Executive Outlook:
In the ever-evolving landscape of technology, keeping our products competitive requires continuous changes and adaptability. Redeploying backend services becomes a routine part of this journey, and staying ahead means addressing challenges head-on. In this article, we’ll explore eight obstacles that slowed down our backend workflow and detail the strategic solutions we implemented for improvement.
1. Streamlining Service Node Setup with Dockerization:
Setting up environments for new service nodes was a repetitive and time-consuming task. To break free from this cycle, we embraced Dockerization. Docker containers encapsulate our services, allowing us to manage them seamlessly across various operating systems with a single DockerFile configuration.
2. Effortless Deployment of New Features:
Transitioning from traditional binary deployments, we adopted Docker and Docker Compose for a more straightforward feature deployment process. Leveraging Kubernetes and Gitops methodologies further streamlined container management and ensured stable service rollouts.
3. Centralized Authentication Handling:
Avoiding duplicate authentication implementations for each service, we introduced a universal login system and a shared library for JWT token verification. Kubernetes Service Mesh and Istio’s Authentication Policy were instrumental in enhancing token verification efficiency.
4. Simplified Service Failure Retries:
Handling failure retries in backend services often involves intricate logic. For projects using advanced languages like Kotlin, syntactic sugar simplifies retry patterns. Kubernetes Service Mesh, particularly with Istio, facilitated the definition of retry policies outside the application code, enhancing readability.
5. Efficient Management of System Properties and Feature Toggles:
Managing configurable variables for system properties and feature toggles can be challenging. Dockerization allowed us to leverage volumes for configuration management. In Kubernetes-based services, we efficiently organized properties and toggles as environment variables.
6. Effective Cloud Infrastructure Management:
Managing cloud infrastructures demands strategic tools. Initially using platform-specific technologies, we transitioned to Terraform as our primary infrastructure management tool. For those preferring imperative code, pulumi or terraform cdk proved to be viable alternatives.
7. Streamlining API and Contracts Tests Across Microservices:
Contract testing across different microservices can be complex. We opted for openAPI definitions to define contracts remotely, allowing both client and server to share the same contract. Customizing openAPI generator logic provided flexibility in test implementation.
8. Balancing Best Practices and Practicality:
Embracing every new best practice can lead to challenges. While the pursuit of innovation is essential, tearing down legacy systems constantly is not practical. Recognizing learning curves, efforts, and migration risks, we adopted a balanced approach, advancing step by step.
Wrap-Up
In the pursuit of a competitive and efficient backend workflow, addressing challenges with strategic solutions is paramount. By leveraging Dockerization, Kubernetes, Service Mesh, and other tools, we’ve navigated obstacles, enhancing our backend workflow and setting the stage for continued innovation. The key lies not just in adopting the latest practices but in striking a balance that aligns with the practical needs of our projects.