[Doc] Add h2 dialect usage documentation

StreamPark supports H2 as a dialect for the convenience of users to experience quickly, but doesn't persist data; the data is stored in memory, so once the platform is restarted, the data is lost. H2 itself supports data persistence, so we need to support H2 data persistence. In this [PR](https://github.com/apache/incubator-streampark/pull/3759) is already supported, We need to complete the update of this part of the document before the release of 2.1.5