Integration-Architect Salesforce Certified Platform Integration Architect (Plat-Arch-204) Questions and Answers

The Batch Data Synchronization pattern is suitable for this business use case because it allows transferring large volumes of data between Salesforce and the scheduling system in batches. This can handle the surge of requests for workshop registrations without overloading the systems or affecting the performance. The batch process can run at regular intervals or on demand to synchronize the data between the systems1

1: Batch Data Synchronization Pattern

 The BatchApexErrorEvent object allows handling errors that occur during Batch Apex execution. By subscribing to this event,the integration can retry the invoice generation for the failed records or perform other recovery actions. This is a resilient and scalable solution that does not require custom code in the Batch Apex job itself or in the middleware1

 Configuring Identity Type, Certificate, and CSRF Protection for OData connection are threeconfigurations that an Integration Architect should recommend to secure requests coming from Salesforce. Identity Type is used to specify the authentication method for the OData connection, such as Basic Authentication, OAuth 2.0, or Named Principal. Certificate is used to enable SSL/TLS encryption for the OData connection, which protects the data in transit from eavesdropping or tampering. CSRF Protection is used to prevent cross-site request forgery attacks, which exploit the trust between the user and Salesforce by sending malicious requests from another website. Configuring Special Compatibility for OData connection is not a configuration that is related to security, but rather to compatibility issues with different versions or implementations of OData.Configuring CSRF Protection on External Data Source is not a configuration that can be done in Salesforce, but rather on the external data source itself. Reference: Salesforce Connect: Custom Adapters Developer Guide, page 9-10

Option A is correct because external gateway products in use can affect the performance and security of outbound integrations from Salesforce to on-premise servers. External gateway products are software or hardwaredevices that act as intermediaries between Salesforce and the on-premise servers, such as firewalls, proxies, load balancers, or VPNs. They can have different configurations, features, and limitations that can impact the speed, reliability, and security ofthe data transmission. For example, some external gateway products may require authentication, encryption, or compression of the data, which can add overhead and latency to the integration. Some external gateway products may also have bandwidth or throughput limits, which can affect the scalability and availability of the integration. Therefore, the architect should consider the external gateway products in use before recommending a solution.

Option B is incorrect because a default gateway restriction is not a factor that can affect the performance and security of outbound integrations from Salesforce to on-premise servers. A default gateway restriction is a feature that allows administrators to restrict outbound requests from Salesforce to a specific IP address or domain name. This can help prevent unauthorized or malicious requests from Salesforce to external systems. However, this feature does not affect the performance or security of the outbound requests themselves, as it only acts as a filter for the destination of the requests.

Option C is incorrect because considerations for using deterministic encryption are not relevant for outbound integrations from Salesforce to on-premise servers. Deterministic encryption is a type of encryption that produces thesame ciphertext for the same plaintext input. This can help preserve some functionality and performance of encrypted data in Salesforce, such as filtering, sorting, and indexing. However, deterministic encryption is not applicable for outbound integrations from Salesforce to on-premise servers, as it is only supported for custom fields and not for standard fields or attachments. Moreover, deterministic encryption does not affect the security of the data transmission itself, as it only encrypts the data atrest in Salesforce.

Option D is incorrect because Shield Platform Encryption limitations are not relevant for outbound integrations from Salesforce to on-premise servers. Shield Platform Encryption is a feature that allows administrators to encrypt sensitive data at rest in Salesforce using AES 256-bitencryption. This can help protect data from unauthorized access or theft. However, Shield Platform Encryption limitations are not relevant for outbound integrations from Salesforce to on-premise servers, as they only affect the functionality and performance of encrypted data in Salesforce, such as searching, reporting, or validation rules. Shield Platform Encryption does not affect the security of the data transmission itself, as it only encrypts the data at rest in Salesforce.

Leverage Change Data Capture to track changes to the Contact object and write a CometD subscriber on the homegrown application. This solution meets the requirements of keeping Salesforce and the homegrown application synchronized for consistency, using event-driven and asynchronous integration. Change Data Capture is a feature that allows you to receive near real-time notifications of changes to Salesforce records, such as inserts, updates, deletes, and undeletes. You can use Change Data Capture to track changes to the Contact object, and publish them as events to an event bus. CometD is a protocol that allows you to subscribe to events from the event bus using a streaming API. You can write a CometD subscriber on the homegrown application, and use it to receive the events and update the customer data accordingly. References: Certification - Integration Architect - Trailhead, [Change Data Capture Developer Guide], [Streaming API Developer Guide]

Option B is a correct finding, because it means that the company can use their cloud-based ETL tools to extract data from the legacy systems via the APIs exposed by the on-premise middleware, transform the data as needed, and load the data into Salesforce using queues. Queues are a way of managing asynchronous operations, such as bulk data loading, by placing them in a queue and executingthem when resources are available1.

Option C is also a correct finding, because it meansthat the company can expose their legacy data to the cloud via their on-premise middleware APIs. APIs are a way of enabling communication and data exchange between different systems using standard protocols and formats2. By providing APIs to access their legacy data, the company can use any cloud-based tool or application that can consume those APIs toretrieve and manipulate the data.

Option A is not a correct finding, because it means that the company cannot access their legacy data from the cloud, which prevents them from using their cloud-based ETL tools or any other cloud-based integration solution. If only on-premise systems are allowed to access the legacy systems, then the company would need to use an on-premise integration tool or application to copy the data to Salesforce.

Option D is not a correct finding, because it does not affect the ability to copy legacy data in Salesforce. Queues are useful for on-premise integration, but they are not necessary for copying data to Salesforce. The company can use other methods, such as direct API calls or batch processes, to load data into Salesforce without using queues.

The Connect REST API is designed to access Chatter feeds and social data such asusers, groups, followers, and files. By using this API, the integration can query the Chatter posts that include the #thanksNTO hashtag and post them to Twitter using another API. The REST API, Streaming API, and Apex REST are not specific to Chatter and do not provide the same level of functionality and ease ofuse as the Connect REST API2

The maximum calloutsin a single Apex transaction is one of the factors that should be considered when choosing Apex code to make callouts to an external system. The limit is 100 callouts per transaction, and if this limit is exceeded, a LimitException is thrown. Therefore, the integration architect should evaluate the expected volume and frequency of the callouts and design the solution accordingly2

Answer D is valid because using Bulk API 2.0 for both inserting and querying data canmeet the requirements of loading and extracting large volumes of data in a day. Bulk API 2.0 is a RESTful API that allows creating, updating, deleting, or querying millions of records asynchronously by uploading or downloading CSV or JSON files. Bulk API2.0 has several advantages over Bulk API 1.0, such as:

It does not count against the daily limit of 5,000 batches per rolling 24-hour period.

It supports PK Chunking for queries, which automatically splits large data sets into manageable chunks based onthe primary key.

It simplifies the job lifecycle and reduces the number of API calls needed to perform a bulk operation.

Answer A is not valid because using Bulk API 2.0 for inserting data and REST API for querying data is not an optimal solution for handling large volumes of data in a day. REST API is a synchronous API that allows creating, updating, deleting, or querying individual records or small batches of records using HTTP methods. REST API has several limitations and drawbacks for this use case, such as:

It counts against the daily limit of 15,000 API requests per 24-hour period.

It does not support PK Chunking for queries, which means that large data sets may exceed the query timeout or heap size limits.

It requires more API calls and processing time to perform a bulk operation than Bulk API 2.0.

Answer B is not valid because using Bulk API 1.0 for both inserting and querying data is not a feasible solution for handling large volumes of data in a day. Bulk API 1.0 is a SOAP-based API that allows creating, updating, deleting, or querying millions of records asynchronously by uploading or downloading XML or CSV files. Bulk API 1.0 has several limitations and drawbacks compared to Bulk API 2.0, such as:

It counts against the daily limit of 5,000 batchesper rolling 24-hour period, which may not be enough to load and extract 90M and 30M records respectively.

It does not support JSON format for data files, which may not be compatible with some external systems or applications.

It requires more API calls andcomplexity to manage the job lifecycle and handle errors or retries than Bulk API 2.0.

Answer C is not valid because using Bulk API 1.0 for inserting data and REST API for querying data is not a suitable or reliable solution for handling large volumes ofdata in a day. As explained above, both Bulk API 1.0 and REST API have limitations and drawbacks that may affect the performance, efficiency, and scalability of the integration.

 The integration architect should require bulk load for Einstein Analytics and callouts to RESTful services for the Community Cloud portal and Einstein Analytics. Bulk load is a capability that allows loading large volumes of data into Einstein Analytics from external sources, such as the Patient Prescriptions System. This can be done securely by using encryption in transit and at rest, and by masking or removing any identifiable patient information from the data. Callouts to RESTful services are a capability that allows making requests to external web services from Salesforce, such as the Patient Prescriptions System. This can be done securely by using identity tokens or certificates for authentication and authorization, and by using encryption in transit and at rest12

 The OAuth protocol is a standard way to authorizeaccess to web resources. By configuring a connected app with an authorization endpoint of the API gateway, Salesforce can obtain an access token from the API gateway and use it to invoke the external API securely. This avoids the need to manage certificates oruser credentials for authentication2

The Data Virtualization pattern allows Salesforce to access external data sources without storing the data in Salesforce. This reduces the data storage and synchronization costs and enables real-time access to the historical data. Salesforce Connect can be used to consume OData endpoints exposed by the ESB tool and display the external objects as related lists or custom components on the Case object

 Both Platform Events and Outbound Message offer declarative means for asynchronous near-real time needs. They aren’t best suited for real-time integrations (A). In both Platform Events and Outbound Messages, the event messages are retried by and delivered in sequence, and only once. Salesforce ensures there is no duplicate message delivery (B). Message sequence is possible in Outbound Message but not guaranteed with PlatformEvents. Both offer very high reliability. Fault handling and recovery are fully handled by Salesforce ©. Number of concurrent subscribers to Platform Events is capped at 2,000. An Outbound Message configuration can pass only 100 notifications in a single message to a SOAP end point (D). Both Platform Events and Outbound Message are highly scalable. However, unlike Outbound Message, only Platform Events have Event Delivery and Event Publishing limits to be considered (E). Reference: Salesforce Integration Architecture Designer Resource Guide, page 23

The mobile device should make a REST Apex inbound call to meet the requirement. A REST Apex inbound call allows the mobile device to invoke custom Apex logic on Salesforce using a RESTful interface. The Apex logiccan then access the geolocation field using the user’s OAuth token and update it with the GPS coordinates from the mobile device. The geolocation field is secured with Field Level Security, so the user can view the value without changing it. A SOAP API inbound call or a REST API inbound call would not be able to access the custom geolocation field, as they are not supported by the standard Salesforce objects and fields. Reference: Salesforce Integration Architecture Designer Resource Guide, page 20

Using Salesforce custom object, custom REST API and ETL is a better solution than using Salesforce standard object, REST API and ETL. A custom object can store the order information and provide more flexibility and control overthe data model, validation rules, triggers, etc. A custom REST API can be used to create orders in real time from Salesforce, using the ESB as a proxy to communicate with the legacy system. An ETL tool can be used to sync the order history and status fromthe legacy system to Salesforce, using the last modified date as a filter. This way, sales reps can see the order history and the most up-to-date information on current order status, and managers can run reports on order volumes and fulfillment timelines.A standard object, such as Order, may not meet all the business requirements and may require more customization than a custom object. A standard REST API may not provide enough security and error handling for creating orders in real time. Reference: Salesforce Integration Architecture Designer Resource Guide, page 18

 The best solutions for minimizing potential downtime for users in this situation are to use named credentials and an enterprise ESB. Named credentials allow you to store the URLand authentication parameters of an external service in one definition, and then use it in Apex callouts or declarative integrations. If the URL of the external service changes, you only need to update the named credential once, and all the integrations that use it will be updated automatically. An enterprise ESB is a middleware layer that acts as a broker between different systems and services. It can provide routing, transformation, orchestration, and monitoring capabilities for integrations. If the URL of an external service changes, you only need to update the ESB configuration once, and all the integrations that use it will be updated automatically. Option B is not correct because remote site settings only allow you to specify the domains that are allowed for callouts from your org. They do not store any authentication parameters or provide any routing or transformation capabilities. Option C is not correct because content security policies are used to control what resources can be loaded on a Visualforce or Lightning page. They do not store any authentication parameters or provide any routing or transformation capabilities. References:

NamedCredentials

Streamline and Secure: Mastering Named Credentials in Salesforce

 Invoking middleware service to retrieve valid shipping methods is a solution that can allow the sales representatives to select between valid services for the customer’s country. The middleware service can act as a single point of entry for all shipping services and provide routing and transformation capabilities. Using middleware to abstract the call to the specific shipping services is a solution that can allow the sales representatives to request shipping estimates from the selected service. The middleware can hide thecomplexity and heterogeneity of the shipping services and provide mediation and orchestration capabilities. Using Platform Events to construct and publish shipper-specific events is not a solution, as Platform Events are used for event-driven integration,not for web-service integration. Storing shipping services in a picklist that is dependent on a country picklist is not a solution, as it does not address how to request shipping estimates from the shipping services. Reference: Salesforce Integration Architecture Designer Resource Guide, page 16-17

 A trigger is a programmatic way of executingsome logic when a record is inserted, updated, deleted, or undeleted in Salesforce. A trigger can invoke an Apex class that contains the code to perform aREST callout to an external system. A REST callout is a way of sending an HTTP request to a service endpoint and receiving a response. A REST callout can be synchronous or asynchronous, depending on whether the Apex code waits for the response before continuing the execution. A synchronous callout is suitable for scenarios where the response is needed immediately, such as order fulfillment1.

An Apex proxy class is a class that is generated from a WSDL (Web Service Description Language) document of an external SOAP web service. An Apex proxy class can be used to make a SOAP callout to the external web service, but not a REST callout. A SOAP calloutis another way of sending an HTTP request to a service endpoint and receiving a response, but it uses a different format and protocol than REST2.

Process Builder is a declarative tool that allows you to automate business processes by defining criteria and actions. Process Builder can invoke an Apex class that implements the Process.Plugin interface, which allows you to extend the functionality of Process Builder with custom logic. However, Process Builder does not support synchronous callouts, because it runs in the background and does not wait for the response from the external system. Process Builder only supports asynchronous callouts, which are executed after the transaction is committed3.

Batch Apex is a way of processing large volumes of data asynchronously by breaking them into smaller batches of records. Batch Apex can be used to perform complex or long-running operations on data, such as data cleansing, archiving, or integration. Batch Apex can make callouts to external systems by implementing the Database.AllowsCallouts interface in the batch class. However, Batch Apex is not suitable for scenarios wherethe callout needs to be synchronous, because it runs in the background and does not wait for the response from the external system. Batch Apex also has some limitations, such as the maximum number of batches in the queue, the maximum number of records per batch, and the maximum number of callouts per batch4.

A Lightning Component is a reusable unit of user interface that can be used to build modern web apps with Salesforce. A Lightning Component can make a callout to an external system by using JavaScript code or by invoking an Apex controller class that contains the logic for the callout. A Lightning Component can make a synchronous or asynchronous callout, depending on whether the JavaScript code or Apex code waits for the response before continuing the execution. However, a Lightning Component is not a good choice for scenarios where the callout needs to be triggered by arecord change, such as when an opportunity is closed/won. A Lightning Component requires user interaction, such asclicking a button or loading a page, to initiate the callout5.

Therefore, the correct answer is A, because writing a trigger that invokes an Apex class to make a REST callout to the Order Management System is the only option that satisfies the requirements of making a synchronous callout when an opportunity is closed/won.

The three systems from the current system landscape that can beretired with the implementation of Salesforce are Case Management System, Sales Activity System, and Email Marketing System. These systems can be replaced by Salesforce Service Cloud, Sales Cloud, and Marketing Cloud respectively, which provide similar orbetter capabilities for managing cases, sales activities, and email campaigns. Option A is not correct because Order Management System cannot be retired with the implementation of Salesforce. The company wants to reuse their existing enterprise capabilities for order management processes, which are likely to be complex and customized. Salesforce does not provide a native order management solution for B2C commerce scenarios. Option E is not correct because Quoting System cannot be retired with the implementation of Salesforce. The company wants to reuse their existing enterprise capabilities for quoting processes, which are likely to be complex and customized. Salesforce does not provide a native quoting solution for B2C commerce scenarios.

Change Data Capture is a feature that enables you to receive near-real-time changes of Salesforce records, including create, update, delete, and undelete operations. Change Data Capture retainschange events in the event bus for up to three days, so you can resume data replication from the point of failure. Change Data Capture also provides a consistent and reliable way to synchronize data changes with external systems, without the need for custom triggers or code. Reference: Salesforce Integration Architecture Designer Resource Guide, page 24

Answer A is valid because the Remote Process Invocation fire and forget pattern allows sending a message from Salesforceto the ERP system without waiting for a response or acknowledgement. This enables near real-time integration and reduces the latency and complexity of the integration. The Flow Builder is a low code tool that can be used to invoke an Apex actionthat makes the HTTP callout to the REST endpoint on insert of the order object12

Answer C is valid because the Change Data Capture feature allows capturing changes in Salesforce data and publishing them as eventsto a streaming channel. The middleware can subscribe to the channel and receive the events that contain the order details. The replay Id is a unique identifier for each event that can be used by the middleware to resume from the last processed event in case of failures or interruptions. This ensures reliable and consistent integration between Salesforce and the ERP system3

Answer B is not valid because the Outbound Messaging feature uses SOAP messages to send notifications to external systems based on workflow rules. This requires configuring a SOAP endpoint and a WSDL file for the ERP system, which may not be compatible with the REST endpoint requirement. Moreover, Outbound Messaging does not guarantee delivery or order of messages, which may affect the accuracy and timeliness of the integration.

Answer D is not valid because the Platform Event feature allows publishing custom events from Salesforce to external systems or vice versa using a publish/subscribe model. However, this requires creating a platform event object and defining its fields and permissions, which may notbe a low code solution. Moreover, using the record type as the platform event name may not be a valid or meaningful option, as record types are used to define different business processes or user interfaces for an object.

1:Remote Process Invocation—Fire and Forget 2: Invoke an Apex Class That Makes a Callout 3: Change Data Capture : Replay Events Using Middleware : Outbound Messaging Implementation Guide : Platform Events Developer Guide

Theintegration architect should recommend enforcing separate security protocols and return formats at the first tier of the API-led architecture. The first tier is also known as the experience layer, which is responsible for providing a tailored interface for each system of engagement. By enforcing security and format at this layer, the integration architect can ensure that each system of engagement can access the data in a secure and consistent way, without affecting the other layers.

The best solution for this scenario is to use a custom object to store the file location URL and redirect the community user to the on-premise file store when they click on the URL.This way, the community user can access the large design files without having to download them or use any external data source. Option A is not feasible because the Request and Reply integration pattern is not suitable for large files and would cause performance issues. Option B is not correct because Salesforce Connect cannot upload files to an external object, only data. Option D is not possible because Salesforce Files has a limit of 2 GB per file, and the design files are 2.5 GB in size. References:

Salesforce Connect Developer Guide

Salesforce Files Developer Guide

The considerations that an Integration Architect should do prior to the implementation of Shield Platform Encryption are:

Review shield platform encryptionconfigurations.

Encrypt all the data so that it is secure.

Shield Platform Encryption is a feature that allows you to encrypt sensitive data at rest in Salesforce, such as social security numbers, without compromising critical platform functionality. Before implementing Shield Platform Encryption, you should review the shield platform encryption configurations, such as the encryption key management, the encryption policy, and the encrypted fields and files. You should also encrypt all the data that is subject to encryption, not just the data using the most current key. Encrypting all the data ensures that your data is secure and compliant with your business requirements. Encrypting the data using the most current key is not a valid consideration because Shield Platform Encryption uses a deterministic encryption scheme that does not allow you to rotate or re-encrypt your data with a new key. Using Shield Platform Encryption as a user authentication or authorization tool is not a valid consideration because Shield Platform Encryption is not designed for that purpose. Shield Platform Encryption only encrypts data at rest, not in transit or in use.

Using middleware queuing and buffering is a solution that can handle errors during service outages by storing the messages in a queue until the service is available again. This way, Salesforce does not loseany data or encounter any failures when the service is down. Using @future jobld and custom scheduled apex process to retry failed service calls is not a good solution because it can consume a lot of governor limits and create a lot of duplicate records. Using Outbound Messaging to automatically retry failed service calls is also not a good solution because it has a limited number of retries and a fixed retry interval, which may not be sufficient for long service outages. Using Platform Event replayld and custom scheduled Apex process to retrieve missed events is not applicable for this scenario because Platform Events are used for event-driven integration, not for web-service integration. Reference: Salesforce Integration Architecture Designer Resource Guide, page 29-30

Using a Connected App for each external system integration is a good solution because it can provide security, auditing, and monitoring features for each integration. A Connected App is an application that can connect to Salesforce using APIs and OAuth as an authentication protocol. A Connected App can also enforce policies such as IP restrictions, login hours, and session timeout for each integration. Using a shared integration user for the three external system integrations is not a good solution because it violates the principle of least privilege, as well as makes it difficult to audit the activities of each system. Using a shared Connected App for the three external system integrationsis also not a good solution because it does not allow for granular control and visibility of each integration. Using a unique integration user for each external system integration is not enough to meet the security and auditing requirements, as it does not provide any mechanism for authentication, authorization, or encryption. Reference: Salesforce Integration Architecture Designer Resource Guide, page 20-21

Option C is correct because including selective criteria in query filters can improve the performance of the integration. Selective criteria are filters that reduce the number of records that need to be scanned by the query optimizer, such as indexed fields, standard indexes, or custom indexes. Selective criteria can help the query run faster and avoid hitting the query timeout limit or the query row limit.

Option D is correct because removing the sharing restrictions can improve the performance of the integration. Sharing restrictions are filters that limit the access to records based on the user’s role, profile, or sharing rules.Sharing restrictions can add complexity and overhead to the query execution, as they require additional joins and calculations. Removing the sharing restrictions can simplify the query and reduce the number of records that need to be processed.

Option A is incorrect because including non-selective criteria in query filters can degrade the performance of the integration. Non-selective criteria are filters that do not reduce the number of records that need to be scanned by the query optimizer, such as non-indexed fields, formula fields, or OR conditions. Non-selective criteria can cause the query to run slower and hit the query timeout limit or the query row limit.

Option B is incorrect because removing the query filters can degrade the performance of the integration. Query filters are conditions that specify which records to retrieve from the database, such as WHERE clauses or LIMIT clauses. Query filters can help the query run faster and avoid retrieving unnecessary or unwanted data. Removing the query filters can increase the number of records that need to be processed and transmitted by the integration.