Apache Ignite's distributed clustering model and data grids move data closer to the application and help us to achieve high performance by keeping objects in memory. However, we cannot keep the entire dataset in memory because the RAM is volatile in nature. If the entire cluster goes down, we will lose our precious data. Also, keeping the entire application data in memory can affect the system performance.
The optimal design is to keep the subset of data, the most recently used or active data, in memory and the other set (inactive or not so frequently used data) can be stored in a persistent data store. Apache Ignite supports the following two persistence modes:
- Native disk-based persistence
- Third-party data store persistence: RDBMS and NoSQL
In this section, we are going to examine the default native persistence, third-party MySQL...
