The Struggles with Guidewire Cloud Upgrades
Navigating multiple Guidewire Cloud upgrade projects offered many learning opportunities. Two key takeaways emerged: the necessity of thorough pre-migration preparation and the critical importance of automated testing. Tackling migration complexities head-on should yield substantial dividends, outweighing any perceived costs of doing so. However, the journey also shed light on the realities of Virtual Machines (VMs) – instead of delivering promised benefits, they introduced unexpected challenges and delays, underscoring their potential risk to project success.
Let’s show you how we came to these conclusions.
Taming the Virtual Machine(VM) Beast
We have faced significant challenges on all our projects using Virtual Machines. VMs are envisioned as efficient workhorses but are more like stubborn mules, slow and resource-intensive. Their challenges required careful configurations and strategic resource allocation, and we achieved only two workable VM environments across several simultaneous projects in two years. Even then, the high cost of maintaining them deterred business from fully adopting the solutions.
Paradoxically, to avoid the “high” direct costs associated with an effective VM, business unintentionally incurred a significantly higher cost by reducing developer effectiveness. Once such example, simple compiles took well over an hour on the VM. This trade-off emphasized the need for balancing upfront investment with long-term productivity gains in any cloud migration.
The Missed Opportunity of Data Cleanup
The forest of data migration was a harsh landscape not properly navigated. It contained orphaned records and legacy issues which acted as undergrowth and impeded the team’s progress. Addressing these issues with data cleanup tools and strategies early in the process would have facilitated a smoother migration and is a crucial step that should not be overlooked.
Cracking the Source Code Puzzle
Guidewire’s Cloud Assurance Assessment (CAA) standards were not effectively leveraged lengthening our journey. A significant portion, as much as 90%, of the necessary changes could have been addressed even before the migration started. What’s more, a preemptive cleanup could have been handled by the internal team reducing the financial burden, and letting the more expensive specialists focus on the complex issues that cannot be resolved before the migration begins.
Navigating Unclear Requirements
As we all know: over time, things change. Requirements that once existed might now be lost or, worse yet, contradict the current operations. Amidst these transformations, a solution emerged: effective automated testing based on the original codebase could serve as an anchor, ensuring consistency and reliability throughout the migration. Then even when the present seems to contradict the past, test automation can ensure the new does what the old did, helping the team manage uncertainty and maintain progress in complex migrations.
Learning from Unexpected Hurdles
Unexpected challenges shed light on the need for foresight. For example, loading oversized rate books consumed hours instead of minutes. Use only a subset in development but continue loading everything in deployed environments to ensure it still works. Missing admin data also caused linking issues when loading environments, so taking the time to create a reusable process paid for itself and more. And introducing requirement changes during migration led to many complications; hence, separating migration from new requirements proved invaluable. These were just some of the hurdles encountered, but recognizing and addressing them kept them from becoming mountains.
Surviving the Trials of Insufficient Testing
Just as the team thought they reached the end of their journey, performance issues under heavy load and non-thread-safe code emerged. These unexpected challenges underscore the critical importance of comprehensive and rigorous testing from the beginning of a migration project, to preemptively expose potential performance bottlenecks and problematic code.
Conclusion
These hard-learned lessons underline the importance of strategic planning, rigorous pre-migration preparation, and a strong emphasis on automation and testing in any Cloud migration project. It’s a reminder that every challenge encountered is an opportunity to learn and refine your current process.
We are here to help and offer our insight so your project can avoid these obstacles.
Kim Filiatrault
Founder and President
Watch or read more CenterTest related posts, understand all features of the CenterTest Automated Testing Framework for Guidewire, or Connect with us to understand how we can help improve the effectiveness of your testing process.
