Java Persistence: From DB over JDBC to Transactions
An airhacks.fm conversation with Vlad Mihalcea (@vlad_mihalcea) about: accessing database from Java, the four Java Database Connectivity (JDBC) driver levels, JDBC-ODBC bridge, native JDBC driver via Java Native Interface (JNI), JDBC middleware driver, the JDBC thin driver, the CloudscapeDB, the JDBC Driver initialization sequence, physical va. logical connections, connection pools and HikariCP, p6spy - the JDBC pool decorator, JDBC made databases more portable, evaluating project's age by the version of JDBC driven in Apache Maven's POM, statement caching, execution plan reuse, table scanning and index, execution plan is binding parameter dependent, PreparedStatements are not always preparing the statement, keeping connection timeout short, the JDBC "isValid" method, client side metadata caching, JDBC SQL statement compression, JDBC network data compression with Oracle, and MySQL, SSL encryption and fault tolerance, JDBC and transaction routing, primary and secondary node selection on the JDBC level, MySQL fault tolerance and fail over: Java EE, Jakarta EE, Helidon and quarkus programming models, GraalVM vs. Java / JVM mode, the Hibernate Types project, dependency reduction and ThinWARs, backward compatibility on NeXT generation runtimes, JDBC and auto-commit on, JDBC and Isolation Levels, disabling transactions and auto-commit mode, BASE vs. ACID, the trend to more correctness and consistency, SAGAs, compensative transactions, and the MiB flashlight, strict serialiazablity, read committed isolation level and data drift, isolation levels anomalies, serializable vs. snapshot isolation, the "High-Performance Java Persistence" book, the Hypersistence Optimizer, flexy pool, Vlad Mihalcea on twitter: @vlad_mihalcea