{{>toc}}

h1. Demo scenario

h2. E-commerce website

ECommerce website depending only on Redis to garanty performances for end-users experience.

h2. Product catalog 

Products are stored in MongoDB. Which is slow and expensive. 

h3. Read-through

* Add a connection to Mongo database
* Add a Read-Through rule with transformation to fetch products from Mongo to Redis on demand
* Add a long TTL to Redis stored products
* Add auto-refresh in Redis products on TTL expiration
* Add a Redis Query Engine index to search the products
* Add prefetch to prefetch the catalog in Redis at startup and make search more meaningful

h3. Write-through

When a change is made the in the online catalog, it needs to be immediately reflected in the backoffice.

h2. Price management

Prices are managed in a PostgreSQL database from a legacy system.

h3. Read-Through

* Add a PostgreSQL connection
* Add a Read-Through rule with transformation to fetch product prices
* Add auto-refresh in Redis stocks on read operation (first read is fast but maybe stale, second read is slower but accurate)
* Add prefetch to prefetch the stocks in Redis at startup

h2. Stocks management

Stocks are managed by a REST API microservice to automatically trigger alerts and restock

h3. Read-through

h3. Write-through

* Add a REST API webservice connector
* Add a write-through rule to update the stock management system synchronously