A discussion community project. It not only includes registration, login, posting, commenting, liking, and replying functions, but also realizes sensitive word filtering using prefix tree, generates growth charts and pdfs using wkhtmltopdf, realizes website UV and DAU statistics, and stores user avatars and other information on the cloud server.
- Used Spring Security to implement permission control, instead of the interceptor, and used their own authentication scheme instead of the Security authentication process, which made permission authentication and control more convenient and flexible.
- Used Redis set for likes, zset for follows, and used Redis for storing login tickets and authentication codes to solve the distributed session problem.
- Used Redis advanced data type HyperLogLog to count UV (Unique Visitor) and Bitmap to count DAU (Daily Active User).
- Built a powerful asynchronous messaging system using Kafka to handle sending system notifications such as comments, likes and follows, and encapsulating them using events.
- Used Elasticsearch to implement the global search and add features such as keyword highlighting.
- Used distributed cache Redis and local cache Caffeine as multi-level cache to avoid cache avalanche for the hot post ranking module, which increases QPS by 20 times; used Quartz to update the hot post ranking regularly.
- Index
- Message
