* Introduce new h2Settings config block which overrides h2Port
* H2 server listens on localhost by default
* Change is backward compatible and old h2Port option can still be used but that always listens on localhost now
* Update changelog and docs with H2 changes
Enterprise tarball Corda distribution can only load JDBC jars from relative ./drivers directory.
The Capsule version can use directory specified by jarDirs config property and jarDirs was documented as a way to add JDBC drivers.
To make it consistent between tarball and Capsule version enterprise docs will be updated to use ./drivers dir, still preserving jarDirsbut state this option is not advisable and depending on the distribution may not work.
When specifying data source for NonStop database custom data source properties needs to be passed, which was always feasible but never documented (as there was no need for it).
ENT-1575 Add db migrations for db attachment changes,
ENT-1566 Fixes for db integration tests -following changes in CORDA-926 (obligatory node properties), upgrade db schema setup in draining mode/RPC tests, set runMigration=true in properties files for db tests, add new tables for deletion in SQL setup scrips
Updated db integration tests description in testing.rst
* Sql setup scripts for Oracle and datasource configuration.
* CashSelection for Oracle.
* Workaround of forbidden distinct for BLOBs.
* ojdbc8.jar driver dependency from Oracle Maven repository, ojdbc6.jar from lib (it's not in Maven repo).
* allow to use random Port in node names and trim it from schema name, remove schema with port numbers from db setup (to cater for new test, non Oracle specific change)
* Removed unnecessary code from ConfigUtilities (non Oracle specific change)
* Removed db integration tests setup for RPCStabilityTest as it doesn't start any nodes
* Change JDBC drivers dependencies to integrationTestRuntimeOnly
* Added smoke test validating JDBC driver registration located in "./plugins" directory
* Some docsite tweaks
* For Postgres, if a schema name is provided then it's wrapped in double quotes before passing to Hibernate to preserve case-sensitivity, Liquibase does it out-of-box.
* Sql setup scripts for PostgreSQL and datasource configuration.
* Improved SQL scripts for SQL Server and Azure to drop user/permissions on class setup not on test setup
* Set Micorsoft JDBC driver as compile time dependency.
* Database testing description.
* New table node_mutual_exclusion added to SQL test setup scripts.
* Coin selection for SQL Server/Azure
* Transaction isolation level change for updateInfoDB in PersistentNetworkMapCache from "repeatableRead" to "readCommitted"
* Configuration option to prefix all Hibernate generated SQL with a schema name via a configuration property `database.schema`
* Fix negative value in SELECT TOP query (fix for Oracle db)
