Data caching involves storing frequently accessed data in memory to improve query performance and reduce the need to retrieve data from disk.

Caching Strategies

Buffer Pool

• Data Pages:

  • Data is organized into fixed-size units known as data pages when read from or written to disk;
  • These data pages are temporarily stored in the DBMS's memory within the buffer pool, also referred to as the buffer cache.

• Cache Replacement Policies:

  • Due to limited space in the buffer pool, a cache replacement policy determines which data pages to retain in memory when new pages are loaded.

• Cache Management:

  • When a query requests data, the DBMS checks if the required data pages are already in the buffer pool;
  • If the data pages are found in memory (cache hit), they are retrieved from the buffer pool, bypassing slower disk access;
  • If the data pages are not in memory (cache miss), the DBMS fetches them from disk and loads them into the buffer pool for future access.

Query Result Cache

• Some DBMSs provide a query result cache, storing frequently executed query results in memory;
• When a query is executed, the system checks if the same query with identical parameters has been executed previously. If found, it returns the cached result rather than re-executing the query.

Index Caching

• Besides caching data pages, DBMSs may cache index pages in memory to accelerate index-based queries;
• Index caching reduces the need to traverse the entire index structure on disk, enhancing query performance for indexed columns.