Free AD0-E117: Adobe Experience Manager Architect Master Exam Questions and Answers

A company runs its website on AEM as a Cloud Service. A custom OSGi service calls an address validation API with an API key. The key is different for the Dev, Stage and Production environments and must not be stored in the Git repository. What should the architect recommend?

• AA. Set a secret environment variable per environment in Cloud Manager and reference it in the OSGi configuration with $[secret:ADDRESS_API_KEY]
• BB. Store the key as a property under /conf/global/settings and restrict read access to the service user of the OSGi service
• CC. Define the key in a run mode specific OSGi configuration folder, such as config.prod, and limit who can review that folder
• DD. Encrypt the key with the Crypto Support service on a local instance and commit the encrypted value in the OSGi configuration

An AEM 6.5 Author instance replicates to two Publish instances, publish1 and publish2. Authors report that newly activated pages appear on publish1 but not on publish2. In the Author, the replication agent for publish2 shows a queue of 37 items with the state blocked. The Author error.log contains repeated java.net.ConnectException: Connection refused entries for the publish2 transport URI. What should the architect do FIRST?

• AA. Find out why publish2 refuses connections and restore it so the agent resumes the queued items in order
• BB. Disable the publish2 agent so that activations are processed only by publish1 until the next maintenance window
• CC. Clear the publish2 queue and re-activate the affected page tree once the agent is working again
• DD. Increase the maximum parallel jobs in the Apache Sling Job Queue Configuration used for replication

A hotel chain wants seasonal offer banners to appear on its AEM Sites pages and in marketing emails sent from Adobe Campaign. The emails need styled HTML with absolute links. Authors want to create each banner once and update it in a single place. What should the architect recommend?

• AA. Author each banner as a Content Fragment and let the email platform request it through a GraphQL endpoint
• BB. Create a Live Copy of the home page and let the email platform read the rendered offer section from it
• CC. Author each banner as an Experience Fragment with a plain HTML variation for email and reference the same fragment from site pages
• DD. Develop a servlet on Publish that converts each offer page into an email template whenever the email platform requests it

A customer on AEM as a Cloud Service collects newsletter sign-ups on Publish and must pass each one to a third-party CRM. The CRM accepts calls only from allow-listed IP addresses, enforces a limit of 50 requests per second, and has a weekly maintenance window of 30 minutes. No submission may be lost. Which two design elements should the architect include? (Choose two.)

Select 2 answers.

• AA. Place submissions on a message queue in a middleware layer that retries and throttles delivery to the CRM
• BB. Enable advanced networking in Cloud Manager and configure a dedicated egress IP address for the CRM allow list
• CC. Call the CRM REST API synchronously from a Sling servlet on Publish and return its response to the visitor
• DD. Store each submission in the Publish repository and reverse replicate it to Author for a workflow to forward

An architect is assessing a customer's AEM 6.5 on-premise platform before proposing a move to AEM as a Cloud Service. The inventory shows static templates under /apps, OSGi configurations in run-mode folders such as config.publish.prod, a custom Oak index used for author-side search, and publish instances that send form submissions back to the author instance. Which finding requires a change to the solution architecture rather than only a code refactoring?

• AA. A custom Oak index that supports search queries run on the author instance
• BB. OSGi configurations kept in run-mode folders such as config.publish.prod in the code package
• CC. Form submissions transferred from publish instances to the author instance by reverse replication
• DD. Static templates stored under /apps that are used to author pages across the site

A bank's marketing team wants to show a different hero banner to returning visitors than to first-time visitors on pages served by AEM as a Cloud Service. The banner variations must be authored in AEM, pages must stay cached at the CDN and Dispatcher for anonymous traffic, and the team already licenses Adobe Target. Which implementation approach should the architect recommend?

• AA. Author the variations as Experience Fragments, export them to Adobe Target as offers, and let the Target client library swap the offer into the cached page.
• BB. Add a Sling Dynamic Include for the banner with a server-side condition on a session attribute, and cache the result on the CDN for one hour.
• CC. Configure the Dispatcher to bypass the cache for every page that contains the hero banner and select the variation on Publish for each request.
• DD. Create a separate page copy for each audience and use a cookie-based rewrite rule at the Dispatcher to route visitors to the matching copy.

A utility company runs AEM 6.5 on premise with a single TarMK author instance used by 40 authors. A new continuity policy requires that if the primary data center is lost, authoring resumes in the secondary data center within 30 minutes, with minimal loss of recent edits. The team accepts a documented manual failover procedure and wants the lowest operational complexity. What should the architect recommend?

• AA. Run a TarMK cold standby author in the secondary data center and document the steps to promote it and repoint the load balancer
• BB. Schedule an hourly content package export from the primary author and install the packages on a second author in the secondary data center
• CC. Replace the author tier with a MongoMK cluster of three active authors, with one MongoDB replica set member in each data center
• DD. Take a nightly backup of the primary author repository and restore it in the secondary data center when a failure is declared

A manufacturer runs AEM 6.5 on-premise with eight years of accumulated customizations. The company plans to move to AEM as a Cloud Service and needs an estimate of the migration effort. Before estimating, the architect must inventory the repository content and code patterns that are incompatible with AEM as a Cloud Service, such as design configurations under /etc, classic UI dialogs, and custom replication agents. What should the architect do?

• AA. Run the Content Transfer Tool in a dry-run extraction and use the size of the migration set to estimate the effort
• BB. Run the Dispatcher Converter against the existing Apache configuration and count the warnings it reports
• CC. Run the Cloud Manager code quality scan against the existing codebase and review the issues it raises
• DD. Run the Best Practices Analyzer against the source instance and review the pattern findings in Cloud Acceleration Manager

A bank is building a customer portal on AEM 6.5 with Dispatcher. Pages under /content/portal/accounts are visible only to authenticated customers who belong to a specific group, but the pages are identical for every permitted user. The bank wants these pages to be cached by Dispatcher to protect the publish instances, without allowing a cached copy to be served to an unauthorized visitor. What should the architect recommend?

• AA. Restrict the accounts path in the Dispatcher filter section to requests that carry a session cookie and cache the response as usual
• BB. Add the accounts path to the Dispatcher cache rules as non-cacheable and let every request be rendered by the publish instance
• CC. Cache the pages normally and set a short time-to-live so that unauthorized visitors can see a stale copy only briefly
• DD. Enable permission-sensitive caching so Dispatcher calls an authorization check on the publish instance before serving a cached page

A publisher wants to know whether Content Fragment models and GraphQL persisted queries can meet the needs of its new mobile app team before it commits to a headless approach. The proof of concept has a two-week time box, uses sample content only, and has no budget for new cloud environments yet. What should the architect recommend?

• AA. Build custom Sling servlets that return JSON for the sample content and compare their output with the mobile team's expectations
• BB. Migrate a representative content set to the existing Stage environment with the Content Transfer Tool and build the models there
• CC. Create a new production program, deploy a prototype through a full-stack pipeline, and load a copy of the existing content
• DD. Run a local AEM as a Cloud Service SDK instance, build sample models and fragments, and call persisted queries from a small prototype client