Talk Oracle Code One 2019
0 likes87 views
Java server-side technologies have evolved over 20 years from J2EE in the late 1990s and early 2000s to today's Jakarta EE and MicroProfile standards. Key developments included the introduction of EJB, JSP, Servlets, and later Java EE additions such as CDI, JAX-RS, and security improvements. Standards have faced criticism but also drive innovation while providing portability. Cloud-native applications now emphasize fast startup, observability, fault tolerance, and other features aligned with containers and microservices. Jakarta EE and MicroProfile aim to continue advancing Java for the cloud.
Talk Oracle Code One 2019
- 1. Java server-side 20 years old and still innovative! @jefrajames @sebi2706
- 2. J2EE Servlet JSP EJB 1-2 JMS 1999-2003 2006 2009 2013 2017 2019 Java EE 5 Ease of Dev EJB 3 JPA JSF JAX-WS Java EE 6 CDI JAX-RS Bean Validation Web Profile Java EE 7 JBatch Web Socket JSON-P Concurrency The origins Sun Enterprise times From J2EE to Jakarta EE A user experience Java EE 8 Java SE 8 JSON-B Security Jakarta EE Java EE 8 Equivalent GlassFish 5.1 Jakarta EE 8 Eclipse Cloud times Oracle Enterprise times
- 3. Current status Fragmented yet consistent and innovative MicroProfile 1.x Java EE 8 MicroProfile 2.x, 3.x Jakarta EE 8Java EE 7 Jakarta EE 9 MicroProfile 3.x Innovation Innovation Innovation Java SE 7,8 Java SE 9, 10, 11, 12, 13 … Amazon Coretto, Alibaba Dragonwell, AdoptOpenJdk, Azul HotSpot, OpenJ9, GraalVM Innovation
- 4. • Standards are slow – Broad consensus and getting things right takes time • Standard don’t guarantee portability – Still better than vendor lock-in • Standards don’t have feature XYZ – The core of an ecosystem can’t be bloated or complex • Standards don’t innovate – In fact they do and over-standardizing the unproven or niche case is a bad idea • It’s just a bench of vendor experts • You can on-board Standards?
- 5. Shifting to cloud-native applications 2006 2008 2009 2010 2011 2012 2013 2014 2015
- 6. Shifting to cloud-native applications To Private on premise data centers Cloud platforms Bare metal infrastructure Infrastructure as code Mutualized full-blown application servers Just enough runtime Monolith Microservices SQL-only Polyglot persistency ACID transactions Eventual consistency Synchronous & imperative programming Asynchronous & functional programming Standalone applications API-first platforms From
- 7. ● Runtime efficiency: fast startup, small system footprint, container awareness ● Built-in “Cloud-native” features: ○ Observability ○ Fault Tolerance ○ External configuration sources ○ Distributed security ○ Service mesh readiness ● Agile specification process Innovation needed!
- 8. MicroProfile support The application servers (r)evolution Java/Jakarta 8 EE support
- 9. Demo time! Simple Card Payment System Easy to Pay Check amount, POS, card Ask authorization If amount >100$ Fallback delegation if amount <=200$ Smart Bank Gateway Accept if amount<=500$ TomEE (Port 8082) Merchant Back Office Payment Request Payment Response Authorization Request Authorization Response
- 10. OpenJ9? Memory footprint: -30% Startup time (optional class sharing) : -15%
- 11. • Renew the Java EE “legacy” image • MicroProfile and Jakarta EE integration? • javax packages renaming: big bang or incremental? • Jakarta EE release cadence • Evolution of the ecosystem: IBM, Red Hat, Microsoft, Amazon, Alibaba ... • Effective contribution of the user community to the success of Jakarta EE Challenges
- 12. Conclusion
- 13. Thank you!