Case Study – Insurance

Project background:

A company from the insurance industry, the leader of direct insurance sales. A company that is realizing a number of modern ventures, for which the time to deliver business applications is paramount. The company's vision is to determine the direction for future insurance.

Challenges:

The IT environment consists of one basic application, being developed by their own development team, and many supporting applications. This causes the tests of new functionalities to be laborious and finding the causes of the errors takes a long time.

Business departments expect IT to accelerate ad-hoc changes and shorten strategic project delivery times. They also expect the system to work more efficiently.

Solution:

DynaTrace solution was implemented. A very specialized lightweight agent (installed on a machine running Java or .NET application) allows to obtain the type of information as follows:

What methods are invoked within the monitored application?
What is the time to invoke each method?
Where is the application bottleneck?
What SQL queries are generated by the Java/.NET application and what is the timing of their performance?
Does the application contain any memory leaks that may cause it to stop working suddenly?

The above information provides a very detailed insight into the performance of Java/. NET applications - allowing to identify the application elements (to the level of the method and the SQL query) that do not meet the performance assumptions.

Implementation:

Unlike traditional solutions of this type, DynaTrace is based on the record format of the so-called transaction paths, so called PurePath. It allows the tool to store information about all transactions of all end-users passing through all the layers of the monitored service.

The uniqueness of PurePath is that the information gathered in this way is not averaged in any way. It stores information about each single transaction even on very large production environments. This unique functionality allows to obtain a completely new quality of operational, test and development business data. Using the functional analysis of each individual can be in the lifetime of the application – from its development, through tests till production monitoring.

The analytic module and dashboard-based reporting engine in the solution provide the following information to individual recipients:

Business owners obtain reliable, non-average data of SLA execution for key business transactions, broken down into synthetic traffic and real traffic, end-customer type, geographic location, and transactions business parameters.

The maintenance department obtains a tool for rapid analysis of problems sources, which is based on appropriately correlated information from all layers of the application service:

- Full transaction paths and time measurement from code level,
- Application server and JVM/CLR layers – metrics related to Garbage Collector, Heap, CPU, JMX/PMI, WMI,
- Virtual operating system layers.

Result:

Using this tool, the time to solve performance problems decreased by up to 90%.
It was possible to analyze and optimize the operation of the application under the account of the system response times, and thus the business time of service of a single client was shortened,
Architects have been given access to UML sequence diagrams, which enables them to verify the actual flow of information between the various components of application services.
Testers receive information about application issues when carrying out performance tests.
Data in the form of files containing individual paths can then be forwarded to programmers who can view them in a free program. This allows to „play back” the saved path, so that the programmer receives a tool that allows him to find quickly the source of the problem and come up with the solution.
Programmers analyze bottlenecks, memory leaks in the code already at the development stage, so that many unpleasant performance surprises can be avoided after production deployment, at the same time accelerating software production process.

Phone number:

Opening times:

E-mail address:

Case Study – Insurance

Project background:

Challenges:

Solution: