Developer forum

Dynamicweb will ask the price provider everytime it needs to calculate something - i.e. if the amount of a orderline changes, or new products are added to the cart, everything is recalculated, so you have to take performance issues into consideration.

With 15.000 users and I dont know how many products would give a hole lot of records in the database? How many are we talking about - is caching in i.e. application scope a possibility?

We are hosted on a custom server, so anything should be possible...
There'll be arround 600 products. And there'll be 2 groups (with 100 and 50 users) that will have unique prices for all products. 3 groups will have some unique prices - lets say 50-100 products.
To the above - some of the 100 users (10-20%) in the one group will have unique prices on some of the products (10-20%). These products prices will depend on the amount added to the cart.

We are planning to create a new table containing prices for groups only. The table will contain a row for each product in combination of a indexed cell for each groupid (int) + a indexed cell holding the productid (int) and of course the price it self.

For the useres prices we will create another table with a row for each price with a cell for price, userid, productid and amount.
We'll use stored procedures...

I dont know about caching - because a price could change every hour by import. Every price are validated on date/time before the are imported.

FYI. Potientially we have 150.000 users. Hopefully will have 10% or so converted to actual useres. Theses users will have unique prices - but only on group level.


/Morten