I85PREP
⌘ K
Informatica CAI R42 Professional study materials
Informatica CAI R42 Professional study resources

Informatica CAI R42 Professional practice questions

One question at a time. Build your Informatica CAI R42 Professional understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Cloud Application Integration (CAI) Implementation

In Cloud Application Integration, which asset type represents a reusable, versioned unit of business logic that can be invoked from a process, a service connector, or another process via a Service step?

  1. A connection
  2. A guideline
  3. A business service Correct answer.
  4. A deployment package

Explanation

A business service is the reusable, independently versioned unit of logic that other processes and service connectors call through a Service step. Guidelines are design-time rule definitions, connections hold endpoint credentials, and deployment packages are release containers, so none of those can be invoked as logic.

Question 2Cloud Application Integration (CAI) Implementation

An integration developer needs a process to pause and wait for an external approval event that may arrive hours later, without consuming a thread while waiting. Which CAI construct is designed for this?

  1. A parallel path inside a loop
  2. A decision step with a timer branch
  3. A subprocess call with a retry policy
  4. A Receive step bound to a wait-and-notify event Correct answer.

Explanation

Receive steps suspend the process until a matching event arrives, and the process state is persisted so no thread is held during the wait. Loops, decision branches, and subprocess retries all keep the process executing rather than parking it for an asynchronous event.

Question 3Cloud Application Integration (CAI) Implementation

Which statement accurately describes the role of a Service Connector in a CAI process?

  1. It compiles process definitions into deployable Java artifacts.
  2. It exposes a process as an API endpoint for external callers.
  3. It defines the typed request and response contract used when a process calls a business service or app connection. Correct answer.
  4. It stores the runtime logs generated by process executions.

Explanation

A Service Connector is the design-time contract that declares input and output fields for a call, decoupling the calling process from the implementation. Endpoint exposure is handled by a service or API asset, logs are kept in execution monitoring, and compilation is not a connector concern.

Question 4Cloud Application Integration (CAI) Implementation

A process must call an external REST endpoint that requires an OAuth 2.0 client-credentials token. Where should the client ID and secret be configured so they are not hard-coded in the process logic?

  1. In a temporary variable declared inside the process
  2. In the process's fault handler
  3. In the connection associated with the app connection Correct answer.
  4. In a process input parameter default value

Explanation

Credentials such as OAuth client IDs and secrets belong in the connection definition, which is secured and environment-specific. Input defaults, process variables, and fault handlers are visible in the process logic and are not appropriate secret stores.

Question 5Cloud Application Integration (CAI) Implementation

When designing a process that must handle a failure from a Service step and continue with a compensating action, which pattern should be used?

  1. Enable automatic retry on the Service step and ignore the error
  2. Wrap the Service step in a scope with a fault handler that performs the compensation Correct answer.
  3. Convert the Service step into a Receive step
  4. Place the Service step on a parallel path with no join

Explanation

A scope with an attached fault handler lets the process catch the step failure, run compensating logic, and then decide whether to continue or rethrow. Retry alone does not provide compensation, an unjoined parallel path cannot guarantee ordering, and Receive steps wait for events rather than call services.

Question 6Cloud Application Integration (CAI) Implementation

Which CAI capability allows a process to transform an incoming JSON payload into a canonical XML structure before calling a downstream service?

  1. A Data Transformation step using a mapping Correct answer.
  2. An Assignment step copying a field
  3. A Notification step with a template
  4. A Decision step using an expression

Explanation

The Data Transformation step applies a mapping between source and target structures, including JSON-to-XML conversion with expressions and functions. Decision steps branch on conditions, Assignment copies values without structural mapping, and Notification steps send messages.

Question 7Cloud Application Integration (CAI) Implementation

A developer wants to reuse the same error-handling logic across many processes. Which approach best supports this in CAI?

  1. Disable fault handling and let executions fail silently
  2. Copy the fault handler into every process manually
  3. Define the error handling in a shared business service and invoke it from fault handlers Correct answer.
  4. Rely on the runtime to retry every failed step indefinitely

Explanation

Encapsulating error handling in a business service makes the logic reusable and versioned, and fault handlers can call it. Manual copying creates drift, indefinite retries are not a default behavior, and disabling fault handling removes observability.

Question 8Cloud Application Integration (CAI) Implementation

Which statement best describes how a human task step differs from a Service step in a CAI process?

  1. A human task step can only be used inside a fault handler.
  2. A human task step calls an external system, while a Service step waits for a person.
  3. A human task step always executes asynchronously without persisting process state.
  4. A human task step creates a work item for a user or group and suspends the process until it is completed. Correct answer.

Explanation

Human task steps generate a task assigned to a user or group and persist the process until the task is completed or expired. Service steps call automated endpoints, so the reversed description, the asynchrony claim, and the fault-handler restriction are all inaccurate.

Question 9Cloud Application Integration (CAI) Implementation

An integration must be triggered on a recurring schedule every night at 02:00. Which CAI mechanism should be configured?

  1. A fault handler with a timer
  2. A guided process started by a schedule trigger Correct answer.
  3. A service connector invoked by an external caller
  4. A human task with a due date

Explanation

A schedule trigger starts a process automatically on a defined cadence, which suits nightly batch windows. Service connectors are call contracts, fault-handler timers manage error timing, and human task due dates govern task expiry rather than process initiation.

Question 10Cloud Application Integration (CAI) Implementation

A process receives a list of order records and must call a fulfillment service once per record. Which step configuration achieves this?

  1. A single Service step with the whole list as input
  2. A Decision step with one branch per record
  3. A Data Transformation step applied to the list
  4. A Service step inside a loop that iterates over the collection Correct answer.

Explanation

Iterating over the collection with a loop and invoking the Service step inside it produces one call per record. Passing the whole list assumes a batch API, decision branches do not scale to variable record counts, and transformation does not invoke services.

Question 11Cloud Application Integration (CAI) Implementation

Which CAI feature lets a developer inspect the field-level values passed between steps during a specific execution for debugging?

  1. The asset dependency graph
  2. The execution details view with step-level input and output Correct answer.
  3. The connection configuration page
  4. The process design canvas

Explanation

Execution monitoring exposes each step's recorded inputs and outputs for a given run, which is how field-level values are inspected. The design canvas shows structure only, connection pages hold endpoint settings, and dependency graphs show asset relationships.

Question 12Cloud Application Integration (CAI) Implementation

A developer needs a process to branch based on whether an order total exceeds a credit limit stored in a process variable. Which step and configuration is appropriate?

  1. A Decision step with a condition comparing the total field to the limit variable Correct answer.
  2. A Receive step waiting for a credit event
  3. An Assignment step that overwrites the total
  4. A Data Transformation step with a filter function

Explanation

Decision steps evaluate boolean conditions and route execution down the matching branch, which fits comparing a runtime field to a variable. Transformation, assignment, and receive steps do not provide conditional routing.

Question 13Cloud Application Integration (CAI) Implementation

When a CAI process is deployed to a runtime environment, which statement about process versions is accurate?

  1. Each deployment references a specific process version, and new versions are created for changes Correct answer.
  2. Processes have no versioning and are overwritten on every save
  3. A deployed version can be edited in place without affecting running instances
  4. Only business services are versioned; processes are not

Explanation

Processes are versioned assets, and a deployment targets a specific version so running instances are not disturbed by later changes. In-place editing of a deployed version, overwrite-on-save, and process-level exemption from versioning all contradict the versioning model.

Question 14Cloud Application Integration (CAI) Implementation

An outbound REST call intermittently returns HTTP 503. The developer wants the process to retry a limited number of times with a delay before failing. Which configuration supports this?

  1. An infinite loop around the Service step
  2. A fault handler that restarts the entire process
  3. A decision step that always selects the retry branch
  4. A retry policy on the Service step with a bounded count and interval Correct answer.

Explanation

A bounded retry policy with a delay handles transient 503 responses without unbounded execution. An infinite loop risks runaway calls, restarting the whole process repeats completed work, and an always-retry decision branch never terminates.

Question 15Cloud Application Integration (CAI) Implementation

Which statement about using a temporary variable versus a process variable in a CAI process is correct?

  1. Temporary variables are stored in the connection definition
  2. Temporary variables are scoped to their containing step or scope, while process variables are visible throughout the process Correct answer.
  3. Temporary variables persist across process restarts, while process variables do not
  4. Temporary variables are visible to external callers as API fields

Explanation

Temporary variables exist within the scope where they are declared, whereas process variables are available across the process. Persistence across restarts, external API exposure, and storage in connections are not properties of temporary variables.

Question 16Cloud Application Integration (CAI) Implementation

A developer must ensure that a process only proceeds when an incoming payload satisfies a schema with required fields. Which CAI mechanism enforces this before processing begins?

  1. A parallel path that skips validation
  2. A notification step at the end of the process
  3. A loop over the payload fields
  4. A guideline applied to the process input Correct answer.

Explanation

Guidelines define validation rules that can be applied to inputs so invalid payloads are rejected before logic runs. Notifications occur after processing, loops do not inherently enforce required fields, and skipping validation defeats the requirement.

Question 17Cloud Application Integration (CAI) Implementation

Which CAI asset would you create to expose a process as a REST endpoint that external applications can invoke over HTTPS?

  1. A guideline
  2. A service asset with an API definition Correct answer.
  3. A business service
  4. A service connector

Explanation

A service asset publishes a process behind a defined API contract so external callers can invoke it over HTTPS. Business services are internal reusable logic, service connectors are call contracts, and guidelines are validation rules.

Question 18Cloud Application Integration (CAI) Implementation

A process must send an email to an operations group when a critical step fails. Which step type is intended for this?

  1. A human task step
  2. A Receive step
  3. A Data Transformation step
  4. A Notification step configured with recipients and a message template Correct answer.

Explanation

Notification steps send messages to configured recipients using a template, which fits alerting on failure. Transformation restructures data, Receive waits for events, and human tasks assign work items rather than broadcast alerts.

Question 19Cloud Application Integration (CAI) Implementation

Which consideration is most important when choosing between a synchronous and an asynchronous invocation pattern for a long-running CAI process exposed as an API?

  1. The version number of the business service
  2. Whether the caller can tolerate waiting for the full execution duration or needs an immediate acknowledgment Correct answer.
  3. The color scheme of the process canvas
  4. The number of variables declared in the process

Explanation

Synchronous invocation holds the caller until completion, so long-running work risks timeouts; asynchronous invocation returns an acknowledgment and lets the caller poll or receive a callback. Canvas appearance, variable counts, and version numbers do not determine the invocation pattern.

Question 20Cloud Application Integration (CAI) Implementation

A developer needs to map a repeating JSON array to a repeating XML element, preserving order and applying a currency conversion to each amount. Which approach is correct?

  1. Use a Data Transformation mapping with a repeating group and an expression for the conversion Correct answer.
  2. Use a Decision step to select the first element
  3. Use an Assignment step for each array element
  4. Use a Notification step with the array serialized

Explanation

A mapping with a repeating group iterates the source array into repeated target elements, and an expression on the amount field applies the conversion per item. Assignment cannot iterate, Decision selects branches, and Notification does not transform structures.

Question 21Cloud Application Integration (CAI) Implementation

Which statement correctly describes the relationship between a connection and an app connection in CAI?

  1. Connections are only used by human tasks
  2. Connections and app connections are identical and interchangeable
  3. A connection stores runtime credentials and endpoint details, while an app connection defines the metadata and operations for a target application Correct answer.
  4. An app connection stores credentials, while a connection defines operations

Explanation

The app connection supplies the operation metadata and interface for a target system, and the connection binds that metadata to a specific endpoint with secured credentials. The reversed description, the interchangeability claim, and the human-task restriction are all incorrect.

Question 22Cloud Application Integration (CAI) Implementation

A process must guarantee that either all of two related updates succeed or neither is committed. Which design consideration is most relevant in CAI?

  1. Use two independent Service steps with no compensation
  2. Use a Notification step to announce the updates
  3. Use a single Service step that performs both updates transactionally on the target system Correct answer.
  4. Use a parallel path so both calls run simultaneously

Explanation

CAI processes are not distributed transactions, so atomicity must be delegated to a single transactional call on the target system or handled with explicit compensation. Independent calls, parallel execution, and notifications do not provide all-or-nothing semantics.

Question 23Cloud Application Integration (CAI) Implementation

Which CAI capability should be used to define a process that a business user can start from a guided experience with a form-driven interface?

  1. A guided process with user input steps Correct answer.
  2. A scheduled service connector
  3. A data transformation mapping
  4. A fault handler scope

Explanation

Guided processes present form-driven user input steps so business users can start and interact with a process. Scheduled connectors, fault scopes, and transformation mappings are backend constructs without a guided user interface.

Question 24Cloud Application Integration (CAI) Implementation

During testing, a process fails because a mapped output field is null even though the source field is populated. Which troubleshooting step is most direct?

  1. Increase the retry count on the Service step
  2. Convert the process to a guided process
  3. Recreate the connection from scratch
  4. Check the mapping expression and the execution details to confirm the source value and target path Correct answer.

Explanation

Null outputs usually stem from a mismatched source path or an expression that evaluates to null, so reviewing the mapping alongside recorded step inputs and outputs isolates the cause. Recreating connections, changing retries, or changing process type does not address mapping defects.

Question 25Cloud Application Integration (CAI) Implementation

Which statement about fault handling in a parallel path is accurate?

  1. Parallel branches cannot raise faults
  2. Fault handling applies only to human task steps
  3. A fault handler attached to the containing scope can catch the fault and define compensating behavior for the branches Correct answer.
  4. A fault raised on one branch always terminates all branches immediately with no handler involvement

Explanation

Attaching a fault handler to the scope that contains the parallel paths allows the process to catch branch faults and run compensation or cleanup. Faults are not automatically fatal without handler involvement, branches can fail, and fault handling is not limited to human tasks.

Question 26Cloud Application Integration (CAI) Implementation

In Cloud Application Integration, which asset type is the container that holds the process steps, variables, and fault handlers that execute at runtime?

  1. A business service
  2. A process object Correct answer.
  3. A guide
  4. A connection

Explanation

A process object is the design-time container that holds steps, process variables, fault handlers, and event subscriptions, and it is what gets deployed to a runtime environment. Connections only describe endpoint configuration, and guides are user-facing experiences rather than executable orchestration logic.

Question 27Cloud Application Integration (CAI) Implementation

A developer is building a process and needs a value that every step can read and that persists for the whole process run but is not stored in the database. Which item should be created?

  1. A process variable Correct answer.
  2. A service connector input field
  3. A temporary variable
  4. A guide input parameter

Explanation

Process variables are scoped to the entire process instance, so any step in the run can read or assign them, and their values are retained for the life of that instance. Temporary variables exist only inside the scope of a single step and are discarded afterward.

Question 28Cloud Application Integration (CAI) Implementation

Which CAI step type should a developer add when the process must invoke an operation on an external application such as a REST API or database?

  1. Service step Correct answer.
  2. Notification step
  3. Assignment step
  4. Human task step

Explanation

A Service step invokes an operation defined through a connection to an external application, such as a REST endpoint or a database operation. Assignment steps only set variable values, and human task and notification steps involve people rather than system calls.

Question 29Cloud Application Integration (CAI) Implementation

Which CAI component is used to store the URL, authentication details, and other endpoint settings required to reach an external application?

  1. A process variable
  2. A connection Correct answer.
  3. A fault handler
  4. A decision step

Explanation

A connection holds the endpoint URL, authentication method, and related settings for an external application, and process steps reference that connection when they invoke operations. Variables and fault handlers are process constructs and carry no endpoint configuration.

Question 30Cloud Application Integration (CAI) Implementation

When a process must wait for a person to review and approve a record before continuing, which step type provides that behavior?

  1. Assignment step
  2. Human task step Correct answer.
  3. Decision step
  4. Service step

Explanation

A Human task step pauses the process and creates a task for an assigned user or group, and the process resumes only after the task is completed. Service steps call systems, while assignment and decision steps execute automatically without human interaction.

Question 31Cloud Application Integration (CAI) Implementation

Which CAI feature should be configured when a process must start automatically on a defined recurring interval without any external caller?

  1. A listener on an application event
  2. A scheduled trigger on the process Correct answer.
  3. A guide launch from the home page
  4. An API endpoint exposed by a business service

Explanation

A scheduled trigger starts a process automatically at configured times or intervals, with no external caller required. Listeners react to application events, guides are user-initiated experiences, and business service endpoints start processes only when a consumer calls them.

Question 32Cloud Application Integration (CAI) Implementation

A developer needs to expose a process so that an external system can invoke it over HTTPS as a REST-style operation. Which asset should be created?

  1. A process variable
  2. A connection
  3. A business service Correct answer.
  4. A guide

Explanation

A business service publishes process operations as an API that external consumers can invoke over HTTPS, with defined request and response structures. Guides are UI experiences, connections are outbound endpoint definitions, and variables are internal storage.

Question 33Cloud Application Integration (CAI) Implementation

Which statement correctly describes the relationship between a connection and the operations a process can call in CAI?

  1. A connection must be recreated for every Service step that uses it.
  2. Each connection exposes exactly one operation that any process may call.
  3. Operations are defined on process variables, and connections only supply credentials.
  4. A connection groups related operations for one application, and a Service step selects one of those operations. Correct answer.

Explanation

A connection represents one external application and groups the operations available on it, such as multiple REST resources or database procedures. A Service step then selects the specific operation to invoke, and the same connection can be reused across many steps and processes.

Question 34Cloud Application Integration (CAI) Implementation

A developer wants to branch on whether an incoming order total exceeds a stored credit limit. Which step should evaluate the condition and route execution?

  1. Service step
  2. Assignment step
  3. Notification step
  4. Decision step Correct answer.

Explanation

A Decision step evaluates one or more conditions and routes the process down the matching outgoing path. Assignment steps only set values, and Service and Notification steps perform calls or messaging rather than conditional routing.

Question 35Cloud Application Integration (CAI) Implementation

A process must call a fulfillment service once for every order in an incoming array. Which step configuration achieves this?

  1. A single Service step with the array mapped to a scalar input field
  2. A Service step placed inside a loop step that iterates over the array Correct answer.
  3. A Decision step with one branch per array element
  4. A parallel path with one branch per possible order status

Explanation

Placing the Service step inside a loop step causes the process to execute the call once per element of the array, with loop variables exposing the current element. A single call with an array mapped to a scalar field cannot produce one invocation per record.

Question 36Cloud Application Integration (CAI) Implementation

An outbound REST call intermittently returns HTTP 503. The developer wants the process to retry a bounded number of times before failing. Which approach fits CAI?

  1. Add a Notification step after the Service step
  2. Increase the process variable size limit
  3. Place the Service step on a parallel path so it runs twice
  4. Wrap the Service step in a loop with a counter and a decision that exits after the retry limit Correct answer.

Explanation

CAI does not apply retries implicitly, so bounded retry behavior is built explicitly by looping the Service step, tracking attempts in a counter variable, and exiting the loop when the call succeeds or the limit is reached. Parallel paths and notifications do not implement retry logic.

Question 37Cloud Application Integration (CAI) Implementation

A developer must confirm that an incoming payload contains all required fields before any downstream step runs. Which approach enforces this in CAI?

  1. Log the payload and continue processing
  2. Rely on the downstream service to reject malformed data
  3. Map missing fields to empty strings and proceed
  4. Validate the payload against the defined input schema at the process start event and route invalid payloads to a fault path Correct answer.

Explanation

Defining a start event input schema with required fields causes CAI to reject payloads that do not conform, and the fault path can return a clear error to the caller. Deferring validation downstream or silently substituting empty values lets invalid data propagate.

Question 38Cloud Application Integration (CAI) Implementation

A process must transform an incoming JSON order into a canonical XML structure before calling a legacy service. Which CAI capability is designed for this?

  1. A Human task step
  2. A guide screen
  3. A mapping or transformation step that applies field-level expressions Correct answer.
  4. A scheduled trigger

Explanation

Mapping and transformation steps convert between structures such as JSON and XML, applying field-level expressions to reshape and reformat values. Human tasks, schedules, and guide screens do not perform structural data transformation.

Question 39Cloud Application Integration (CAI) Implementation

A developer wants to reuse identical error-handling logic across many processes. Which approach best supports this in CAI?

  1. Define a shared fault handler that processes reference Correct answer.
  2. Add a Notification step to each process
  3. Place all logic in a single guide
  4. Copy the fault handler into every process manually

Explanation

A shared fault handler can be defined once and referenced by multiple processes, so changes to error handling propagate consistently. Copying handlers duplicates logic and invites drift, while notifications and guides do not centralize fault logic.

Question 40Cloud Application Integration (CAI) Implementation

During testing, a mapped output field is null even though the source field is populated. Which troubleshooting step is most direct in CAI?

  1. Redeploy the process to a different runtime environment
  2. Change the process trigger type
  3. Delete and recreate the connection
  4. Inspect the step's input and output values in the execution details for that run Correct answer.

Explanation

Execution details record the actual input and output values for each step in a specific run, which reveals whether the source value arrived and whether the mapping expression produced a value. Redeploying, recreating connections, or changing triggers does not diagnose a mapping defect.

Question 41Cloud Application Integration (CAI) Implementation

A developer needs to branch execution based on whether an order total exceeds a credit limit stored in a process variable. Which configuration is correct?

  1. A Human task step assigned to the credit team
  2. A Decision step with a condition comparing the total to the credit-limit variable Correct answer.
  3. A Service step that passes the total to an external system
  4. An Assignment step that sets the total to zero

Explanation

A Decision step evaluates a condition expression that can reference process variables, so comparing the order total to the stored credit limit routes execution correctly. The other options either call external systems, mutate data, or involve people without evaluating the condition.

Question 42Cloud Application Integration (CAI) Implementation

A process receives a repeating JSON array and must map it to a repeating XML element while preserving order and applying a currency conversion to each amount. Which approach is correct?

  1. Use a loop over the array with a mapping that converts each amount and appends the element Correct answer.
  2. Convert only the first element and reuse it for the rest
  3. Rely on the target service to reorder and convert the values
  4. Map the array to a single scalar XML field and convert once

Explanation

Iterating the array in a loop and mapping each element preserves order and lets the conversion expression run per record, producing one XML element per source entry. Collapsing the array to a scalar or converting only the first element loses data.

Question 43Cloud Application Integration (CAI) Implementation

A developer needs to correlate a request with its response when an asynchronous process is invoked. Which design element supports this?

  1. A correlation identifier passed in the request and returned or recorded with the response Correct answer.
  2. A scheduled trigger on the process
  3. A human task assigned to the requester
  4. A temporary variable scoped to the start step

Explanation

Asynchronous invocations return before processing completes, so a correlation identifier carried with the request and recorded or returned with the result lets the caller match the two. A step-scoped temporary variable cannot travel to the caller, and tasks or schedules do not provide correlation.

Question 44Cloud Application Integration (CAI) Implementation

Which consideration is most important when deciding between synchronous and asynchronous invocation for a long-running integration?

  1. Whether the process uses a Decision step
  2. Whether the connection uses REST or SOAP
  3. Whether the process has more than one process variable
  4. Whether the caller can tolerate waiting for the full processing time or must receive an immediate acknowledgment Correct answer.

Explanation

Synchronous invocation keeps the caller waiting until processing finishes, which is unsuitable when work is long-running or the caller has timeout constraints. Asynchronous invocation returns an immediate acknowledgment and delivers the result later, so caller tolerance for waiting drives the choice.

Question 45Cloud Application Integration (CAI) Implementation

A process must guarantee that two related updates either both succeed or neither is committed. Which design consideration is most relevant in CAI?

  1. Rely on the process to roll back automatically across separate systems
  2. Increase the number of retries until both updates succeed
  3. Place both updates on parallel paths so they commit simultaneously
  4. Use compensation logic in a fault handler to reverse a completed update when a later update fails Correct answer.

Explanation

Because the updates target separate systems, CAI cannot enlist them in a single distributed transaction, so consistency is achieved with compensation logic that undoes a completed update when a subsequent one fails. Parallel execution and retries do not provide atomicity across systems.

Question 46Cloud Application Integration (CAI) Implementation

A developer must build a guided experience in which a business user supplies inputs step by step and then starts a process. Which CAI asset is intended for this?

  1. A business service
  2. A fault handler
  3. A guide Correct answer.
  4. A connection

Explanation

A guide presents a sequence of screens that collect inputs from a business user and can launch a process with those values. Business services expose APIs, connections define outbound endpoints, and fault handlers manage errors.

Question 47Cloud Application Integration (CAI) Implementation

A process is deployed to a runtime environment and then modified. Which statement about process versions is accurate?

  1. A new version must be deployed for changes to take effect, and in-flight instances continue on the version they started with Correct answer.
  2. Running instances automatically migrate to the newest version mid-execution
  3. Editing a deployed process immediately changes the behavior of running instances
  4. Deploying a new version terminates all running instances

Explanation

Changes to a process are packaged as a new version that must be deployed, and instances already running continue on the version with which they started. This isolation prevents mid-flight behavior changes and avoids terminating active work.

Question 48Cloud Application Integration (CAI) Implementation

A process has a parallel path with three branches. One branch fails with an unhandled fault. Which statement describes the most likely behavior?

  1. The other branches are restarted from their first step
  2. The failed branch is automatically retried until it succeeds
  3. The process-level fault handling applies and the overall process is marked as faulted Correct answer.
  4. Only the failed branch stops while the other branches continue to completion

Explanation

A parallel path synchronizes its branches, so an unhandled fault in one branch propagates and the process-level fault handling applies, marking the instance as faulted. Branches are not silently retried, restarted, or allowed to complete in isolation from the failure.

Question 49Cloud Application Integration (CAI) Implementation

A developer must send an email to an operations group when a critical step fails. Which step type is intended for this?

  1. Assignment step
  2. Decision step
  3. Notification step Correct answer.
  4. Service step

Explanation

A Notification step sends messages such as email to configured recipients and is the intended mechanism for alerting people about failures. Assignment and Decision steps handle data and routing, and a Service step would require an external messaging API rather than the built-in notification capability.

Question 50Cloud Application Integration (CAI) Implementation

A developer wants to reuse the same error-handling logic across many processes without duplicating it. Which approach is most maintainable in CAI?

  1. Define a shared fault handler referenced by each process Correct answer.
  2. Move all error handling into a guide
  3. Add a Notification step to every process
  4. Copy the handler into each process and keep the copies in sync

Explanation

A shared fault handler defined once and referenced by multiple processes centralizes error logic, so a single change updates behavior everywhere. Duplicated handlers drift over time, and notifications or guides do not provide centralized fault logic.