Traditional Vs SOA testing

SOA promises benefits to organizations by providing a framework for building applications through re-usable components. This helps organizations evolve their IT solutions over time, as well as avoid costly integration and re-engineering efforts. SOA enables this by using service, rather than technology, as the basis of applications.

However, SOA based application development introduces new challenges in testing of applications built on top of them. By building applications based on loosely coupled services, the various phases of testing are required to address function as well as interaction complexities. Testers are required to wear additional hats, including the ability to bring in business analyst and system engineering into the testing process.

Testing of mobile based applications ?

Currently in the testing industry, there are very less tools which are available in the industry for performing mobile based testing. Also, industry available tools works based on the simulator, Optical character recognition (OCR) and is not effective enough for today’s business applications. Looking at the current and future trend, we will have lot of applications which are customized to mobile based applications and would require an extensive testing.


Having this situation, there will be a great demand for mobile application testing tools, which can effectively recognize mobile based controls and perform validations.


From my research i also understand that ,there are different kinds of python interpreters and libraries available which runs on different mobile platforms.Also there are python versions for different embedded devices, palm based devices and linux- arm based devices.


Have you ever experienced such a kind of tool, which can recognize mobile based controls and perform validations.?

SSIS - Data Integration Testing Tool

In today's corporate world, data is spread across multiple data sources (ex: DB2, Oracle, Sybase etc) and business users wish to generate types of business reports from a single source, no matter how the data is distributed. This scenario calls for data integration from multiple data sources and demands for new pertinent technical challenges:
a) Data sources with different formats and organized in structured, semi-structured, and unstructured data.
b) Huge data volumes and data populated from source systems arriving at different times.
c) Data quality and transforming the data into a format that is meaningful to business analysts.

Microsoft SQL Server Integration Services (SSIS) package enables developers to address the business users need. However testing of these packages introduces new challenges in testing. This scenario calls for the following needs:a) How do we validate complex sql queries?b) How do we ensure data workflow validation?c) How do we validate multiple data validation checkpoints?d) How do we ensure data correctness between the source and target, with millions of records?e) How do we approach continuous testing?

Recommended validation focus areas for this includes :

• Business rules validations and Data correctness
• Validate containers and tasks inside the SSIS packages from a functional perspective
• Validate usage of relevant field and table names, based on business rules
• Validate relevant column mapping between the source and the target
• Validate data truncation and data types between source and target
• Check for duplicate records
• Validate row count between the source and target
• Validat data correctness between the source and target

From a testing perspective, we must be able act upon these focus areas using both manual and automated testing.

Manual Testing:

List of functional and business requirements for each individual task needs to be identified and listed. SQL scripts present in the individual task needs to be validated against the functional and business requirements. SQL queries validation for appropriate field name, appropriate joins, relevant column conversion, column type-casting and relevant table names.

Test Automation:

Using test libraries , we can validate the structure of the data and data correctness. Based on the needs , i have created a test automation tool named "SSISTProbe". This tool works on the philosophy of XML tag based test automation approach. Some of key features of this tool is:

• Built-in intelligence, which executes relevant test actions based on the user XML tags
• COM based tasks and component execution , helps to execute tests in a non-UI environment
• Listeners to handle extraction of huge amount of data
• Data comparsion API's, to compare billions of records
• Multiple execution checkpoint support, which helps in enabling checkpoints in packages, databases, logs etc

This tool is open-sourced and the hompage url is http://www.codeplex.com/SSISTProbe