Treasure Island!!!!

Guest stars: Elton John, Beck

Nice event!

Moving Oracle ADF to the Cloud: Development and deployment in the new age

Speaker: Shay Shmeltzer, Dan Goerdt

Reasons for moving to cloud:

• eliminate hw management
• eliminat sw management
• eliminate maintenance costs
• faster provisioning
• pay for what you use

Finally a picture of oracle cloud services

How to develop in the cloud? feature required

• Development infrastructure should provide
• Source control system/Version management
• Automated testing
• Continuous integration
• Deployment services
• Team infrastructure should provide
• task tracking
• documentation/wiki
• code review
• team management

All this features/requirements in one product: Oracle Developer Cloud Service

And more things are provided:

• weblogic, adf libraries, jdeveloper deployment profiles, interfaced with preferred IDE and browser
• complete platform with all together linked and working. Remembers to me Team Foundation Server from Microsoft.

Required tools:

• JDeveloper 12.1.3 and 11.1.1.* and DevCS (Oracle Developer Cloud Service)
• JDeveloper 12.2.1
• native integration with cloud
• interact with projects, tasks, build
• integrate with git

Demo time

• all revision/review workflow is provided by the platform
• all jira tracking too
• it is possible to work via browser or via JDeveloper as well

Deployment in the cloud

• Oracle Java Cloud Service : run time for the java
• supported by JDeveloper 12.1.3

Demo product of FlexDeploy product

• integrated with several kind of application
• could be a competitor for current deployment portal

Summary

• cloud is good
• use cloud for
• develop ADF application (developer cloud service)
• deploy and run ADF application (java cloud service)

Enterprise Mobile Persistence and Security using Oracle Mobile Application Framework

Speakers: Srini Indla, Denis Tyrell

MAF is a framework to develop once and run in multiple platform. Every application has a build jvm in it so the business logic is gone to be built in java. View is done in HTML5 pure application. Integration with features of the device is gone with cordova.

A lot of application have been built on MAF

MAF 2.2 release 2 weeks ago:

• many many new features to support latest mobile patters (swipe, animations, pull to refresh, ...)
• mobile alta ui 1.4 also released
• data binding improvements
• navigation improvements
• full support for android back button
• performance: more than 30% improvement
• jsonp improvements
• optimized jvm of 30/40%

MAF 2.3 features (2016)

• windows 10 support
• integratation easility with oracle mobile cloud service (same presentation of yesterday http://oowpost.blogspot.com/2015/10/the-emerging-enterprise-mobile-cloud.html)

Data persistence

• optimize application for offline
• follow some general principles
• caching of http layer for resource, will boost performances 
• fetch data from cache first if online
• Data types in JDev can be imported from Oracle Mobile Cloud from the REST API defined in that platform. there is a complete integration between the two products even if it is optional.

Security:

• 93% of data loss
• 2/3 time spent in mobile security projects
• 76% stored credential on the device
• 89% uses personal device at work (BYOD)

complexity

• no real existing standard available for multiple platform
• multiple platform to support

security in maf:

• SSL Secure channel
• Oauth 2
• SAML
• native library for ios and android which manages the security implementation packaged to the application
• any  sensitive information is stored in a native key store in the maf container application

EMM enterprise mobile management

ui experience

• sso support for multiple application in the device
• offline authentication 

authentication protocols supported

• basic auth
• offline authentication
• oauth 2
• web based sso

Federated authentication

• saml based sso authentication
• oauth 2
• secure token service (for upcoming release) 

Oracle BigData, Oracle Database In-Memory, Oracle Exadata: one database to rule them all

Speaker: Holger Friedrich

Database/Exadata

• 1977-2008: how db where working in that period
• 2008: exadata

• first version done with HP
• first full engineered system by oracle
• the recipe
• hardware engineering
• local query processing
• database aware of intelligent functional storage layer
• Features
• smart scanning
• flash cache
• storage indexes
• hybrid columnar compression and smart scanning on compressed data
• data mining tasks
• columnar flash cache

Big data

• biggest new area of oracle
• the Hadoop zoo
• a lot of tools around hadoop 
• tools obsolescence problem
• kafka replaced apache storm?
• select the right tool and the right provider!!!

Oracle big data appliance runs cloudera inside

• analytics
• tools from hadoop
• oracle big data analytics
• data integration
• tools from hadoop
• oracle big data sql 2.0
• smart scan features
• it is possible to query via sql hadoop cluster with same feature
• complete full set of features of sql language also with extension (see create table!!)
• oracle fusion middleware data integrator
• the recipe
• hardware engineering
• local query processing
• engine is aware of intelligent big data layer

Oracle In-Memory

• store data in columnar format
• persisted data still in row format
• all oracle business continuity invested continues to be used
• No application changes required
• Best for
• scan large quantities of data
• small set of columns
• compute aggregation on the results
• The recipe
• conceptual advantage of columnar storage
• speed of processing

The database kernel rules them all these solution

• divide et impera: divide the processing and collect results
• coordination for processing and job spawning 
• partial query are executed in sub systems
• data are collected and processed
• security policies are applied
• and final results are delivered
• still a single point of entry anyway!

the most important thing at the end is the optimizer!

• transformer
• estimator: cost estimation models
• execution engine

the dictionary

• the database has to know where the object are
• more improvements will come in this area

Conclusions

• exadata boosts execution for traditional application and analytics
• big data affordable data management for unstructured data
• in-memory fast for scans, joins and aggregation for analytics
• Data silos & isolated solution are being built again!
• oracle provides one integrated solution with a central point which is the oracle database