Write-through Cache Strategy
- Data is written to both Cache Server/CPU Cache and the Database/Main Memory at the same time
Important
This ensures consistency between the cache server and the database. However, it increases write latency because write operations are limited by the speed of main memory. This is a feasible solution if write operations are infrequent. Alternatively, a write buffer (as shown above) can improve write performance.
Write-back Cache Strategy
- Data is written to Cache Server/CPU Cache first, and write to Database/Main Memory only when cache is replaced or evicted
Important
This approach offers low write latency. However, it risks data loss in the event of system failures, as the cache server stores data in volatile main memory.
Boost efficiency
To reduce unnecessary write-backs, each cache block can be assigned a dirty bit. This bit is flipped from
0to1when the data within the cache block is modified. A write-back operation is only performed if the dirty bit is set when the cache block needs to be replaced.
Write-around Cache Strategy
- Data bypasses Cache Server/CPU Cache and goes directly into the Database/Main Memory
Important
This prevents cache flooding.
However, this may increases read latency for new data.
Read-through Cache Strategy
- After receiving a request, Server first checks if the Cache Server has the available response
- If it has, it sends data back to the Client
- If not, it queries the Database, stores the response in cache server, and sends it back to the client