This IBM® Redbooks® Solution Guide shows how to take advantage of enterprise caching for a mobile environment using an integration solution that includes IBM Worklight and IBM WebSphere® eXtreme Scale or the DataPower® XC10. This solution guide describes the business value, integration concepts, and an overview of how to develop a mobile application to take advantage of WebSphere eXtreme Scale caching benefits.
In the current global enterprise business environment, with the millions of applications running across Apple iOS, Android, Windows Phone, and other platforms, mobile marketing brings more business opportunities for value added services but also brings more challenges to existing systems. Data is being accessed at a rapidly increasing rate to service mobile applications and the workload is becoming a burden to back-end servers. This guide shows how you can use IBM® WebSphere® eXtreme Scale or the WebSphere DataPower® XC10 Appliance to cache frequently used data, thereby improving the performance of mobile applications and reducing the traffic to back-end applications.
Figure 1. Mobile applications benefit from caching data
Did you know?
By the end of 2012, there were over 1 billion smart phones in use worldwide. In the next three years, that number is expected to reach 2 billion.
Similar to traditional applications, mobile applications need to cache large amounts of data in memory, which can affect performance. Yet the requirements for better performance, high availability, and linear scalability are critical to mobile services. Improving the user experience is the key issue faced by mobile businesses.
IBM Worklight is part of the IBM Mobile Foundation family and provides an open, comprehensive, and advanced mobile application platform. WebSphere eXtreme Scale is one of the caching infrastructures for the WebSphere family of products that has the ability to use distributed object caching to manage application data and business logic for multiple servers. By integrating these elastic cache capabilities of WebSphere eXtreme Scale with IBM Worklight, you can accelerate mobile businesses and provide a better experience for the user.
Mobile applications are often unique among application types. With mobile applications, extremely high and unpredictable workloads can occur because the applications can be accessed at any time and from anywhere. There can be multiple, frequent peak times. Consider also that the data that is exchanged is small in size, but frequently requested. In this type of scenario, back-end systems can become a bottleneck when the workload is heavy. Caching can relieve the pressure on the back-end system. In this solution, a mobile application caches data that is retrieved from back-end applications in order to respond to subsequent requests for the same data. This increases response time to the user and decreases the number of requests to the back-end server.
WebSphere eXtreme Scale servers and Worklight servers utilize WebSphere Application Server V8.5 Liberty profile servers. The Liberty profile is a composable and lightweight runtime application server; WebSphere eXtreme Scale features can be enabled quickly and easily using the Liberty profile. Without complex configuration and with only minimal code changes, mobile applications can benefit from the WebSphere eXtreme Scale cache seamlessly.
This solution uses the WebSphere eXtreme Scale client to access remote WebSphere eXtreme Scale grids that are then invoked by the Worklight HTTP adapter. As such, this solution focuses on using the WebSphere eXtreme Scale as a side cache.
An overview of the solution architecture is shown in Figure 2.
Figure 2: Solution architecture
This solution includes the following products:
- WebSphere Application Server V8.5 and the Liberty profile
- WebSphere eXtreme Scale or DataPower XC10
- IBM Worklight
- IBM Worklight Studio
Building the cache
The cache is implemented by defining a WebSphere eXtreme Scale grid that spans multiple Liberty profile servers. One set of servers runs the WebSphere eXtreme Scale catalog and the other set runs the grid containers. The container servers hold the application data stored in the cache. A catalog server provides administrative services to the container servers in its domain. The catalog server manages the placement of the data in the cache and keeps track of where the data is. When a request is made for data from the cache, the catalog server directs the requester to the proper location. Multiple catalog servers are used for failover purposes.
Each Liberty profile server is configured using simple XML and property files that define the features that your applications will use, including the WebSphere eXtreme Scale server feature and the location of the catalog servers. The WebSphere eXtreme Scale cache on these servers is also configured using simple XML files. When the Liberty profile servers are started, the WebSphere eXtreme Scale configuration files are detected and the caching grid is made available for use by applications.
Figure 3 shows a sample topology, with three catalog servers and four grid containers. Use one Worklight server as shown, or you can scale out to use a Worklight cluster for your production system.
Figure 3: Components of a Worklight application
Building the IBM Worklight server on a Liberty profile server
Beginning with V5.0, Worklight server ships with a Liberty profile, making implementation simple. The Liberty profile server is configured with an XML file that includes information about the Worklight server application and the Worklight application center. The XML file also enables the WebSphere eXtreme Scale client feature, giving applications on the server access to the cache.
When the server is started, you can access the Worklight console shown in Figure 4.
Figure 4: Worklight console
You can perform the following tasks on the Worklight console:
- Deploy a Worklight application
- Deploy a Worklight adapter
- Preview and test your Worklight server-side application with a browser
Using the cache from mobile applications
Worklight Studio contains an embedded Worklight server for development. To develop a mobile application with Worklight Studio:
2. Generate the artifacts for a specific device.
3. Build and deploy your application for testing, selecting a target device (for example, an iphone).
4. Run and test your mobile application with a simulator.
A typical Worklight application topology looks similar to that shown in Figure 5 and includes a 3-tier architecture with a client, a Worklight Server, and a back-end system, which might be a database, a service bus, and so on.
Figure 5: Worklight application topology
In this solution, the client communicates with the Worklight server through the HTTP protocol. The Worklight server uses JDBC to communicate with a back-end database.
On the server side, Worklight provides an interface to handle the incoming HTTP requests. If the application needs to talk with a back-end system, you can use a Worklight adapter. The adapters can also call server-side Java code; this solution uses this capability.
Consider a simple mobile application called Company Address Book (CAB). From the main screen, the user can search a company address book. Each search result is a list of records that satisfies the search criteria. The user can then drill down into a person’s data, including a photo of the person. The application obtains all of this data from the company’s back-end database. Search results are cached. For example, many people can search for sales reps in the organization or for the CEO. Caching the results of these searches eliminates the need to query the database again for corresponding requests.
The CAB application includes generic code that is common to all platforms and specific code for iOS and Android. The application also contains a server-side adapter for connectivity and a search adapter. All requests from a device go to the search adapter, which parses requests and is invoked by other adapters on behalf of the user.
To use caching, a customization class is created so that every request for data passes from the search adapter through the class. The class checks the cache for the requested data, and, if found, the data is returned from the cache. Otherwise, the class invokes another adapter to retrieve the data from the database, stores it in the cache, and returns the data to the search adapter. Then the data is sent back to the client device. This customization class works either for WebSphere eXtreme Scale or DataPower XC10. All configuration data is stored in Worklight configuration files and the cache server addresses can be changed dynamically without recompiling the code.
The data flow is shown in Figure 6.
Figure 6: Application workflow with WebSphere eXtreme Scale caching
An alternative to using WebSphere eXtreme Scale server running on a Liberty profile, is the use of the DataPower XC10 as a cache. Among other features, the XC10 can be used as a simple data grid. In this solution, you would simply define the cache on the XC10 and when you configure your Worklight server for the WebSphere eXtreme Scale features, you would point to the XC10 catalog and cache.
IBM WebSphere eXtreme Scale is supported on AIX®, HP, Linux, Linux on System z®, Solaris, Windows, z/OS® and z/OS.e. Details about platform support can be found at http://www-01.ibm.com/software/webservers/appserv/extremescale/sysreqs/index.html.
IBM Worklight Server is supported on AIX, HP, Linux, Solaris, and Windows. IBM Worklight Studio is supported on Linux, Mac OS, and Windows. Details about platform support for both IBM Worklight and Worklight Studio can be found at http://www-01.ibm.com/support/docview.wss?uid=swg27024838.
Ordering information is shown in the following table.
Table 1. Part numbers and ordering information
For more information, see the following documents:
- IBM Redbooks: Enterprise Caching Solutions using WebSphere DataPower SOA Appliances and WebSphere eXtreme Scale, SG24-8043
- WebSphere eXtreme Scale V8.6 Sales Manual
- WebSphere eXtreme Scale home page
- WebSphere DataPower XC10 Appliance Sales Manual
- WebSphere DataPower XC10 Appliance home page
- IBM Worklight home page
- IBM Mobile Foundation home page
This material has not been submitted to any formal IBM test and is published AS IS. It has not been the subject of rigorous review. IBM assumes no responsibility for its accuracy or completeness. The use of this information or the implementation of any of these techniques is a client responsibility and depends upon the client's ability to evaluate and integrate them into the client's operational environment.