A caching mechanism based on data freshness

Use of cache is an effective way to improve access to a databases. However, when data in the cache are not updated or invalidated when writing to databases, the possibility increases with time that data in the cache will differ from data in the database. Therefore, we propose to adjust the balance between data freshness and access performance by switching between the use and non-use of a cache based on the load on the databases or on access performance. We also suggest and test a way of implementing the switching.