Increase Efficiencies in DevOps to Accelerate Software Releases

No comments

Background:

Several enterprises have moved towards agile and DevOps model but have challenges in taking the advantages of these models. Most of these enterprises have legacy methodologies and technologies to release software that cause challenges in transitioning to modern efficient models. Our client is a global software company that serves financial services customers. Our client had several inefficiencies in their software release management that were causing delays in software releases and customer disappointment.

Challenges

Client had challenges in releasing the software on time as planned. First, there was no standardization in the release management process. The current release management process was partly automated with weaker accountability. The pass/fail criteria for a release every sprint in the SCRUM process were weakly defined and the overall team lacked transparency to make quick decisions on the issues. Second, there were several independent legacy technologies being used to build and release the software to N+1 environment. The manual handshake was not coordinated well and there was a clear need for automation of not just the handshakes but automating the critical test cases and regression test cases. There was no tool that orchestrated the release process end-to-end that gave transparency not just to developers but also to the product/business team. Finally, there were gaps in understanding between business and development/release teams on how and when the finished releases would be launched to the customers.

Solution and Outcomes

Antra captured and analyzed the current development, build, and release process to identify the key inefficiencies that were hindering high quality software releases to their customers. We participated in client sprint cycles and interviewed the team to understand and identify the process flows, tools used, and the key issues currently faced.

Antra’s solution composed of

  1. Standardize the software release management process to reduce complexity and increase transparency and accountability
  2. Automate the test cases and the release management process using industry standard technologies
  3. Establish a agile framework to prioritize and release the business features with the proposed automated solution

The solution offered:

  • Increased software release throughput by 70%
  • Faster time to market on business features
  • Faster regression cycles and Improved software quality

Antra solution also consisted a clear roadmap with an analytical and ROI model that demonstrates desired software release model with the solution. Antra proposed and implemented a test pilot software release with the proposed solution to demonstrate the benefits in the ROI model presented.

AthenaIncrease Efficiencies in DevOps to Accelerate Software Releases
Read More

Automating Continuous Software Delivery Across Multiple Applications

No comments

Background

Our Client is a financial institution that has several customer and internal software applications. The software development teams have an existing platform and a framework to develop and build these applications and the operations team was responsible for the infrastructure and tools to release the applications. The overall platform had the tools to build, release, track issues and notify, and governance of the releases. Client had automated the tests and used continuous integration tools (Jenkins) to automate the code integration and the build of the applications.

Challenges

The key challenge for the team was that the release process of a single application was taking around 10 to 15 days and multiple application releases made the management of the releases complex. The existing automation tools such as Jenkins helped Client accelerate some of the build and the release process, however, the overall orchestration process took significant time that included change management, issues analysis and decision making, issue resolution approval process, and rapid deployment of the code on to N+1 environments. In case of errors during the build process, rollback, issue resolution, approval, and redo, build were taking significant time. These integrations are at large done manually that created many dependencies and increased wait time. Deployment tickets were raised and the wait times were up to 24 to 48 hours to get the tickets resolved. In addition, there were no feedback and notifications to all the stakeholders. The resolution times sometimes took longer to analyze the issues during the build and make a decision to promote the build to N+1 environment.

Solution and Outcomes

To ensure continuous delivery, all the tools (including change management, release management, and ticket management) had to be integrated to optimize the DevOps model. To reduce wait time between promoting the build, a solution with automated validation of the errors was required. To manage multiple releases effectively and promote collaboration, a visual easy-to-use interface was required for orchestration. The solution involved analysis, design, and deployment of the tool along with the process changes that enabled complex multiple deliveries and orchestration of end-to-end application life cycle. Outcomes include:

  • Reduced build and release time to less than a day from 10 to 15 days
  • Increased deployments across multiple applications (more than 10 a day)
  • Reduced wait time between build promotions
  • Reduced dependencies and improved team collaboration
  • Improved continuous delivery across multiple applications and environments
  • Improved end-to-end process
AthenaAutomating Continuous Software Delivery Across Multiple Applications
Read More