Connecting via the application's REST API is a standard, vendor-supported method for integrating with complex systems like CRMs. This approach ensures data integrity by interacting with the application layer, respecting the system's business logic and security model. It is the required method for most SaaS CRMs (e.g., Salesforce) where the underlying database is inaccessible. Qlik Replicate provides both a generic REST API source and specific application connectors that utilize APIs, making this a robust and widely applicable solution for modern enterprise applications. This method provides a stable integration point that is less likely to be affected by underlying database schema changes.

A. This option is ambiguous as "connecting to the application" lacks the specific technical protocol required to define a Qlik Replicate source endpoint.

B. Using CSV exports is a batch process, not real-time replication, and negates the core Change Data Capture (CDC) functionality of Qlik Replicate.

D. Connecting directly to the RDBMS can bypass critical application business logic, potentially leading to data inconsistency, and may violate vendor support agreements.

1. Qlik Replicate Help, "Salesforce source endpoint": This official documentation describes the configuration for replicating data from Salesforce, a major CRM application. The prerequisites and settings exclusively reference the Salesforce API and user permissions required to access it, confirming that the API is the sanctioned method for this type of source. (Accessed via Qlik Help Portal for Qlik Replicate, specific version documentation).

2. Qlik Replicate Help, "REST API source endpoint": This document details the generic REST API source endpoint in Qlik Replicate. It states, "The REST API endpoint allows you to replicate data from web applications that expose their data through REST APIs." This confirms that using a REST API is a built-in, supported capability for accessing application data. (Accessed via Qlik Help Portal for Qlik Replicate, specific version documentation).

3. Qlik Replicate User Guide, Section: "Application Sources": The product documentation explicitly categorizes certain sources like Salesforce and SAP as "Application Sources," which are distinct from "Database Sources." This architectural distinction highlights that the intended connection method is through the application layer (often an API), not directly to the database, preserving application integrity. (Found in the table of contents for sources in any recent Qlik Replicate User Guide).

After a Full Load test is completed and Change Processing is enabled, the goal is to test only the Change Data Capture (CDC) functionality. The 'Start Change Processing Only' option, found under 'Advanced Run Options,' is specifically designed for this scenario. It instructs Qlik Replicate to bypass the Full Load phase entirely and begin capturing and applying changes from the source system from the point the Full Load ended, or a specified point in time. This provides a direct and efficient method to validate the CDC configuration.

A. 'Reload Target' would unnecessarily re-run the entire Full Load, which has already been successfully tested, wasting time and resources.

B. Re-creating the task is a drastic and inefficient action when the existing task configuration is correct and only needs to be run in a different mode.

C. 'Resume Processing' is the standard operation to continue a task. While it would start CDC after a full load, 'Start Change Processing Only' is the more explicit and precise method for isolating and testing only the CDC part.

1. Qlik Replicate User Guide, "Starting a task in Change Processing only mode": This section explicitly states, "You may want to start a task in Change Processing only mode when the target tables have already been loaded with the source data (either by a third-party tool or by a previous Replicate task)." This directly aligns with the scenario where a Full Load test has already been completed.

2. Qlik Replicate User Guide, "Using the Advanced Run Options dialog box": The documentation details the advanced options available when running a task. It describes the ability to start processing changes from a specific time or log position, which is the underlying mechanism for the 'Start Change Processing Only' mode, allowing the user to bypass the Full Load phase.

In Qlik Replicate, the "Apply Exceptions" feature captures records that fail to be applied to the target endpoint due to errors (e.g., data constraint violations). These failed records are stored in a dedicated control table, typically named attrepapplyexceptions, on the target database. By default, Qlik Replicate is configured to retain these exception records indefinitely. This ensures that no data is lost and allows administrators ample time to investigate, correct, and reapply the failed transactions without a risk of automatic deletion. This default behavior can be overridden in the task's Error Handling settings.

B. 7 days: This is a configurable retention period, but it is not the default setting for Apply Exceptions.

C. 60 days: This is a possible custom value for data retention, not the out-of-the-box default for this specific feature.

D. 30 days: While a common retention period for logs or other temporary data, it is not the default for Apply Exceptions.

1. Qlik Replicate User Guide (May 2024): In the section "Defining advanced error handling settings," under the "Store changes" tab, the setting "Delete exceptions older than (days)" is described. The documentation implicitly confirms the default is indefinite by not having a default numerical value checked or pre-filled, requiring the user to explicitly enable and set a time limit. The default state is to not delete. (Reference path: User Guide -> Setting up a replication task -> Using the Replicate Console -> Defining and managing tasks -> Defining advanced error handling settings -> The Store changes tab).

2. Qlik Replicate User Guide (May 2024): The section "Managing Apply Exceptions" describes the attrepapplyexceptions table and the process for reloading or deleting records. The entire process is manual by default, reinforcing that automatic deletion is not the standard behavior. (Reference path: User Guide -> Monitoring tasks -> Managing Apply Exceptions).

In Qlik Replicate's CDC (Change Data Capture) mode, the "Batch optimized apply" method is designed to enhance performance by grouping multiple individual changes into a single transaction before applying them to the target. The task can be configured to apply these batches based on two primary triggers that work in tandem: a time interval and data volume. The batch is applied whenever the specified time limit is reached OR the accumulated size of the change data reaches the defined volume threshold, whichever occurs first. This provides a flexible way to balance throughput and latency.

A. Source connection processes are configured at the source endpoint level, not within the change processing apply mode settings.

B. While you can limit the maximum number of records in a batch, the primary triggers for applying the batch are time and total data volume.

D. This option is incomplete as it only mentions the time-based trigger and omits the equally important volume-based trigger.

1. Qlik Replicate User Guide: In the section "Setting advanced run options," under the "Change Processing Tuning" tab, the configuration for "Batch optimized apply" mode explicitly lists the options: "Apply changes every seconds" (Time) and "Apply changes when accumulated changes exceed MB" (Volume). This directly confirms that batching is controlled by time and/or volume.

2. Qlik Community Knowledge Base: Article 000038693, "Qlik Replicate batch optimized apply mode and transaction apply mode," explains that in Batch Optimized Apply, "Replicate accumulates the changes and applies them to the target in batches. The batch is applied when either the configured time or size is reached." This corroborates the dual trigger mechanism.

When performing a large-scale, out-of-band data operation like an IBM load on a target database, the Qlik Replicate task must be stopped. This prevents the task from failing due to table locks, unavailability, or other conflicts caused by the load utility. It also avoids data inconsistency, as the load utility's changes are not captured by Qlik Replicate. Once the IBM load is complete, the task should be resumed. Resuming allows the task to continue capturing and applying transactional changes from the source system from the exact point where it was stopped, ensuring ongoing data synchronization after the merge.

A. Stop task, finish IBM load, reload target: This is incorrect because the IBM load has already populated the target. Performing a full reload with Qlik Replicate immediately after would be redundant and extremely inefficient, overwriting the work just completed.

C. Continue to run the task: This is incorrect as it would likely lead to task failure due to resource contention and locking on the target tables. Furthermore, it would cause severe data integrity issues as the massive changes from the load are not replicated.

D. Create a new task after finishing IBM load: This is a possible but less efficient approach. Creating a new task is unnecessary overhead when the existing task can be simply resumed, preserving its configuration, history, and the correct starting point for change data capture.

1. Qlik Replicate User Guide (November 2023), "Controlling tasks," Page 218: The documentation states, "You may want to stop a task temporarily, for example, to perform maintenance on the source or target machine." The IBM load constitutes a major maintenance/update event on the target.

2. Qlik Replicate User Guide (November 2023), "Resuming a task," Page 219: The guide explains, "When you resume a task, processing continues from the point at which it was stopped." This confirms that resuming is the correct procedure to continue replication after a planned stoppage.

3. Qlik Enterprise Manager User Guide (November 2023), "Task Control," Page 102: The guide describes the Stop and Resume operations as standard controls for managing replication tasks. The described scenario is a primary use case for temporarily stopping and then resuming a task to accommodate external processes affecting the target environment.