Showing posts with label eclipsesummit2009. Show all posts
Showing posts with label eclipsesummit2009. Show all posts

Saturday, July 18, 2009

Post Lunch session -OSGI

Lunch was extremely good and felt good after eating. Had only less though so as to keep in make.

Now the topic is on OSGI by Sameera Jayasoma . WSO2 is an open source company hased in Srilanka. Their main product is WSO2 carbon,Which is a fully componentized SOA platform based on OSGI. Then there was some PR on their company.


Modular Systems
Break large system into more smaller,understandable units called modules.Benefits of modular systems are reuse,abstraction,division of labour and ease of repair.


Module should be self contained,highly cohesive and loosely coupled.

Java for building Modular systems
--> provides class level modularity(public, non public methods)
--> we need somthing like external packages, internal packages in a jar

ClassLoader hierarchy : Bootstrap classloader (rt.jar) --> extension classloader (ext.jar) --> application classloader (a.jar,b.jar)


Problems with JARs

--> Jar is unit of deployment in Java and typical java app consists of set of jar files

--> no runtime representation for a JAR.

when java loads a class, it tries to load class one by one in the classpath.

--> multiple version of jar files cannot be loaded simultaneoulsy
--> Jar files cannot declare their dependencies on other jars

Java lacks true modularism and dynamism. Java has given flexibility to build system on top of it. This is called OSGI (dynamic module system for java)

OSGI

--> Bundle is a unit of modularization of OSGi. OSGI app is a collection of Bundle.
Bundle is similar to jar and contains additional metada in Manifest.mf file. In osgi , java package is the unit of information hiding. unlike in jar when class is the
unit of information hiding.

--> bundle can share and hide packages from other bundles.

sample manifest.mf file

Bundle-MainfestVersion:2
Bunle-Name :
Bundle-SymbolicName:
Bundle-Version:1.0.0 (major,minor,micro version)
Export-Package
Import-package:

symbolic name and version are unique for a bundle. default value for version is 1.0.0
all packages in bundle are hidden from other bundle by default. if we need to share package explicitly mention the name.

Bundles & Class Loaders
OSGI gives separate classloaders per bundle,thereby eliminating hierarchial classloading in java

System bundle is a special bundle that represents the framework and registers system services

Tooling for osgi : eclipse pde is the best of the lot

osgi provides command line interface

ss -> lists all the bundles in the system

b/bundles - gives the list of informaiton about the bundle.export information.

b 0 --> gives system bundl information

export-hundle and import-bundle
require-bundle : import all exported packages from another bundle. but this is discouraged.

Issues with require bundle
Split packages,bundle changes over time,require bundles chain can occur.

fragment bundles : attached to a host bundle by the framework.Shares parent classloader.

runtime class loading in osgi: order of loadin

1) call parent clas loader (only for java.*)
2)imported packages
3)required bundles
4)load from its own internal classpath
5)Load from fragment classpath

Usage of fragment bundle 1) provide translation files to different locales

OSGI specifications --specifies framework and specificaions

osgi alliance

current version is 4.1

osgi can be considered as a layer on top of java.can also use jni to talk to OS. functionality of framework is deivided into several layers

1)module 2)lifecycle 3(

Lifecyce layer : manage lifecyle of api.
bundle starts --> installed,resolved,starting,active,stopping,uninstalled.

Register a service using bundle service using registerservice on bundlecontext

Using a service --> find a serviceReference,get the object,cast it to a proper type and use the service

Events and Listeners

framework fires serviceevents during registering a service,unregistering a service

services are dynamic. monitor services using service listeners,service trackers,declarative service,iPOJO and blueprint services.

Service Tracker

declarative service

SCR (service component runtime)

Powerful Reporting with BIRT


This is the sesison that i hope to get the maximum benefits from especially since we have started using BIRT for our projects. The session is by Krishna Venktraman from Actuate, who is the director



Background

--> Most applications have some kind of data visualization need. Real world applications don't consider reporting when they design the application. BIRT provides a framework that manages all the reporting requirements.

Traditional approach :

Buy closed source commercial products or build custom developed solution
With open source based products things become much easier.

Emergence of BIRT Project
BIRT was initiated in 2004 as a top level eclipse project. IBM,Innovent and Actuate are the key members.
Focus was BIRT was to make report development easy.Its open and standards based and provides for rich programmatic control. It has both simplicity and well as Power to create complex reports. BIRT supports of concept of reporting libraries that promotes reuse and reduces changes.

The main audience for BIRT was report developers,advanced report developers who use scripting, runtime integration developers who use birt viewer and engine apis, report design integrator who use design engine apis, extension developers who develop extention points and finally core development who do eclipse development itself.

There were five major releases since project launch with 1.0 ,2.0,2.1,2.2,2.3,2.5 as the versions. It was built from ground up and lot of community feedback was take into account

Key capabilities
--> Simple to complex layouts
--> Comprehensive data access
--> Output formats
--> Reuse and developer productivity
--> Interactivity and linking
--> Multiple usage and productivity aids

Some key features added in 2.x versions
--> ability to join datasets
--> enhanced chart activity
--> multiple master page support
--> Dynamic crosstab support
--> CSS files can be linked
--> Emitters for XLS,Word , PPT and Postscript
--> Webservices can act as datasource
--> Javascript/Java Debugger

BIRT design gallery : some of the displays llok really good.

High level BIRT acrhitecture

Two key components
1) Report Design Engine
2) Report runtime Engine

at the end of design process we get a .rptdesign. report engine then looks at the file, interprets it, fetches the data and goes about the generation process. we get an .rptdocument. Key services of report engine are generation services,data services,charting engine and presentation services.

BIRT exchange is a community site for posting all BIRT related issues.

Key API
a) Design Engine API b) Report Engine API c) Chart Engine API

Extension points
--> data source extensibility
--> custom business logic extensibility using scripting and acess java code
--> Visualization extensibility
--> Rendering content for output (emitters)

Deployment options

--> Commercial Report Server
--> J2EE Application server
--> Java application

Actuate provides Actuate BIRT report designer,BIRT report studio, BIRT Viewer,BIRT Interactive Viewer,deployment kits, iServer Express,iServer Enterprise.

Now for the actual report designs...

Just had an overview of the birt tool. Going through the series of demos on basic report, now basic charts, then crosstab/matrix report.

Book Marks and hyper links

Click the element and then pickup bookmark from properties . Now go to the place where you want to place hyperlink and link up the bookmark.

Filters
Limit what to display . You can limit at dataset level or at table level.

Sub Reports

Main Report --> Outer table
Sub Report --> nested table
Pass data value from outer table to nested table

In BIRT we need to nest tables in order to create sub reports.
Use Dataset parameter binding on the child set to get the data from parent dataset

BIRT Scripting

Mozilla Rhino script engine is embedded in BIRT.
Scripting = Expressions + Events
It users server side scripting. All the export options will get the same output.


Event Handling
Initialization Report level events (-initalize,-beforeFactory) --> Data source Open (-beforeOpen,-open) --> Data Set Generation (--beforeOpen,--open,fetch) --> Dataset Generation

Generation phase : Report level,datasource level, element level
Render : report level,element level

can implement powerful logic using scripting

Report Libaries
Just use export library by right clicking on the rptdesign file. Then click on use library and say use library.It has all the data sources, data sets and report items.
Library is a container of reporting items.
Can do local overrides on things imported from libraries

Templates
File --> New Template.Serves as starting points for speedy report development.
Display a visual image. then register this template with the new wizard.

Last piece was a demo on how to deploy the reports.

One big disappointment was i couldn't get any idea on how to integrate report engine and birt viewer with the custom applications.

On the whole it was a good session.

Keynote II



Enhancing the developer productivity using RAD for websphere


--> Interesting the two main competitors give the keynotes one after the other

IBM Rational Architecture and construction
--> for solution,software and enterprise architects : Rational Data Architect,Rational Software modeler
--> for architects and designers who code in java,c++,xml --> Rational software architect Standard edition
--> Rational application developer --> for developers


IBM RAtional Applicaiton Developer for websphere

--> accelerates development
--> do even traceablity matrix
--> support for jpa and code generation

Comprehensive JEE5 tools

Unit testing is provided out of box
Visually layout JPA and EJB 3.0 classes
Convert UML to WSDL using RSA product
Provides excellent AJAX support

Enhanced support for Web 2.0
Declare POJOs and make it as REST service. Call Rest service as JSPs.
Javascript editor and debugger

JSF support
visual development of JSF pages
Data abstraction objects for easy data connectivity - SDO
Automatic code generation for data validation,formatting
Integration of third party JSF libraries

Portal development support is also excellent.

One of the very key features is to Automate performance and memory testing . This is built on top of eclipse TPTP.

Automates azpplication testing using WebSphere Application Server 6.0,6.1 and 7.0

IBMs strategy is to deliver high quality solutions by moving towards flexible architecture,automation layer,reduce onboarding

JAZZ platform : is meant for open community.

Mail id of the presenter is :bhrengen@in.ibm.com

Day 2 live blogging - EclipseSummit India

Participants are quiely settling in. Strength has considerably dwindled compared to yesterday. There was unexpected drizzle in Bangalore , causing me to get partly wet and making me feel miserable in an AC room.

Wireless Internet connection does not seem to work as expected :(. Hope to post all these one by one when i am back in the network.


Expectations for today
Today I am eagerly looking forward to attend BIRT workshop in the morning followed by a sumptuous lunch and then another workshop on OSGI.

Friday, July 17, 2009

JPA 2.0 New Features

JPA 2.0 is releasing 2009 fall. (JSR 317)

Goal

--> Fil in ORM mapping gaps
--> Make object modeling more flexible
--> offer simple cache control abstraction
--> allow advanced locking settings
--> JPQL enhancements

More standardized properties
--> some properties are used by each and every driver

Persistent unit properties like javax.persistence.jdbc.driver, javax.persistence.jdbc.url

JPA 2.0 supports join table as well as f key relationships.
Collections of embeddables or basic values.

This is made possible using elementcollection annotation

OrderedLists
order is maintained even if you move things around by using an index.

More Map flexibility
Map Keys and values can be : Basic objects,embeddables, entities

Enahcned embedded support
Embeddables can be nested and can have relationships

Access Type options
--> mix access modes in a hierarchy (field or properties)
@Access annotation

Derived Identities
JPA 1.0 relationship field cannot be part of the id

JPA 2.0 @Id + @OneToOne

Second Level Cache
API for operating on entity cache which is accessible from EntityManagerFactory
Supports only ver basic cache operations , which can be extended by vendors

@Cacheable annotation on entity (default is true)

There is also a property by the name shared-cache-mode to denote per PersistenceUnit whether to cache.
--no entities
--only cacheable entities
-- all entities

Properties cacheRetreiveMode and cacheStoreMode per EM method call whether to
read from cache

Locking
1.0 allows only for optimistic locking
2.0 provides for optimistinc locking by default
pessimistic locking can be used per entity or query

Lock mode values introducted
Optimistic (=READ)
optimistic_force_increment(=write)
pessimistic_read
pessimistric_write

API Additions
Lockmode parameter added to find,refresh
Properties parameter added tofind,refresh,lock
Other useful additions
void detach(Object entity)
unwrap
Query API addition
getFirstResult
getHints
Enhanced JP QL
Timestamp literals
Non-polymorphic queries
IN expression may include collection parameter --> IN (:param)
Ordered List indexing
CASE statement

Criteria API
Similar to eclipselink expressions,hibernate criteria
Criteria API - Canonical Metamodel
--> For every managed class , there is a metamodel class


Load State Detection
JPA can integrate with Bean validation (JSR 303) : Hibernate Validator is an implentation

JPA 2.0 shpped as part of Java EE6 release

JOptimizer

Back after a sumptuous lunch

Tool by Embarcadero technogies : http://www.embarcadero.com/products/j_optimizer/


Available standalone as well as eclipse plugin

Uses
--> Detecting excessive object allocation

--> memory leaks

--> detecting bottle necks

--> code coverage/code quality

--> thread debugging

--> Breakdown JEE requests

--> request analyzer

remotely connect and find which layer is causing problems. Can go into any level of detail.

code coverage

how to analyze and find out threading issues ?

--> Detecting deadlocks and visually analyze the deadlocks.

richard.davies@embarcadero.com

Patterns in eclipse

Three rules of eclipse

plugin may not change screen,task and create objects unless user asks for it

1) User owns the screen

2) User owns the CPU

3) User owns the memory

--> Do not use singleton pattern.Memory allocation never goes away

Building the UI withoug loading plugins

Decoupling using adapters
--> dynamic implementation of an interface
--> class can be adapted to any interface
--> best example is the properties view

4) most of the items are implemented as services
Located,Scoped and Destroyed

5) Separation of concerns
same class should not do a,b,c,d,e
use handlers
6) scalable UI
browse instead of combo
filters instead of presenting a large set of info
always use group data in relevant sections
provide deselect /select all buttons
differentiate and place common buttons appropriately

Eclipse and Building Data Centric RIA

Demo of portal type application using flex
Interactive charts --> looked really good with nice drilldowns

Rich Internet Applicaions
1) REal time data push
2) resizable
3) Rich Data entry
4) Chat
5) Data Synchronization
6) Audio and Video
7) Offline

Open Screen platform : Adbobe flash platform

clients --> AIR and flash player

Servers/services --> Blazeds,data services

Flex framework

tools to design and develop : flex builder

Understanding flex

written in mxml markup

flex sdk is free and opensources

flash builder : eclipse based professional IDE

benefits :

UI goes to client only once . after that only data changes..


Approach for Developing RIAs

Design focused
Data focused

Coding

Testing and Deployment

Demo

Model Driven development

Steps

1) Create a new project

2) Create a data model (fml file)

3) Deploy model to LCDS (Live cycle data services)

4) import fml file

5) create mxml and add datagrid.

6) Dnd datamodel to data grid

7) Run application.

Business logic can be written in custom assembler. Normal one is fiber assembler.

Even tomcat is fine. Internally fml creates java classes at runtime during startup.


Data Centric Development using Flash Builder

1) Define Service (CF,php,java,soap,rest)

2) Model Service (flash builder examines service, builds design-time model)

3) Bind Operations to Flex UI Components (data binding,UI generation,Paging,Data-management)

AMF : Action script messaging format used for sending and reciving data , more efficient than json and xml.

Testing and deployment

Network monitor
unit testing framework
Command line builds ( coming soon)

Demo of Ruby service plugin

Data Centric Development Extensibility

Extension points
Key interfaces and classes for custom service
IServiceWizard,IServiceIntrospector,IServiceProperties
FlexService,FlexServiceOperation,FlexServiceConfig
Key interfaces and classes for componentConfigurator


Flash Catalyst

Imports photoshop files and analyzes it. Based on eclipse platform. takes mockups and creates flex application based on them. Generates flex code behind the scenes.

Wow.. thats too good

Require flash builder license and run time license for LCDS

Resource

www.flex.org
Tour de flex : www.adobe.com/devnet/flex/tourdeflex/

EclipseLink : High Performance Persistence

I have used eclipselink on one of the key projects lastyear. It was donated by Oracle to Eclipse foundation. Toplink became eclipselink. I was really impressed by its implementation of JPA and customizations such as Criteria Queries. Hoping to learn more from this session.

EclipseLink Architecture

Supports JavaSE,EE,OSGi and Spring

Eclipselink solution comprises JPA implentation,Moxy,EIS,SDO(service data objects),DBWS (database webservice --> want xml from relational data ).

JPA
--> supports JPA 1.0 with many advanced features
--> simplified configuration of using annotations and/or xml.
--> Best ORM for Oracle database

Tool Support
-->EclipseLink is a runtime project supported by IDEs.
-->Eclipselink supported by Dali
-->MyEclipse
-->Oracle Jdeveloper 11g,Enterprise Pack for eclipse
--> Netbeans
--> Standalone workbench(Swing project)

Eclipselink distributions
--> Eclipse,Oracle toplink,weblogic,glassfish,spring framework,JOnAS

JPA persistence Provider(eclipselink) sits between Java SE/EE and rdbms


Core JPA Mappings

ID (primary key)
Basic (field mappings)
Relationships
OneToOne,ManytoOne,OneToMany,ManyToMany

Configurations
Persistence.xml
Mapping Annotations
Query hints
Advanced Mappings
Eclipselink orm.xml

Examples of Advanced Mappings
class level --> @Converter(name="",converterClass="")
field level --> @Convert(name="")

@PrivateOwned (coming in JPA 2.0) : doesnt exist without parent

Caching :
@Cache(type=SOFT_WEAK,coordinationType=SEND_OBJECT_CHANGES)

Optimistic locking :
@OptimisticLocking(type=CHANGED_COLUMNS)

Custom Data type conversions:
@Converter,@TypeConverter,@ObjectTypeConverter
@StructConverter

Query Framework

5 different ways

Entity Model : JPQL
Expressions
QBE
Native : SQL
Stored Procedures

Customizations in queries
--> Locking,Cache usage
--> Optimizations (batching,joining)
--> Result shaping/conversions
--> Stored Proc support

SQL and Stored Procedure directly hit db.

For other 3 query framework is used. Checks for Cache hit,Cache result is used using Object builder)

Annotations for stored procedures
@NamedStoredProcedureQuery and @NamedStoredProcedureQueries

e.g. @NamedStoredProcedureQuery(name="",procedureName="",parameters={})

Hints
eclipselink.read-only : useful in read only screens

Lazy loading & Fetch Groups
Two fetch groups defined automatically --> eagerly loaded fields and lazily loaded fields.Enabled by byte code weaving

Create your own fetch groups and added to query as hint. Overrides default fetch groups.

Caching Advantage :

Shared L2 and persistence context caching
--> Entity cache and not data cache like ehacache that comes with hibernate

Cache Invalidation --> time to live,fixed times, programmable

Cache Coordination --> Messaging,Type specific configuration (invalidate,sync,sync + new,none)

Invalidate is the most used and cheapest one

Annotations for caching : @Cache,@TimeofDay,@ExistenceChecking

Concurrency Protection :
@OptimisticLocking : Numeric,Timestamp,All Fields,Selected fields,Changed field

Pessimistic locking : using hit eclipselink.pessimistic-lock


Dealing with DB Triggers

@ReturnInsert and @ReturnUpdate : efficiently reread modified data in Oracle db.

Change Tracking
@ChangeTracking annotation to configure the way changes to Entities are computed
Attribute level ,object level, deferred ,auto

Performance Options summary
--> ChangeTracking
--> Read only queries
--> parameter binding
--> joining
--> inheritance
--> concurrency
--> dynamic expressions

Moxy : Object XML Binding

JavaApp --> Objects --> EclipselinkOXM --> xml

Rich set of mappings provides complete control to map objects to any XSD.

Eclipselink JAXB2 Annotations , Moxy XML

twice as fast as sun ri and xml beans.

API steps
1) Create factory,context
2) Marshal and UnMarshal

SDO : Xml Centric

Wow.. too fast session .. but left me lots to explore in JPA.

Eclipselink 1.1.2 released as part of Galileo

Eclipselink 2.0 --> Fall 2009 ; implements JPA 2.0

DBWS : Generated JAX-WS from DB

Keynote address - Day 1

First PR on Saltmarch going on :).

Sponsors : Platinum --> Oracle,IBM and Microsoft
Gold : Actuate,Adobe etc

Key note speaker will be Ramkumar Kothandaram from Microsoft

What is microsoft doing in an eclipse conference !!!

Topic will be interoperability.

Agenda

--> Microsoft's approach to interoperability
--> Open source & Microsoft platform

Need for interoperability


We have diversified client applications (firefox,ms office,Open office , IE8) and server applications (Jboss,.NEt etc..) , storage(EMC2,netapp) , multiple databases, processors(intel,ibm,sun).

Customers buy heterogeneous hardware/sofware. Painful EAI projects used to take a long time. This resulted in interoperability evolution.

Microsofts approach to Interoperatbility

--> Products : All products natively interoperability
--> Colloboration : with Partners,competitors and open source community
--> Standards : Promote interoperability through new and existing standards
--> Developer resources : msdn,codeplex etc

Products

All products like Windows Vista,.net ,Win server 2008,office 2007 support interoperability.

Can consume and expose webservice endpoints.

Colloboration

Apache stonehenge : Project to test for interoperability between various vendors.

MS part of interoperability alliance. Took leadership in webservices interop.

Standards

Participates in over 150 standards bodies

Developer Perspective

PHP on Windows --> PHP FastCGI on IIS

Seems that PHP on windows scales as fast as Apache ..

Then there was introduction to Microsoft Azure which is microsofts cloud computing platform.

Eclipse tools for sliverlight (eclipse4SL) : eclipse plugin for silverlight from soyatech. Looks good to me :). Can build RIA and embed in Html.

Java API for open XML : standard for storing documents. Create a document for server side. http://poi.apache.org

Links to remember

http://www.microsoft.com/web

Codeplex : 80000 open source apps on MIcrosoft application.

On the whole it was a good eye opener on microsoft and its contribution to interoperability.

Live Blogging from Eclipse summit - Bangalore

Here I am back again with live blogging from Eclipse summit Bangalore.

We are minutes away from launch and there seems to be a lot of Java/Eclipse lovers. The ballroom is really full.

The keynote address is set to begin