id = product_inventories.product_id where channel = 'default' and locale = 'en' group by product_flat.product_id
on my 700 000 products it takes 8sec.
please, help to get faster this query on big catalog!Second query become faster than comment code in public function prepareQueryBuilder() in class ProductDataGrid extends DataGrid of ProductDataGrid.php:
comment ->leftJoin('attribute_families'...
comment ->leftJoin('product_inventories'...edit ->select('product_flat.product_id as product_id'... to
->select('product_flat.product_id as product_id', 'product_flat.sku as product_sku', 'product_flat.name as product_name', 'product_flat.status', 'product_flat.price')
-
Hello @george
We gonna apply check on itemPerPage, which can not be more than 100, because due to this it is taking more time.
Thanks
-
-
Hello, Rahul, it's not helped.
Problems on big catalog (over 300 000 products):- "main page" loading about 10sec.
- "/categories/xxxx" not working (blank screen, infinite loading)
- admin panel, products: loading 5-10 sec
- product form: 3-5sec.
Partially has fixed some code, but need some help.
Raw SQL queries with "default items per page" set to 1 or 99 is equally slow.
There are needed some optimization, adding indexes, simplify the queries, and etc.
Could we discuss this problem directly? Then you may add to your marketing phrase "big data" or something. -
Hi @george
Can you send me dump of your database so that we can check and optimise it, because it is a big concern.
Thanks
-
This post is deleted! -
Hi @george
I advice you to generate a ticket for same at https://bagisto.uvdesk.com/en/customer/create-ticket/
Thanks