start portlet menu bar end portlet menu bar
  • Contact us

    • What we achieved

    • Fully

      automate complex scripts

    • Integrating

      with existing toolset

    • Regression

      testing runs more quickly and efficiently

    To who

    • Industry Banking Industry

    • Products: HCL DevOps Test

    • Region Northwestern Europe

    Overview

    • Part 1

      Challenge

      Open-source test automation has come a long way in the past decade, moving beyond the brittle libraries of old, to a more mature, User Interface (UI)-led test authoring experience.
    • Part 2

      Solution

      Six scrum teams were initially onboarded and trained to use DevOps Test with another ten teams planned. These teams are focused on regression and end-to-end use cases, and the testing needs vary from mobile to mainframe with complex, hybrid, and lengthy scenarios – some with desktop clients to browser to databases.
    • Part 3

      Results

      DevOps Test was integrated to the bank’s legacy Jenkins CI/CD pipelines (while they were still in the process of being migrated to cloud), as well as to those that were already in the cloud using the DevOps Test Azure marketplace plugins for Azure DevOps.

    The Challenge

    Open-source test automation has come a long way in the past decade, moving beyond the brittle libraries of old, to a more mature, User Interface (UI)-led test authoring experience. Tools like Postman, Selenium, and others are favoured by developers, often because they are easy to learn with no additional cost to their project. Open-source however, does not address all the needs of a large organization. In many cases, open-source tools still require the tester to write code, and for some this is not an option. In other cases, open-source simply does not accommodate the more complex scenarios, or the breadth of modern and legacy technologies. This was certainly the case for the bank, and they settled on a strategy of blending preferred open-source tooling with vendor products for ‘exotic’ legacy apps like mainframe and SAP, desktop applications, complex test cases that span desktop, mobile and web, and finally for the integration of Application Programming Interfaces (API) and endpoint activities into broader workflows.

    The bank embarked on a multi-vendor product evaluation that began before COVID had hit the international press. After meeting the HCL DevOps Test team at the EuroSTAR conference and being impressed by what they saw of the testing platform, HCL was invited to participate in the bake-off. By the end of the evaluation however, nearly a year had passed, and the testers were working mainly from home.

    Adding to the complication of remote work and partway through the evaluation, the bank announced a company-wide mandate to move all platforms, tools, and applications to the cloud, specifically to Microsoft Azure and Azure DevOps. While some vendors were excluded from the evaluation at this stage, due to their lack of support for cloud native architectures, DevOps Test’s hybrid approach to supporting both legacy on-prem and cloud native deployments shone through. As did its out of the box integration to Azure DevOps Releases, Pipelines, and Test Plans.

     

    Instead of mandating a specific tool, they wanted to support open-source tools along with vendor products.

    The Solution

    Six scrum teams were initially onboarded and trained to use DevOps Test with another ten teams planned. These teams are focused on regression and end-to-end use cases, and the testing needs vary from mobile to mainframe with complex, hybrid, and lengthy scenarios – some with desktop clients to browser to databases.

    The bank decided to accelerate the adoption of DevOps Test by employing a limited number of professional services from the HCLSoftware group. Subject matter expertise was on hand to help initially with the deployment and rollout of the software, in compliance with the bank’s stringent rules and processes, and later to assist with new user questions and issues to smooth the transition from the old automation tools to DevOps Test

    Using a combination of DevOps Test UI and DevOps Test API, the teams were able to fully automate their complex scripts and create the needed scenarios using ‘compound tests.’ The testers also found quite quickly that in some cases, they preferred to use DevOps Test in place of the chosen open-source solution, especially where there was additional value and capability offered by DevOps Test. A good example being the automated import and migration of simple Postman collections into the more powerful DevOps Test API tool.

    An example deployment architecture diagram used during the architecture board approval process is shown below, demonstrating the typical process through which DevOps Test integrates to the bank’s Microsoft Azure DevOps pipelines and Azure Test Plans:

    image-casestudy

    As a DevOps Test customer, the bank was offered access to the HCL Client Advocacy program. This allowed the bank’s team to meet with HCL weekly to discuss best practices and problem resolution, as well as to keep up with updates and new features in the products. The HCL Client Advocacy program ensures the customer always has the DevOps Test expertise needed for quick responses.

    The Results

    DevOps Test was integrated to the bank’s legacy Jenkins CI/CD pipelines (while they were still in the process of being migrated to cloud), as well as to those that were already in the cloud using the DevOps Test Azure marketplace plugins for Azure DevOps. The bank requested several customizations for the integration to their legacy Jenkins pipelines, and the DevOps Test team was happy to provide updates to support this.

    About the company

    HCL DevOps Test belongs to the Secure DevOps product domain of HCLSoftware which is a division of HCL Technologies (HCL) that operates its primary software business. It develops, markets, sells and support more than 20 product families in the areas of Secure DevOps, Automation, Digital Solutions, Data Management, Marketing and Commerce, and Mainframes.

    Read full story