Jet Infosystems advances QlikView to the corporate market

03/05/2012

Moscow – In the first project of its kind in Russia, Jet Infosystems completes load testing of the QlikView Business Discovery platform. The results of the testing show that the platform is suitable for building enterprise class solutions.

The QlikView product is well-known in Russia. It has an excellent record of handling local problems that involve a rather modest amount of data. Due to its flexibility, support for the associative model of data and ease of use, this product was selected by Danone-Unimilk for its business intelligence system. The solution developed by Jet Infosystems monitors performance at one of the client’s plants in real time, allowing for a fast management response in any situation.

QlikTech advertises the most recent QlikView v.11 as an enterprise level platform capable of handling large and sophisticated analytical tasks. Realizing the potential advantages of QlikView, Jet Infosystems offered to load-test the platform and demonstrate its operation to potential clients.

“Our experience shows that companies require efficient yet flexible business intelligence instruments that can analyze large amounts of data fast. QlikView appeared to be an excellent product," comments Anna Kharitonova, chief of the business analysis division at Jet Infosystems’ Software Solutions Center. “Thanks to these tests we had a chance to thoroughly explore the new version of the product and to make sure that it can indeed be deployed as an enterprise level solution.”

An impersonalized database containing about 3 billion records and examples of daily analytical reports was provided for the study by a corporate client with over 5,000 employees. Daily reporting at the company involves about 500 simultaneous users. These numbers are fairly typical for a large business.

Load testing involved three versions of the QlikView server and three benches with different configurations. At the first stage Jet Infosystems’ experts tested the platform on virtual machines. Once it was found that QlikView can efficiently process large amounts of data in the cloud, the tests continued on a physical server and a cluster. The loading scenarios closely approximated real-life conditions. The total number of load tests was over 20.

The platform easily handled 500 simultaneous users and 7 billion records, which is twice as much as the norm for a large business. During these tests Jet’s engineers also developed a methodology for customizing QlikView for large corporate clients.