Google Cloud Digital Leader: Modern Apps, APIs and Portability

Describe the role of an API boundary and API-management proxy; evaluate container portability without assuming that state and permissions move automatically.

Objective: separate a boundary from its implementation

After selecting a compute shape, explain how application components communicate and what must remain stable when one component changes. An API is an interface through which a caller requests a capability. A useful teaching test is to identify the caller, the request, and the expected result before discussing the backend implementation. A stable interface can support a staged change, but stability is a design obligation. Merely exposing a network endpoint does not prove that old callers will still receive the behavior they expect.

API management adds a policy layer

Apigee is Google Cloud’s API management platform. Its documentation describes API proxies between clients and backend services, with controls such as security policies, rate limiting, quotas, and analytics. This layer can provide consistent management around backend interfaces. It does not automatically repair an incorrect pricing calculation or replace the application’s own data model. Distinguish a request-management rule from the business operation behind it. A request that obeys a rate limit can still contain invalid business data, which the application must handle appropriately.

Packaging helps but does not finish portability

GKE uses Kubernetes to orchestrate containerized applications, while Cloud Run can execute containerized code without requiring a customer-created cluster. Containers provide a packaging approach, but portability is broader than moving an image. Check runtime compatibility, network dependencies, identity permissions, configuration, persistent data, and the application’s operating assumptions. Cloud Run instances have disposable local file systems; an image that runs successfully can still lose an improperly stored record when the instance stops. An open orchestration technology is useful evidence about an interface, not proof that every managed feature is identical elsewhere.

Worked example

Fictional case: a wholesaler wants to replace a legacy stock-check backend while distributors continue calling the same interface. First record the current request and response behavior with representative examples, including unavailable stock. Second, investigate an API proxy to apply consistent access and traffic policies while routing to the backend. Third, test the replacement’s business results against the agreed interface. Fourth, inspect persistent data and permissions separately from container startup. The change is successful only if callers receive correct stock information and operations can manage the service, not merely because a new image was deployed.

Change the boundary and re-test callers

Suppose the replacement now returns quantities in cases rather than individual items. Keeping the old endpoint name does not preserve its meaning. Either preserve the original contract or deliberately introduce and communicate the new representation, with callers updated and tested. Similarly, a rate limit that protects a backend can reject legitimate callers if configured without their traffic needs. API management supports change control, but the business still owns compatibility decisions and exceptional cases. Portability claims should name which behavior was tested and which dependencies remain tied to the environment.

Transfer check

A team moves a container image to a second environment and finds that it cannot read the production database. What does this show? Check: image portability alone did not establish identity, connectivity, or database compatibility. Investigate the missing dependency rather than assuming that the image is necessarily defective. Use the cards to distinguish an API policy, a backend rule, and an environment dependency. Next, express the missing database access as an identity-and-resource question in the IAM lesson.

Try it, then explain it.

Complete a short set without looking at the example. After each answer, explain which evidence or rule supports your decision and why the alternatives do not fit.

Practice 4 questions

Sources & authorship

Original OpenExam explanations and examples. References support the concepts; they are not the source of copied questions. Independent educator review is pending.