Community

Blog Events Webinars Tutorials Forum
  1. Blog
  2. Events
  3. Webinars
  4. Tutorials
  5. Forum
Create Account
×
Community Blog Cloud-Native and Application Management with Ease – Part 3: Architecture Principles

Cloud-Native and Application Management with Ease – Part 3: Architecture Principles

Part 3 of this 7-part series discusses the core principles of cloud-native architecture.

By Shantanu Kaushik

Cloud-native offers significant benefits over traditional architecture. Security, performance, scalability are the primary benefits. Cloud-native architecture has to follow some principles that transform a solution into a productive and stable one. In this article, we will discuss the principles followed by the cloud-native architecture and look at Alibaba Cloud’s products and services as examples.

Service-Oriented Architecture

Alibaba Cloud has worked towards building a strong and robust cloud-native solution that focuses on service-oriented architecture. This architecture provides a seamless solution for the development team when the codebase size is considerably large and used to build cloud-native applications.

This architecture splits the service modules by forming a logic layer to divide the services into microservices and mini-services with different lifecycle and management methodologies. Mini-services are a collection of services that share data. Mini-services are usually applied to call each other in local mode transmission since the microservices application can cause call losses due to remote calling. Microservices and mini-services can have multiple iterations with an accelerated iteration progress for extended stability.

The service-oriented architecture follows a fluent and synchronous feature-set, with code-reusability applied within the solution to achieve common functions. This architecture controls the workflow based on server traffic rather than purely network traffic. The architecture is built on a distributed computing platform. Some of the control policies that can be applied using service-oriented architecture are listed below:

  • Data Warehouse Isolation
  • Canary Release
  • Zero-Trust Security
  • Throttling
  • Degradation
  • Circuit Breaking

The extended feature-set using the service-oriented architecture standardizes the service traffic transmission by controlling the flow between different business modules at the architecture level. The correlations between different modules remain unaffected by the way they are deployed. Multiple instances can be deployed across the architecture for every API operation.

These instances can be scaled on-demand. Alibaba Cloud SOA applies policy control over business modules regardless of the programming language(s) used with the backend service. Service splitting increases the O&M workload but also enables the functionality to update each API operation individually.

Automation

Automation is the key to any successful architecture. Automation is a suggestive approach that eases most complex and repeated procedures, so human intervention doesn’t affect operation stability. This way, you can derive a perfect operation stream every time with consistency. The cloud-native architecture reduces the complexities of managing a distributed architecture by utilizing containers, microservices, DevOps tools, and other services.

The software delivery mechanism has a complex structure, but cloud-native can implement scenario-based solutions that utilize service structures based on the Open Application Model (OAM), GitOps, Kubernetes, Infrastructure as Code (IaC), and other tools to facilitate continuous integration and delivery (CI/CD). This enables a streamlined and automated process for application delivery for the enterprises panning environments.

Elasticity

Elasticity is a critical requirement for most businesses. Elasticity is often estimated based on business scale and network load requirements. Elasticity directly reflects the way businesses can scale the backend resources of the infrastructure that are being used, including compute service, storage, security, network, and database connections. The inherent data is collected based on usage patterns indicated by the cloud monitor or by using the Server Load Balancer (SLB) logs. Elasticity refers to the functionality of a system to scale itself as the business scale changes. It can scale up or down. Elasticity is highly useful for cost-saving, resource planning, strategic implementation, and accelerating deployment.

Availability

Availability is one of the most important points for any service or solution. Alibaba Cloud has deep integration among its products and services to enable an array of redundancies to provide maximum availability. There could be various reasons for service downtime. It could be hardware failure, software glitches, network issues, or an attack like DDoS.

Software availability defines the capacity of a solution to resist any possible downtime due to the reasons listed above. The cloud-native architecture enables high-availability scenarios to increase the mean time between failure (MTBF) in an application. When defining an architecture, the design team responds to the most influential design aspects to make the solution capable of delivering high-availability. Some of these are disaster recovery scenarios and active geo-redundancy.

Observability

Sometimes observability is confused with monitoring, application performance management, or business liveness detection. Observability is quite different. When observability is applied within a distributed environment (such as cloud computing), it counts on tracing, logging, and measurement to calculate factors (such as response time, parameters passed, and return values) for any deployed service calls initiated by a single application request.

Some questions need answers before setting up any scenarios for availability with a distributed setup. As the scale of an enterprise grows, the application usage also grows. At this point, if a failure occurs, the questions are, “Why? Which modules were affected? How much of the service is unavailable?”

Observability answers all of these questions for quick recovery and strategy formulation to avoid future outages. Observability monitoring enables O&M personnel to collect metrics in real-time and deliver analysis based on multi-dimensional data. This analysis is highly productive when coupled with real-time data from the Alibaba Cloud Monitor service and collected data from the Server Load Balancer (SLB).

Continued in Part 4

In Part 4 of this series on cloud-native, we will continue exploring cloud-native architecture and models. We will discuss how continuous evolution and zero-trust security enrich the overall architectural robustness of cloud-native architecture. Furthermore, we will deep-dive into Mesh Architecture, Event-Driven Architecture, Observable Architecture, Distributed Transaction Architecture, Storage and Computing Architecture, and Serverless Architecture.

Upcoming Articles

  1. Cloud-Native and Application Management with Ease – Part 4
  2. Cloud-Native and Application Management with Ease – Part 5
  3. Cloud-Native and Application Management with Ease – Part 6
  4. Cloud-Native and Application Management with Ease – Part 7
DevOps Microservices Trends Terraform Digital Transformation Kubernetes cloud native Application Management
0 0 0
Share on

Read previous post:

Cloud-Native and Application Management with Ease – Part 2

Read next post:

How to Learn Alibaba Cloud RDS

Alibaba Clouder

2,599 posts | 764 followers

You may also like

Comments

Alibaba Clouder

2,599 posts | 764 followers

Related Products

More Posts by Alibaba Clouder

See All