I am working on an e-commerce application on which we are predicting around 2 millions products.
We are using Solr for searching, suggestions and filtering.
But for showing the product card data(which has detail like name, image, price, discount and rating) on the Product Listing page(PLP), we are going to Database after fetching the product ids from the Solr.
For E.g:
User went on the our e-commerce application and there on the search box typed "I phone 13 pro".
While user was typing we are showing suggestions to him/her after fetching it from the solr, no db involved yet.
After that user either click or hit enter then we are showing the Product listing page where we first reaching to the solr and fetching the product ids(say 24 per page we are showing) matching the entered text and then we are going to the DB with these ids and fetching the product objects and passing it to the UI with the necessary detail(using DTO object).
Now my question here is:
As I mentioned above we are hitting DB for the Product Listing Page(PLP) for fetching 24 product per page, here I am thinking it may slow down the page when our catalog and user started increasing?
Note:- we are also using Hibernate h2 level caching to stop hitting DB again and again for the same product but cache has limit up to 0.5 million.
Why we are hitting to DB instead of fetching everything from the Solr as we have all product data info which we wanted to show above on the PLP page?
Ans- As we are now keeping important fields of the product object in the Solr not all fields, So we are thinking may be in future we wanted to show something which we don't have in the Solr then we have to Re-index the full catalog to add new field into it and it may be very time-consuming.
I am in a favour of using Solr for showing PLP page instead of going to DB.Of course keeping Re-indexing in mind but not sure it is the best approach or not?
Please provide your suggestion or thoughts on this.
Thanks in advance.
Related
I have created a new category on my store after creating this website taking 15-20 seconds to load, but before that it's taking only 5-6 seconds
Please re-index all the index and if it is still slow you can try by using flat tables for catalog.
Also make sure you have not added anything wrong in category details like special characters in desc etc.
I have a client who is experiencing issues with their Wordpress/WooCommerce website.
Issue: The website is automatically adding products to the cart (Roughly 40-60). The issue occurs for both logged-in & logged-out users. The issue can occur when trying to login to the My Account section, when adding items to the cart or sometimes after adding items to the cart and then visiting the cart it will override cart items with new random items.
I have noticed that in WC_Session_handler the value for _customer_id is often not unique nor is the other session data.. I have removed all server & front-end caching, searched for any other sessions initialized.
Any help would be appreciated as they are losing business due to customers not being able to remove the items from their cart (As the removed items re-appear quickly)..
Domain Name: thecoffeehopper.com :)
The cart is constructed from data in the wp_woocommerce_sessions and wp_usermeta MySQL tables.
I would run the following sql queries to try and find the source of your problem.
select * from wp_woocommerce_sessions;
select * from wp_usermeta where meta_key like '_woocommerce_persistent_cart_%';
The data from these rows are used to construct the cart. Are the random items found in any of these rows? The rows have a user id or customer id to tell which customer the data belongs to.
The data in these rows are serialized strings and is difficult to read directly so I would use the WordPress CLI tool and apply the function maybe_unserialize() to the SQL results.
I work on an educational website in which we show dynamic filters. What does this mean? We now have several course categories that will increase in number in the following weeks.
Right now categories are a string field in each course. I'm planning to model this and create a Categories table. Having done this, it would be pretty easy to load the filters based on the categories we have on the database. However, I see the problem that each time the website is loaded, the query will be made.
I thought of caching these categories but I'm guessing is not the best solution.
Any idea on how can I avoid these queries each time but get this information from the database?
I am running Magento 1.6.2 and have an intermittent error cropping up with some sales orders. The customer is being charged correctly, but the Magento admin page when viewing the sale reports incorrect values. I understand why it is doing it incorrectly - I have a 10% discount off a full section of products on my website, and it is because of this that something is going wrong.
Again, the customer is paying the correct (discounted) amount, but the data must be being stored incorrectly in the magento database because looking at it in the admin system doesn't total the values correctly. The Catalog Price Rule that applies the 10% discount is not accounted for in the displayed totals.
To add to my woes, we use Zynk to bring data in from Magento to our Sage 200 backoffice system, and the data getting imported into that system is also missing the 10% discount, making reconciliation of order payments and Sage invoices a complete pain.
The only thing I can think is that somewhere somehow the data is being stored in the magento database incorrectly (not having had the 10% discount applied to it) so when Zynk collects the data, it is incorrect.
Can anyone help me understand where I can find the order data in the Magento database so I can see what was stored for a specific Magento order number - I can at least begin to debug this then, maybe?
If someone has details of where to find the code used by Magento to perform the order database transactions, specifically where the discounts should be applied before storing the data, that would be a big help too.
Cheers!
Orders data are saved in database tables sales_flat_order and sales_flat_order_item
Discounts are applied in Mage_SalesRule_Model_Validator
You can find orders and order items data in your (tables prefix)_sales_flat_order and (tables prefix)_sales_flat_order_item respectively. but remember that when an order is placed Magento simply copies the data from its respective quote. so in my opinion you should be checking (table prefix)_sales_flat_quote and (table prefix)_sales_flat_quote_item first.
Just wondering if what I want to do is possible with Solr:
I have a Solr index of 'Users' [id, name, join_date] and on my site and I have a page /users which has a Solr query that lists all of the users in the index in alphabetical order.
On my site users can 'follow' other users. I have added another page /followers which first gets all the user_ids which are 'following' the logged in user from MySQL - then uses these IDs to query Solr to only return users which the current user is following.
This all works fine.
But the problem is, I'd like to be able to order the users returned on /followers in order of when they first started following the logged in user - I keep this information in a join table in MySQL. But my Solr documents don't contain this date - as obviously one user can have many followers and vise versa. Also I can't just use MySQL to server this page as I need to keep it Solr query-able.
Would love some insight into this - it's been bugging me for a while now.
you can do that with ExternalField but you need to export your info from the db to a text file (and keep it updated etc).
This guy here discusses a way to do it without ExternalField, but requires customizing Sorl code.