ADF Mobile: Implementing Data Caching and Synching

Implementing Data Caching
Strategies for ADF Mobile
Copyright © 2013, Oracle and/or its affiliates. All rights reserved.1
Strategies for ADF Mobile
Steven Davelaar
ADF/Webcenter A-Team
Oracle Corporation
@stevendavelaar

Disclaimer
The following is intended to outline our general product direction. It is
intended for information purposes only, and may not be incorporated into
any contract.
It is not a commitment to deliver any material, code, or functionality, and
It is not a commitment to deliver any material, code, or functionality, and
should not be relied upon in making purchasing decisions. The
development, release, and timing of any features or functionality
described for Oracle’s products remains at the sole discretion of Oracle.

Agenda
Data Caching and Data Sync Strategies
Implementing Data Caching and Synching Using A-
Team Mobile Persistence Extension

ADF Mobile Architecture
Device Native Container
Web
View
Server
HTML
ADF Mobile
AMX View
Third Party
Web Sites
Third Party
Web Sites
Oracle IDM
Oracle IAM
Oracle IDM
Oracle IAMLocal
HTML
HTML5 & JavaScript Presentation
Configuration
Server
Configuration
Server
ADF Controller
CredentialManagement,
SSO&AccessControl
SSO&AccessControl
App
Config
App
Config
Device
Services
PhoneGap/
Cordova
Server or Cloud
backend
Server or Cloud
backend
Mobile
Device
SSO&AccessControl
SSO&AccessControl
Server
SOAP & REST
Web Services
SOAP & REST
Web Services
Java VM
Business
Logic
ADF Model
Encrypted
SQLite DB
JDBC
SQLite

ADF Mobile Architecture
Device Native Container
Web
View
Server
HTML
ADF Mobile
AMX View
Third Party
Web Sites
Third Party
Web Sites
Oracle IDM
Oracle IAM
Oracle IDM
Oracle IAMLocal
HTML
HTML5 & JavaScript Presentation
Configuration
Server
Configuration
Server
ADF Controller
SSO&AccessControl
SSO&AccessControl
App
Config
App
Config
Device
Services
PhoneGap/
Cordova
Server or Cloud
backend
Server or Cloud
backend
Mobile
Device
SSO&AccessControl
SSO&AccessControl
Server
SOAP & REST
Web Services
SOAP & REST
Web Services
Java VM
Business
Logic
ADF Model
Encrypted
SQLite DB
JDBC
SQLite

Data Sources for ADF Mobile Applications
The data within an ADF Mobile application comes from 1 of 3 places
– Remote web services (SOAP, REST)
– Simple Java POJOs
– Local SQLite database
Typically all 3 are used in combination to allow your application to
– Web services: Retrieve and update data from/with remote servers
– Java POJOs: Cache that data locally for live access when disconnected
– Database: Persist & restore data when the application is stopped & restarted

Data Caching and Synching Challenges
Mobile devices can lose/turn off connectivity
Offline access to data is a common requirement
But it will increase the complexity of your application
If you cache data locally you must consider
– When to cache the data
Security
– When to cache the data
– How much data to cache
– When to synchronize updates
– Recording the order of updates
– How to deal with synchronization conflicts
– Security of the data if the device is lost
Security

1. Online Read/Write
• Needs to be continuously
connected
• Does not cache any data locally
• No synchronization required
• No risk of data theft if the device is
stolen
Data Caching
2. Cached Reads, Online Write
• Caches data as it is accessed
or on startup
• Updates are via web service
calls
• Small risk of data theft
Data Caching
Strategies
3. Cached Reads, Offline Writes
• Edits to cached data are saved
locally
• Edits to the local data are
periodically flushed to the server
• Greater risk of data theft
4. Full Synchronization
• All data is synchronized to the
device on startup
locally

ADF Mobile – Model Layer
Data Object
– Java class to hold attributes of an object
– Represents a single “row” of a collection
– Can contain sub-collections of other Data Objects to form complex object
hierarchies
Service Object
– Java class that provides CRUD operations
– Returns arrays of Data Objects in the get methods
– Exposed as Data Control to create UI using drag and drop

Java VM
ADF Model
Model Layer – Caching Data
Service
Object
DataControl
Data Object
JDBC code
Encrypted
SQLite DB
DataControl

Using the SQLite Database in ADF Mobile
Typically used by a single user
SQLite libraries and JDBC drivers are embedded in ADF Mobile
container
Encryption for the SQLite Database File is provided with ADF Mobile
Zero configuration
No Object-Relational Mapping (ORM) layer provided
– Access using plain JDBC statements

Initialize the Database
call Start()
LifeCycleListenerImpl.java
Start()
1. Creates the DB
2. Creates Connection
3. Connects
4. Populates DB
Application starts up
Note:
Need to add LifeCycleListenerImpl.java to the LifeCycleEvent
Listener Field of the adfmf-application.xml file.

JDBC Example – Get Departments
public void retrieveDepartmentsFromDB() {
try {
Connection conn = DBConnectionFactory.getConnection();
s_departments.clear();
conn.setAutoCommit(false);
PreparedStatement stat = conn.prepareStatement("SELECT * from DEPARTMENTS");
ResultSet rs = stat.executeQuery();
while (rs.next()) {
while (rs.next()) {
int id = rs.getInt("DEPARTMENT_ID");
String deptName = rs.getString("DEPARTMENT_NAME");
int mgrId = rs.getInt("MANAGER_ID");
int locId = rs.getInt("LOCATION_ID");
Department d = new Department(id, deptName, mgrId, locId);
s_departments.add(d);
}
rs.close();…

SQLite More Info
Check out SQLite website : http://www.sqlite.org/
Check out ADF Insider Essentials video by Frederic Desbiens
– http://www.youtube.com/watch?v=-XzE1n_j5Nc

Java VM
ADF Model
Service
Object
DataControl
Data Object
JDBC code
Encrypted
SQLite DB
RestServiceAdapter
DataControl
RestServiceAdapter
REST(JSON/XML)
SOAP/REST-XML

What is JSON?
JavaScript Object Notation
– text-based open standard designed
for human-readable data
interchange. It is derived from the
JavaScript scripting language for
JavaScript scripting language for
representing simple data structures
and associative arrays, called
objects. Despite its relationship to
JavaScript, it is language-
independent, with parsers available
for many languages.

Using JSON REST Service
Create a URL Connection that points to the JSON data host
This URL should be based on the root of all JSON services
– For example, for a JSON service that returns employees and departments
(http://server:port/service/employees and
http://server:port/service/departments), the URL Data Control should point
to http://server:port/service/.
Use RestServiceAdaptor to invoke service

RestServiceAdaptor Example

Java VM
ADF Model
Service
Object
DataControl
Data Object
JDBC code
Encrypted
SQLite DB
restServiceAdapter
DataControl
restServiceAdapter
REST(JSON/XML)
SOAP/REST-XML
SOAP/REST-XML Web Services
Data
Control
AdfmfJavaUtilities
invokeDataControlMethod

Using SOAP XML Service – Run Web Service
Data Control Wizard

Invoking SOAP Web Service Programmatically

Implementing Data Caching Strategies
ADF Mobile provides basic support to implement strategy 2: Cached
Reads – Online Writes
– Java coding required to convert web service payload to Java data objects
and vice versa
– Extensive JDBC coding required when caching should survive application
stop/start, and/or flexible data filtering is needed
Would be really nice to have an ORM mapping framework that auto
generates the JDBC code
Implementing strategies 3 and 4 with offline writes is much more
complex

Agenda
Data Caching and Data Sync Strategies
Team Mobile Persistence Extension

A-Team Mobile Persistence Extension
Sample code created by Oracle Fusion Middleware A-Team
Significantly speeds up implementation of data caching and data
synching
Provided “as-is”, no support, no updates
Provided “as-is”, no support, no updates
Installable as Free JDeveloper extension
Will be included in ADF Mobile later this year

Local and Remote Persistence
Extension runtime library contains generic Java code to perform CRUD
operations against SQLite database and against remote web services.
– Service objects extend EntiyCRUDService class
– Service objects use DBPersistenceManager for CRUD operations
– Service objects use DBPersistenceManager for CRUD operations
against SQLite database
– Service object can use a remote persistence manager for CRUD
operations against web service
The generic code in EntiyCRUDService class and persistence
managers is driven by metadata stored in persistence mapping XML
file

Remote Persistence Managers
Classes that perform appropriate web service
calls based on the CRUD operation performed
by the user and the persistence mapping
information
Currently persistence managers are available
Currently persistence managers are available
for REST web services in both JSON and XML
format, and SOAP (ADF BC) web services
You can easily create custom persistence
managers as needed
– Extend from abstract class that contains
convenience methods

ADF Model
Runtime Persistence Architecture
DepartmentService
Data Control
Department
DepartmentServiceEntityCRUDService
extends
DBPersistence
Manager
uses
REST-JSON
PersistenceManager
Persistence
Mapping XML
DEPARTMENTS
table
SQLite
JDBC
Statements
references
REST(JSON/XML)
HTTP
Requests

Other Runtime Persistence Features
Encryption of database
Auto-cleaning of unused database segments
Lazy loading of child collections (a.k.a “Indirection”)
Entity caching to minimize object creation
Entity caching to minimize object creation
FindEntityByKey
– Checks cache first, then database
Designed for customization
– Easy to override and extend default behavior

Mobile Persistence Extension - JDeveloper
Wizards
Three wizard are provided to create the artefacts needed for this
runtime persistence architecture:
– Mobile Business Objects From Web Service Data Control Wizard
– Mobile Business Objects From REST Web Service Wizard
– Mobile Business Objects From Database Tables (Local persistence only)
Another wizard is provided to generate a default CRUD user interface
on top of the business layer created by one of the above wizards.

Rest Wizard Demo - Toplink Data Services
/ToplinkRest/persistence/v1.0/Model1/query/Department.findAll

Wizards for Creating Runtime Persistence
Artefacts

Business Objects From REST Web Service

SOAP Demo – ADF BC SDO Services

Run SOAP Web Service Data Control Wizard

Mobile Business Objects from WS Data Control

Generated Data Object Classes

Generated Service Object Classes

Generated SQL DDL Script

Generated Persistence Mapping File

RESTful resource calls

SOAP method calls

Generated Config File

Configured InitDBLifecycleListener

Create Data Control For Service Classes

Creating the Mobile User Interface
Build Using Drag and Drop
from DataControl Palette
Generate Using ADF
Mobile User Interface
Two Options
Mobile User Interface
Generator Wizard

Building the User Interface

Using the Mobile User Interface Generator

Generated User Interface Artefacts

1. Online Read/Write
• Needs to be continuously
connected
• Does not cache any data locally
• No risk of data theft if the device is
stolen
Data Caching
2. Cached Reads, Online Write
• Updates are via web service
calls
• Small risk of data theft
Data Caching
Strategies
3. Cached Reads, Offline Writes
locally
• Greater risk of data theft
4. Full Synchronization
• All data is synchronized to the
device on startup
locally

A-Team Mobile Persistence Extension- Offline
Writes and Data Syncing
If a remote persistence manager is configured, and CUD service call
fails, the service call is registered as “pending” data synch action
– Failure reason registered (device offline, service not available, etc)
On next service call, the pending synch actions are processed first in
order of creation
– When failing again, the still pending synch action is updated with date of
last synch attempt and last error message
Out-of-the-box feature: NO Java coding required

Viewing Pending Data Sync Actions
A reusable “DataSync” feature is provided that can be added to your
mobile application
Register the feature in adfmf-application.xml

Viewing Pending Data Sync Actions
Add a button that navigates to the data sync feature

Viewing Pending Data Synch Actions

Notes on Data Sync Functionality
Pending changes can become obsolete because of updates in server
data by other clients
– The server–side code should identify this and throw an error when the data
sync action sends stale data
Stale data detection can be done by including a data revision number or
– Stale data detection can be done by including a data revision number or
‘last-modified” timestamp in payload. This number or timestamp should
match with server-side data record
Data sync actions that keep failing need to be removed manually
– Local database might be out-of-sync, app needs to offer reconcile action

More Info
A-Team Chronicles: http://www.ateam-oracle.com/?p=22331
Includes download links to
– A-Team Mobile Persistence Extension - JDeveloper Install File
– A-Team Mobile Persistence Videos
– A-Team Mobile Persistence Demo applications

Summary
Using the A-Team Mobile Persistence Extension significantly speeds
up and eases implemention of data caching and syncing using the on-
device SQLite database
Using the A-Team Mobile Persistence Extension significantly speeds
up and eases the use of RESTful web services
up and eases the use of RESTful web services
You need to think about how your end users should handle failed data
sync actions, and how to prevent updates based on stale data
Essential parts of A-Team Mobile Persistence Extension will be
included in future version of core product

ADF Mobile: Implementing Data Caching and Synching

More Related Content

What's hot (20)

Similar to ADF Mobile: Implementing Data Caching and Synching (20)

Recently uploaded (20)

ADF Mobile: Implementing Data Caching and Synching