Service Virtualization Use Instances for Cloud-Native Purposes

That is the second article in a two-part sequence. The primary half is right here.

Service virtualization is uniquely suited to help the wants of cloud-native functions. The answer’s rules help all the important thing attributes of cloud-native architectures. I’d contend that service virtualization and cloud-native architectures are constructed for each other. Simply as service virtualization is thought-about important for agile improvement, it’s simply as mandatory for supporting cloud-native functions. 

On this part, we’ll focus on how Broadcom Service Virtualization helps every of the ten key attributes of cloud-native utility techniques. 

Steady Microservices Utility Improvement and Testing 

One of many key attributes of cloud-native functions is the usage of loosely coupled parts, equivalent to microservices. As a result of component-based techniques have quite a few dependencies on one another, use of service virtualization is particularly essential to help the agile improvement, testing, and steady supply of such functions. This is without doubt one of the hottest use circumstances for service virtualization. See determine under. 

 

Service virtualization can be utilized to help each side of microservices testing, together with unit, element, integration, contract, and system testing. Plus, testing will be achieved constantly all through the CI/CD pipeline. For extra particulars on utilizing service virtualization for component-based functions, please seek advice from my earlier blogs (Steady Service Virtualization Half 1 and Half 2). 

The next sections spotlight a number of key use circumstances:

a. Era of Artificial Digital Providers from Microservice APIs

Service virtualization can be utilized to generate an artificial digital service for a dependent element that doesn’t but exist, or is unavailable, utilizing artificial request-response pairs primarily based on the API’s definition of the service.  That is particularly helpful throughout improvement and unit testing. 

The artificial digital service could also be progressively enhanced to help subsequent testing (equivalent to integration and system assessments) alongside the CI/CD lifecycle.

b. Assist for Microservices Contract Testing

A service client can check a service supplier utilizing artificial request-response pairs developed from its API specification. A service supplier can in flip check its interactions with a client utilizing a validated contract—which will be applied utilizing a recorded digital service. See determine under. 

c. Assist for Steady Microservices Efficiency Testing

This is without doubt one of the most essential use circumstances for service virtualization. With service virtualization, we will really shift microservices efficiency testing left. This considerably reduces the necessity for time consuming, end-to-end load assessments that need to be carried out earlier than launch. Service virtualization allows restricted, scaled efficiency testing on the element stage to validate related service stage targets (SLOs). This additionally allows scaled testing of transactions throughout a number of providers utilizing their APIs. For extra particulars on this, please seek advice from my prior weblog on steady efficiency of microservices.

d.  Assist for Simpler Chaos and Unfavourable Testing

Digital providers present a method to help repeatable and structured chaos and detrimental situations testing. For instance, they will simulate such situations as non-responsiveness, downtime, or gradual response time. That is a lot simpler and much much less time consuming than having to energy off servers or take down bodily computing cases.

e. Assist for Steady Reliability Engineering

Use circumstances (c) and (d) above permit us to constantly validate the reliability of cloud-based functions by making use of the rules of steady testing on the element stage. On this means, we will check each element change early within the lifecycle to see if its SLOs are met. With service virtualization, we will simulate dependent parts together with their SLOs (or SLIs if SLOs usually are not obtainable). For extra data, please seek advice from my weblog on Steady Reliability.   

Assist for Self-Service, Digital, Shared, and Elastic Infrastructure 

That is the second key attribute for cloud-native functions. As mentioned in part III, groups can use digital providers as a stand-in for actual parts. Digital providers can simply be packaged into light-weight container-based VSEs, which will be deployed on-demand into ephemeral cloud environments and so they can scale routinely as wanted. 

The truth is, libraries of digital providers might themselves be provided as a service (see “Service Virtualization as a Service”)—with all of the capabilities of a cloud-native service—in order that they might ruled and consumed throughout a number of groups and functions. 

Assist for Isolation from Server and Operation System Dependencies 

Digital providers will be packaged into containers that could be ported throughout a wide range of pc environments, no matter the place the true finish factors are hosted. Plus, they are often deployed throughout a number of hybrid computing environments with several types of {hardware} and OS. This not solely contains cloud-native functions, however legacy techniques (equivalent to mainframes) and different advanced techniques that these functions have to work together with. 

One use case of this attribute is service virtualization’s help for testing cloud-native functions with function-as-a-service (FaaS) dependencies. Usually talking, FaaS implementations usually are not very transportable throughout a number of cloud suppliers. This makes it troublesome to include FaaS code into testing environments which will reside in a special cloud. Service virtualization can be utilized to virtualize the FaaS element so it may be deployed into an area testing surroundings with routinely scaling VSEs. This permits groups to simulate the FaaS conduct for an utility beneath check that depends upon a FaaS endpoint. 

Assist for Unbiased Lifecycle Administration Utilizing Agile/DevOps 

Assist for unbiased life cycle administration is without doubt one of the key use circumstances for service virtualization. (See my earlier weblog for extra on this matter.) Service virtualization helps to optimize steady supply. The truth is, digital providers are key to supporting the principal of getting a devoted CI/CD pipeline for every microservice. 

Since microservices have dependencies on one another, because of this particular person CI/CD pipelines could also be impeded when a dependent microservice shouldn’t be obtainable or is present process parallel improvement. Digital providers assist to take away these dependencies between parallel pipelines.  Digital providers will be made obtainable to different providers as wanted. See determine under. 

Notice: This precept can be utilized to help simpler canary testing. Cloud-native functions are designed to permit frequent micro releases, for instance, of a single element. This enables us to get quick suggestions by deploying solely the modified element right into a canary surroundings. Service virtualization makes this a lot simpler and less expensive by virtualizing the remainder of the applying ecosystem, enabling groups to deal with the conduct of solely the modified element.  

Service virtualization helps a wealthy set of APIs that can be utilized to combine with CI/CD and different instruments for automation of deployment, orchestration, updates, and extra.

Assist for Light-weight Containers 

Digital providers can simply be packaged into lightweight containers. As well as, VSEs will be deployed on-demand into ephemeral cloud environments, and scale routinely as required  in Kubernetes clusters. 

Assist for Finest-of-Breed Languages and Frameworks

Digital providers are constructed on the protocol stage, and due to this fact are typically in a position to help functions, regardless of the programming language they’re developed in. This enables us to construct digital providers for functions which were created with a broad vary of languages and platforms. 

For builders, digital providers could also be even be developed as code and this strategy is supported by many programming languages.  

Assist for API-Based mostly Interplay and Collaboration 

As mentioned earlier than, service virtualization permits us to create artificial digital providers from service API specs. It additionally helps intensive API-based testing (together with what’s known as “headless” testing) throughout layers of APIs sometimes utilized in cloud-based functions. 

As well as, digital providers don’t simply help easy API protocols like REST. Digital providers can help a wide range of API varieties (equivalent to gRPC) throughout a number of kinds of utility techniques. 

The determine under reveals a standard API structure that has expertise, course of, and system APIs. Service virtualization can be utilized to not solely virtualize every of the sub-services that help the API, however the whole API layer by virtualizing the “nearest neighbor.” 

For instance, we will run user-experience assessments on front-end gadgets by virtualizing the “omni-channel” API—with out having to arrange a check surroundings for the all of the difficult stacks under it! 

As well as, service virtualization helps a wealthy set of APIs that can be utilized to combine with CI/CD and different instruments for automation of deployment, orchestration, updates, and extra.

Service virtualization can be built-in with API Gateways to permit clear entry to API back-end providers—whether or not applied by an actual service or a digital service. See determine under. 

Assist for Stateless and Stateful Providers

Digital providers will be simply created to be stand-ins for stateless providers by simulating their conduct. 

When we have to virtualize a stateful service, we will create digital providers which are supported by intensive check information that represents a sub-set of the stateful service’s information. We are able to achieve this by integrating with a check information administration system. For extra on the interplay between digital providers and check information within the context of microservices, please seek advice from my prior weblog on steady check information administration. 

Assist for Automation and Infrastructure-as-Code

Digital providers are extremely amenable to automated deployment (for instance in CI/CD pipelines), particularly when packaged as containers. These providers will be outlined as a part of infrastructure-as-code surroundings recipes, equivalent to Helm Charts, for automated provisioning and deployment.  

Assist for Governance Fashions

Digital providers are sometimes deployed into Digital Service Environments. This enables us to outline governance insurance policies of digital providers that mimic these of corresponding functions. 

Abstract and Conclusion 

We’ve got examined how Broadcom Service Virtualization helps all kinds of use circumstances for cloud computing—from cloud migration to help for cloud-native computing. 

Our view is that service virtualization and cloud capabilities complement one another. By combining  Service virtualization and cloud providers, groups can set up a stage of really agile utility improvement and supply, which might merely not be attainable with solely one among these capabilities by itself. The truth is, groups want to make use of service virtualization to help the necessities of cloud-native techniques. 

—