Application development and deployment
Examine patterns and practices of application development, configure Azure Pipelines, and implement site reliability engineering (SRE) best practices. Site reliability engineering (SRE) is a software engineering approach to application development and deployment, change management, monitoring, emergency response, and more.
Plan, train, and proof
As you get started, use the checklist and application development resources in the following sections to help you plan your application development and deployment. You should be able to answer these questions:
- Have you prepared your application development environment and setup workflow?
- How will you structure the project folder to support Kubernetes application development?
- Have you identified state, configuration, and storage requirements of your application?
Site reliability engineering checklist:
Prepare your development environment. Configure your environment with the tools you need to create containers and set up your development workflow.
For more information, see:
Containerize your application. Familiarize yourself with the end-to-end Kubernetes development experience, including application scaffolding, inner-loop workflows, application-management frameworks, CI/CD pipelines, log aggregation, monitoring, and application metrics.
To learn more, see:
Review common Kubernetes scenarios. Although Kubernetes is often thought of as a platform for delivering microservices, it's becoming a much broader platform. For more information about common Kubernetes scenarios such as batch analytics and workflow, see Overview of common Kubernetes scenarios (video).
Prepare your application for Kubernetes. Prepare your application file system layout for Kubernetes and organize for weekly or daily releases. Learn how the Kubernetes deployment process enables reliable, zero-downtime upgrades.
For more information, see:
Manage application storage. Understand the performance needs and access methods for pods so that you can provide the appropriate storage options. Also, plan for ways to back up and test the restore process for attached storage.
To learn more, see:
Manage application secrets. Use a key vault to store and retrieve keys and credentials. Don't store credentials in your application code.
For more information, see:
Deploy to production and apply best practices
As you prepare the application for production, use the following checklist. You should be able to answer these questions:
- Can you monitor all aspects of your application?
- Have you defined resource requirements for your application? How about scaling requirements?
- Can you deploy new versions of the application without affecting production systems?
SRE best practices checklist:
Configure readiness and liveness health checks. Kubernetes uses readiness and liveness checks to know when your application is ready to receive traffic and when it needs to be restarted. Without defining such checks, Kubernetes won't be able to determine whether your application is running. For more information, see Liveness and readiness checks.
Configure logging, application monitoring, and alerting. Monitoring your containers is critical, especially when you're running a production cluster, at scale, with multiple applications. The recommended logging method for containerized applications is to write to the standard output (
stdout) and standard error (stderr) streams.For more information, see:
Define resource requirements for the application. A primary way to manage the compute resources within a Kubernetes cluster is to use pod requests and limits. These requests and limits tell the Kubernetes scheduler what compute resources a pod should be assigned. For more information, see Define pod resource requests and limits.
Configure application scaling requirements. Kubernetes supports horizontal pod autoscaling to adjust the number of pods in a deployment depending on CPU utilization or other select metrics. To use the autoscaler, all containers in your pods must have CPU requests and limits defined. To learn more, see Configure horizontal pod autoscaling.
Deploy applications using an automated pipeline and DevOps. The full automation of all steps between code commit to production deployment allows teams to focus on building code and removes the overhead and potential human error in manual steps. Deploying new code is quicker and less risky, helping teams become more agile, more productive, and more confident about their running code.
To learn more, see:
Optimize and scale
Now that the application is in production, how can you optimize your workflow and prepare your application and team to scale? Use the optimization and scaling checklist to prepare. You should be able to answer these questions:
- Are cross-cutting application concerns abstracted from your application?
- Are you able to maintain system and application reliability, while still iterating on new features and versions?
Application deployment checklist:
Deploy an API gateway. An API gateway serves as an entry point to microservices, decouples clients from your microservices, adds another layer of security, and decreases the complexity of your microservices by removing the burden of handling cross-cutting concerns. For more information, see Use Azure API Management with microservices deployed in Azure Kubernetes Service.
Deploy a service mesh. A service mesh provides capabilities like traffic management, resiliency, policy, security, strong identity, and observability to your workloads. Your application is decoupled from these operational capabilities and the service mesh moves them out of the application layer and down to the infrastructure layer.
For more information, see:
Implement site reliability engineering (SRE) practices. Site reliability engineering (SRE) is a proven approach to maintain crucial system and application reliability while iterating at the speed demanded by the marketplace.
To learn more, see:
Feedback
Submit and view feedback for