01. A client's PIM vendor exposes a REST API that returns one product per request. The vendor's documentation states a request limit per hour for each account and returns throttling errors above it. At that limit, retrieving all 400,000 products one at a time takes longer than the nightly import window. The vendor also publishes a daily bulk export file of the full catalog.
Which design should the Architect recommend for the nightly catalog load?
a) Call the per-product endpoint from a chunk-oriented job step with a smaller chunk size
b) Configure the service profile's rate limiter to match the vendor's hourly limit
c) Import the vendor's daily bulk export file through a scheduled job
d) Enable the circuit breaker so that throttled requests stop the job early
02. During migration, Wrenfield Garden Supply's nightly PIM feed contains new products plus only the attributes that changed on existing products. Attributes that merchandisers maintain in Business Manager, such as search placement, are not in the feed and must survive every import.
Which import mode should the job use?
a) Merge, which creates new products and changes only listed attributes
b) Replace, which overwrites each product in the file with the file's version
c) Delete, followed by a full import of every product listed in the feed
d) Merge for the new products, then Replace for products already on the instance
03. Redwood Bicycles launched four country sites at once. The combined checkout error rate stayed under its alert threshold all day, yet on the second day the Austrian team reported that nearly every Austrian checkout had failed. Austria carries a small share of total traffic.
What change to the launch monitoring would have exposed the problem on the first day?
a) Add an external uptime monitor that checks the home page of each country site every minute
b) Track and alert on each health indicator per site and locale, not only in total
c) Review Log Center once a day for each site instead of relying on live alerts
d) Raise the log level for the checkout category to debug on all four country sites
04. A tax provider used by Corvane Optics offers both a REST/JSON API and a SOAP API. The client's compliance team requires every request to be signed at the message level, so the signature can still be verified after the message passes through the provider's intermediary gateways. The provider supports that signing only on its SOAP API, through WS-Security.
Which approach should the Architect choose?
a) A REST service over HTTPS, relying on transport encryption between the two endpoints
b) A REST service with an OAuth bearer token that identifies the sender of each request
c) A SOAP service from the WSDL in webreferences2, signed with WS-Security using a key-store key
d) A SOAP service from the WSDL in webreferences2, with its signing key stored in a site preference
05. Granite Peak Tools replicated a new code version from staging to production during its launch window. Within minutes, checkout errors spike. The team has already started hotfix work in the active code version on staging, and the business wants the storefront back to its previous behavior as quickly as possible.
What is the fastest rollback?
a) Upload the previous code version to production over WebDAV and activate it
b) Revert the hotfix work on staging, then run code replication to production again
c) Undo the most recent data replication process on production
d) Activate the previously active code version on production
06. Three hours after Tidewater Sports went live, finance reported that almost no orders had been placed. Throughout that time the launch dashboard showed normal response times, a low error rate and steady Log Center traffic. The cause was a payment configuration that declined every authorization with a well-formed response.
Which gap in the launch monitoring explains why the issue went undetected?
a) It lacked an external uptime monitor checking the home page from several regions
b) It covered technical indicators only, with none for order rate
c) It sampled Log Center instead of raising every log category to debug level on production
d) It did not keep the Code Profiler running for the whole launch window
07. An Architect reviews the installation guide of a LINK cartridge that is already certified for the site's SFRA version. Installation adds custom attributes to the Product system object, a set of scheduled jobs and two service definitions. The client has two concerns: the nightly job window is already full, and the Product object carries many attributes delivered by the PIM feed.
Which two checks address the client's concerns?
(Choose 2 answers)
a) Whether the cartridge's service definitions use SOAP or REST
b) Whether the cartridge's controllers are compatible with the site's SFRA version
c) Whether the cartridge's storefront templates match the site's style guide
d) Whether the new Product attribute IDs collide with attributes the PIM feed uses
e) Whether the cartridge's job schedules fit alongside the existing nightly jobs
08. Larkspur Gear is integrating its SFRA checkout with a loyalty provider's REST API. Each call needs an OAuth bearer token from the provider's token endpoint, obtained with client credentials, and a token stays valid for one hour. The provider limits how many tokens each client may request per hour, its security terms forbid persisting tokens to any database, and the checkout latency budget leaves no room for a token round trip on each call.
How should the Architect design token handling?
a) Request a new token through a separate token service before each loyalty call
b) Store the issued token in a custom object, refreshed by a job that runs every 50 minutes
c) Keep tokens from a token service in a custom cache that expires before each token does
d) Keep the issued token in the shopper's session and request a new one when each session starts
09. A legacy credit-check endpoint used by Quarry Hill Furniture accepts only HTTP POST requests with the content type application/x-www-form-urlencoded and named fields. The Architect wants the integration to need as little custom request-building code as possible.
Which service type should the Architect specify?
a) HTTP Form, which URL-encodes the call arguments as name/value pairs
b) Generic, with execute posting the form through dw.net.HTTPClient
c) SOAP, with the endpoint's fields mapped into a generated WSDL client
d) HTTP, with createRequest building a JSON request body for the endpoint
10. Velora Jewelry runs a main site and an outlet site that both sell from one warehouse. Each site has its own inventory list, and both lists are loaded from the same warehouse feed every hour. Between feeds, both sites occasionally sell the last unit of the same item.
What should the Architect recommend?
a) Load the warehouse feed into both lists every five minutes
b) Assign one shared inventory list to both sites
c) Enable Default In-Stock on the outlet site's inventory list
d) Copy the main list's levels to the outlet list after each order