NGINX Full Version

微服务参考模型（MRA），第 5 部分：《十二要素应用》的微服务实践

Author’s note – This blog post is the fifth in a series:

  1. Introducing the Microservices Reference Architecture from NGINX
  2. MRA, Part 2: The Proxy Model
  3. MRA, Part 3: The Router Mesh Model
  4. MRA, Part 4: The Fabric Model
  5. MRA, Part 5: Adapting the Twelve‑Factor App for Microservices (this post)
  6. MRA, Part 6: Implementing the Circuit Breaker Pattern with NGINX Plus

All six blogs, plus a blog about web frontends for microservices applications, have been collected into a free ebook.

Also check out these other NGINX resources about microservices:

The Twelve‑Factor App

Software is increasingly delivered over the Internet as a service. Originally called Software as a Service (SaaS), similar software – with a much stronger emphasis on mobile interaction – is now usually referred to as web apps.

The Twelve‑Factor App is a praiseworthy effort by Heroku, a platform as a service (PaaS) provider, to establish general principles for creating useful web apps. However, the original principles are somewhat specific to Heroku’s PaaS platform. They aren’t an exact fit for a microservices architecture.

In implementing the NGINX MRA, we’ve extended the Twelve‑Factor App with our own additions and microservices‑specific modifications. We’ve found the amended version extremely useful.

For instance, the Twelve‑Factor App specifies that configuration code be stored in environment variables, rather than in configuration files. This is an extremely useful principle for the MRA, which has three different models, with the only difference between them being their configuration code and the number of NGINX Plus servers they use.

Building on what we’ve learned, we in the Microservices practice here at NGINX have developed the following set of principles. Our principles adapt the core ideas in The Twelve‑Factor App to a general‑purpose microservices architecture that is optimized for continuous delivery.

We invite you to use these principles in the development of your own apps. We’ve kept repetition from that document to a minimum.

The Twelve Factors Applied to Microservices

1 – Codebase

One codebase per service, tracked in revision control; many deploys

The Twelve‑Factor App recommends one codebase per app. In a microservices architecture, the correct approach is actually one codebase per service. Additionally, we strongly recommend the use of Git as a repository, because of its rich feature set and enormous ecosystem. GitHub has become the default Git hosting platform in the open source community, but there are many other excellent Git hosting options, depending on the needs of your organization.

2 – Dependencies

Explicitly declare and isolate dependencies

As suggested in The Twelve‑Factor App, regardless of what platform your application is running on, use the dependency manager included with your language or framework. How you install operating system or platform dependencies depends on the platform:

Note: We recommend that you choose a dependency management mechanism in the context of your comprehensive Infrastructure‑as‑Code strategy, not as an isolated decision. See Martin Fowler’s writings on Infrastructure‑as‑Code and download the O’Reilly report Infrastructure as Code by Kief Morris.

3 – Config

Store configuration in the environment

Anything that varies between deployments can be considered configuration. The Twelve‑Factor App guidelines recommend storing all configuration in the environment, rather than committing it to the repository. We recommend the following specific practices:

4 – Backing Services

Treat backing services as attached resources

The Twelve‑Factor App guidelines define a backing service as “any service the app consumes over the network as part of its normal operation.” The implication for microservices is that anything external to a service is treated as an attached resource, including other services. This ensures that every service is completely portable and loosely coupled to the other resources in the system. Additionally, the strict separation increases flexibility during development – developers only need to run the service(s) they are modifying, not others.

5 – Build, Release, Run

Strictly separate build and run stages

To support strict separation of build, release, and run stages, as recommended by The Twelve‑Factor App, we recommend the use of a continuous integration/continuous delivery (CI/CD) tool to automate builds. Docker images make it easy to separate the build and run stages. Ideally, images are created from every commit and treated as deployment artifacts.

6 – Processes

Execute the app in one or more stateless processes

For microservices, the important point in the Processes factor is that your application needs to be stateless. This makes it easy to scale a service horizontally by simply adding more instances of that service. Store any stateful data, or data that needs to be shared between instances, in a backing service.

7 – Data Isolation

Each service manages its own data

As a modification to make the Port binding factor more useful for microservices, we recommend that you allow access to thepersistent data owned by a service only via the service’s API. This prevents implicit service contracts between microservices and ensures that microservices can’t become tightly coupled. Data isolation also allows the developer to choose, for each service, the type of data store that best suits its needs.

8 – Concurrency

Scale out via the process model

The Unix process model is largely a predecessor to a true microservices architecture, insofar as it allows specialization and resource sharing for different tasks within a monolithic application. In a microservices architecture, you can horizontally scale each service independently, to the extent supported by the underlying infrastructure. With containerized services, you further get the concurrency recommended in the Twelve‑Factor App, for free.

9 – Disposability

Maximize robustness with fast startup and graceful shutdown

Instances of a service need to be disposable so they can be started, stopped, and redeployed quickly, and with no loss of data. Services deployed in Docker containers satisfy this requirement automatically, as it’s an inherent feature of containers that they can be stopped and started instantly. Storing state or session data in queues or other backing services ensures that a request is handled seamlessly in the event of a container crash. We are also proponents of using a backing store to support crash‑only design.

10 – Dev/Prod Parity

Keep development, staging, and production as similar as possible

Keep all of your environments – development, staging, production, and so on – as identical as possible, to reduce the risk that bugs show up only in some environments. To support this principle, we recommend, again, the use of containers – a very powerful tool here, as they enable you to run exactly the same execution environment all the way from local development through production. Keep in mind, however, that differences in the underlying data can still cause differences at runtime.

11 – Logs

Treat logs as event streams

Instead of including code in a microservice for routing or storing logs, use one of the many good log‑management solutions on the market, several of which are listed in the Twelve‑Factor App. Further, deciding how you work with logs needs to be part of a larger APM and/or PaaS strategy.

12 – Admin Processes

Run admin and management tasks as one‑off processes

In a production environment, run administrative and maintenance tasks separately from the app. Containers make this very easy, as you can spin up a container just to run a task and then shut it down.

Conclusion

Use the Twelve‑Factor App and these additional principles to help you create robust microservices‑based apps that are optimized for continuous delivery. In addition, the MRA is like a cheat code to help you go further, faster, than if you had to start from scratch.

To try NGINX Plus, start your free 30-day trial today or contact us to discuss your use cases.