DB caching mechanism

QueryRowIndex

• No query criteria to Primary mapping cache
  • Goes to DB for row records via query criteria, then
    • write the cache of Primary to row records to redis
    • save the query condition to Primary mapping to redis and the framework's Take method does it automatically
  • possible expiration order
    • Query condition to Primary mapping cache not expired
      • Primary to row record cache not expired
        • Direct return of cached row records
      • Primary to row record cache is expired
        • Get row records from Primary to DB and write to cache
          • The problem at this point is that the cache of the query condition to Primary may be about to expire, and a query within a short time will trigger another database query
          • To avoid this problem, you can make the bold part above the first expiration time slightly longer than the second, for example, 5 seconds
    • The cache of the query condition to Primary mapping has expired, regardless of whether the cache of Primary to row records has expired
      • The query condition to Primary mapping is re-fetched, and the new Primary to row record cache is automatically written during the fetching process, so that the expiration time for both caches is just set
• Cache with query condition to Primary mapping
  • Cache without Primary to row record
    • Query to row records via Primary to DB and write to cache
  • Cache with Primary to row records
    • Return cached results directly