ScienceSoft. Professional Software Development. ScienceSoft. Professional Software Development.
Testing and QA

Comprehensive Guide to Effective Microservices Testing

Effective Microservices Testing - ScienceSoft

ScienceSoft has been providing software testing services for 32 years and test automation services for 20 years.

Microservices Testing: The Essence

Microservices testing is a combination of QA activities to ensure each microservice’s functioning and performance is stable, its failure doesn’t result in severe functional disruptions of entire software, all microservices smoothly function as one application.

Microservices testing types: multi-level functional testing, regular regression testing, performance and security testing.

Microservices testing team: a QA manager, manual and test automation engineers, a test automation architect.

Microservices Testing Setup Plan

A microservices testing setup plan depends greatly on the current SDLC stage of your microservices project and the software requirements specifics.

Still, based on ScienceSoft’s portfolio of microservices testing projects, we can outline some universal steps to follow.

1. Microservices testing process design

Duration: Up to 1 week

  • A QA manager should be assigned to ensure the testability of your microservices software requirements.
  • A set of KPIs relevant for testing microservices should be drawn up. You can consider such KPIs as the number of defects found and test cases/scripts developed per iteration, test cases’ granularity, test scripts’ maintainability and robustness, etc.
  • A risk mitigation plan also should be designed to oversee the relevant risks of your microservices testing project.

Best practice: As every microservice is developed and deployed independently, ScienceSoft usually integrates a test engineer into each cross-functional microservice team and establishes collaboration scenarios between the test engineers for better coordination and prioritization of the testing activities.

2. Microservices testing preparation

Preparation for the microservices testing process differs based on the sourcing model you opt for: in-house testing or outsourced testing.

2.1. In-house microservices testing preparation

Duration: Up to 4 weeks

If you opt for an in-house microservices testing option, your QA manager should:

  • Prepare a test strategy, plan, and efforts estimation (typically, a test engineer is assigned to each microservice, still sometimes more test engineers may be required depending on the microservice architecture and functionality complexity).

Best practice: ScienceSoft’s QA consultants usually opt for test automation in a microservices testing project as the microservices-based architecture fits test automation well.

  • Assemble a test team, hire test automation talents experienced in microservices testing and consider additional training for test engineers.

While a test automation manager should:

  • Consider what test scenarios are repeatable, stable, and involve a high data volume to be automated first. Design a test automation architecture.
  • Choose and configure relevant tools and frameworks for automated testing.
  • Decide on service virtualization and API mocking tools essential for testing microservices.

NB! ScienceSoft’s QA consultants advise to check if the following software testing types vital for microservices testing are well-represented

Integration, component, contract testing

The testing team should develop and execute:

  • Integration test cases (validating communication paths and interactions between components within each microservice).
  • Component test cases (verifying a microservice behavior with the help of API mocking tools).
  • Contract test cases (checking the real communication between microservices).

These testing types should comprise the majority of microservices test suites.

It should include load, stress, endurance, and scalability testing at both:

  • Microservice level (when a microservice is deployed).
  • Application level (when all the microservices are deployed).

Regression testing

To avoid application breakdowns upon introducing new services or API evolution, a test engineer within each microservices team should verify how their microservice interacts with its dependencies.

Expert tip: ScienceSoft’s test automation managers strongly recommend automating the regression test suites for all microservices, as the number of regression test cases to be performed per iteration can be multiplied by the number of microservices. Besides, a comprehensive regression test suite for entire software should be run each time a new microservice is deployed.

2.2. Vendor selection for outsourced microservices testing

Duration: Up to 12 weeks

If you decide to outsource your microservices testing, during a testing vendor selection you should:

  • Consider QA vendors with experience in microservices testing or ample experience in test automation.
  • Look through the portfolio for microservices software testing or test automation projects.
  • Consider the testing vendor’s tool and resources stacks to meet your microservices application’s specifics and scale.
  • Short-list 3-5 vendors with relevant hands-on experience and resources.
  • Request microservices testing cost estimate and service-specific presentation from the shortlisted vendors. This way, you can get an understanding of the vendor’s approach to microservices test team lineup, the planned test automation involvement, testing toolkit, and therefore, decide on microservices testing ROI.
  • Decide on the most appropriate microservices testing vendor.
  • Negotiate a SLA and microservices testing costs, sign a collaboration contract.

3. Microservices testing launch

Duration: Up to 2 weeks

To launch the microservices testing process, either an outsourced or an in-house testing team should:

  • Choose and configure the test environment.
  • Start to develop test cases and test scripts.
  • Get down to manual or automated test data generation.

Consider Professional Microservices Testing Services

With 32 years in software testing and 20 years in test automation, ScienceSoft promptly grasps your microservices architecture specifics to ensure the solution’s quality at all levels.

Microservices testing consulting

  • The analysis of the microservices application architecture and requirements, the existing QA process.
  • QA strategy, microservices test plan, test automation strategy and architecture design.
  • The selection of tools and frameworks optimal for your microservices testing project.
  • Microservices testing cost and ROI calculation.
  • Optimal sourcing model selection.

Microservices testing outsourcing

  • QA strategy, microservices test plan, test automation architecture design and optimal testing tools and frameworks configuration.
  • Test environment setup and automated test data generation.
  • Test automation script development and maintenance.
  • Microservices test artifacts compliant with ISO/IEC/IEEE 29119-3:2013, including regular detailed defects and test closure reports.

What our customers say about ScienceSoft’s microservices testing

Collaborating with skilled IT vendors is crucial to adhere to our high work standards. Owing to ScienceSoft’s experience in distributed modular systems, we decided to cooperate with them. ScienceSoft quickly provided a complete team that included QA specialists. They ensured the software functions and performs stably even under heavy load, which helps provide high quality user experience for our customers. ScienceSoft proved to be a reliable tech partner.

Wadih Pazos, COO, Paramount WorkPlace

Team Required for Microservices Testing

The following roles are universal for microservices testing projects and typically present within ScienceSoft’s microservices testing teams.

QA manager

  • Manages the microservices testing process and team.
  • Checks that each microservices software requirement is clear, measurable, and complete.
  • Draws up a test strategy and plan based on business and project goals.
  • Selects test management software.

Test automation manager

  • Designs a test automation architecture.
  • Selects and configures test automation tools and frameworks.
  • Manages test automation engineers to improve maintainability and granularity of automated tests and decrease test execution time.

Test automation engineer

  • Sets up the test environment and test data generation.
  • Develops, executes, and maintains automated test scripts.
  • Reviews automatically generated defect reports.
  • Collaborates with other cross-functional team members to improve maintainability and granularity of test scripts.

Test engineer

  • Designs and maintains test cases.
  • Executes manual testing.
  • Reports defects via outlined tools.

Microservices Testing Sourcing Models

QA management and implementation are in-house

You have complete control over the microservices testing process. At the same time, there are risks bound to incorrect testing efforts estimation, the lack of test automation talents, an unbalanced test automation strategy, the irrelevant selection of testing tools and frameworks.

QA management is in-house; a test team is completely or partially external

You can optimize microservices testing costs by quickly scaling up and down the number of QA talents. Still, you need a well-versed QA manager to design the testing process, manage and regularly assess the microservices testing progress and the external QA team’s performance.

QA management and team are outsourced

You get well-established microservices testing with experienced QA management, test automation excellency, and microservices testing best practices. Still, you should be careful during the vendor selection process and ready to eliminate possible communication issues within microservices teams.

Microservices Testing Tools

As multi-level functional testing comprises the largest share of microservices testing, here are some functional testing tools recommended by ScienceSoft’s test engineers to streamline your microservices testing project.

Postman

Highlights

  • Gartner Peer Insights Customers’ Choice 2020 for full life cycle API management.
  • A comprehensive automated API functional testing tool.
  • REST, SOAP, HTTP API calls and responses testing.
  • Compatible with popular CI/CD tools (Jenkins, Travis, GitLab, etc.) to add testing into a new or existing CI/CD pipeline.
  • Light mocking capabilities without actual service virtualization, thus inappropriate for component microservices testing.

best for

Automated integration, contract, end-to-end functional testing.

Pricing

Open-source.

SoapUI

HIGHLIGHTS

  • Belongs to SmartBear family – 2019 Gartner leader for software test automation.
  • All-encompassing automated API functional testing.
  • Support for HTTP, SOAP, RESTful testing.
  • SOAP, REST Mock Service, and service virtualization (with ReadyAPI) to enable API mocking and service virtualization.

BEST FOR

All-encompassing automated functional testing empowered by API mocking and additional service virtualization tools.

Pricing

  • SoapUI - open-source.
  • ReadyAPI Virtualization – $1,378/user/year.

Apache JMeter

Highlights

  • Tool for testing software behavior and performance under varying continuous loads.
  • Supports functional test suites via features like HTTPS sampler.
  • User-friendly GUI.
  • 71 customizable plugins to extend core capabilities.

Best for

Automated load, stress, and functional testing.

Pricing

Open-source.

Microservices Testing Costs

Each microservices testing project is as different as each microservices solution. Let’s take as an example a microservices project with the following characteristics:

  • 10-month long.
  • 2-week iterations with testing going in parallel.
  • 100 new test cases per iteration.
  • 80% coverage by automated tests.

The rough estimate for such testing project including senior test engineers’ rates, testing tools’ licenses and test environment costs, will be $50,000.

Microservices testing cost factors include:

For an outsourced option

  • The number of iterations.
  • The number of new test cases per iteration.
  • The percentage of automation test coverage.
  • Test cases/script development and maintenance efforts (complex end-to-end tests require much more build and maintenance time).
  • Test team members’ rates.

For an in-house option

  • The number of microservices and the pace of their deployment.
  • The complexity of each microservice’s functionality.
  • The number of required third-party integrations.
  • Microservices performance requirements.
  • The number of test engineers and test automation engineers.
  • The cost of employed tools (for example, licenses and test servers).

Get to Know Your Microservices Testing Cost!

ScienceSoft will carry out microservices testing cost calculation and ROI analysis for you to knowingly decide on a testing budget and lay the base for microservices testing planning.

About ScienceSoft

ScienceSoft is a global IT consulting, software development, and QA company headquartered in McKinney, TX, US. We deliver test automation services for microservices testing projects and guide our customers on how to ensure maximum quality of microservices-based software, balance testing budget, and reduce testing time. Being ISO 9001 and ISO 27001 certified, we rely on a mature quality management system and guarantee that cooperation with us does not pose any risks to our customers’ data security.