Perspective

In today’s software engineering, as never before, tests have become a critical ingredient. The stability and functionality of the software is now at the core of the software development process. Methodologies such as TDD (Test Driven Development) or BDD (Behaviour Driven Development) clearly indicate the trends in the future development of software engineering.

Such tests can be subdivided into many different types – depending on their intended use, application design, test duration and testing method. Consequently, a single test can be on the one hand a functional test, and on the other hand a part of the acceptance test, while some software developers will also regard it as an element of the regression test.

The names of the tests determine their design and logic.  A test should primarily be a tool to create, maintain and use a system whose quality is continuously being improved.

Approaches

Tests in terms of the time of creation of the tested parts of the application:

• Manual tests of newly developed parts of the system,  including both functional testsand non-functional tests, integration tests, etc.  concerning the current sprint/evolution/prototype/iteration.

• Automatic regression tests of parts of the system developed during previous sprints/evolutions/prototypes/iterations, which were tested manually during the previous cycles.

We perform the following tests:

• Functional black-box tests from the level of the user interface (WWW)
• Structural white-box tests
• Integration tests
• Performance tests.

Testing procedure

The cyclical testing procedure usually looks like this:

• Creating test scenarios
• Providing the Continuous Testing service implemented using automatic tests
• Repeating the automatic testing process (regression tests) after completing a particular fragment of software development
• Manual tests based on the previously defined test scenarios
• Determining test scenarios to be automated
• Implementing test scenarios as an element of automatic regression tests in the next test run.

Perspective

Computer virtualisation has significantly contributed to the use of available computing power and – to an extent – made applications and operating systems independent of the particular complexity of hardware. However, virtualisation is not without its drawbacks. The most important of them is the loss of performance to make multiple copies of identical operating systems.

Containerization is a type of virtualisation that is free of this problem. A container is a single application, including all of the necessary elements (executable files, libraries, configuration, etc.). In general, the container does not include the operating system. If the entire system consists of multiple applications (e.g. database server, application server or web server), each of them can form a separate container. Containers can be combined into larger, interacting structures.

Individual containers and entire sets of containers, or even computer networks, can be run in the cloud. Although containerization is generally implemented using the Docker system, sets of such containers are usually created, run and controlled using the Kubernetes system.

Thanks to Docker/Kubernetes, it is possible, for instance, to define efficient systems based on micro-services, which can perform even the most complex processes in a planned, secure and fully manageable way.

Scope

We develop software through practical application of containerization (Docker/Kubernetes) in the full spectrum of software production problems:

• Continuous Integration
• Continuous Testing
• Continuous Delivery

We have extensive expertise and practical experience in the following areas:

• efficient container creation
• container distribution
• defining the security context
• Kubernetes cluster resources management
• work automation.

We routinely resolve our clients’ problems, helping them create and manage local and cloud-based Kubernetes clusters.

We also have our own server resources which can be provided in order to containerize individual applications or even entire systems composed of hundreds of applications and systems.

Services

Our services:

• analysing the existing systems and applications in terms of containerization opportunities
• containerising applications and entire systems in the Docker environment
• orchestration, i.e. developing sets of containerized applications and creating virtual networks to combine these applications and create complete systems
• planning and designing micro-service environments – based on Docker/Kubernetes or PTR Atom micro-services
• creating, managing and maintaining Kubernetes clusters both in our infrastructure and in the client’s infrastructure, and also in an environment such as the Google Kubernetes Engine
• monitoring containerized environments
• defining the security context
• full range of consulting and training regarding the Kubernetes/Docker technologies in all of the above areas.

Software as a Service (SaaS)

Comprehensive service consisting of developing, implementing and maintaining the client's software at eSourcing. The systems function in a highly efficient and optimised technical infrastructure of eSourcing that guarantees maximum security, and all processes are managed entirely by our specialists.

Infrastructure as a Service (IaaS)

Outsourcing model where the client is provided with IT infrastructure – hardware and dedicated software. In the IaaS service, similarly to SaaS, the hardware is located in a secure server room of eSourcing, but it is the client that is responsible for managing the software installed there. In this model, we take over services directly related to the hardware layer, e.g. monitoring of devices or back-ups.

Platform as a Service (PaaS)

This service consists of eSourcing providing a fully functional hardware and system platform to be used to run the client's applications. Additionally, as part of the PaaS service, eSourcing ensures tailor-made package of maintenance and administration services and recovery of lost data, if necessary.