Oracle 1Z0-1084-26 Certification Sample Questions and Answers

1Z0-1084-26 PDF, 1Z0-1084-26 Dumps PDF Free Download, 1Z0-1084-26 Latest Dumps Free PDF, Cloud Infrastructure Developer Professional PDF DumpsThe Oracle Cloud Infrastructure Developer Professional (1Z0-1084-26) Sample Question Set is designed to help you prepare for the Oracle Cloud Infrastructure Certified Developer Professional certification exam. To become familiar with the actual Oracle Certification exam environment, we suggest you try our Sample Oracle 1Z0-1084-26 Certification Practice Exam.

This Oracle Cloud Infrastructure Developer Professional certification sample practice test and sample question set are designed for evaluation purposes only. If you want to test your Oracle 1Z0-1084-26 knowledge to identify your areas of improvement and get familiar with the actual exam format, we suggest you prepare with the Premium Oracle Cloud Infrastructure Certified Developer Professional Certification Practice Exam. Our team of Oracle Cloud Infrastructure experts has designed Questions-Answers for this premium practice exam by collecting inputs from recently certified candidates. Our premium Oracle 1Z0-1084-26 certification practice exam will boost your confidence as well as your actual Oracle Cloud Infrastructure Developer Professional exam result.

Oracle 1Z0-1084-26 Sample Questions:

01. A mobile banking app's account service restarted after a brief outage. Thousands of app clients had been retrying failed balance requests every two seconds, so they retried in synchronized waves, and each wave overloaded the service again as soon as it came back. The clients must keep retrying transient failures automatically.
Which retry strategy should the clients adopt?
a) A fixed retry interval shortened to half a second
b) Immediate retries with a higher maximum number of attempts
c) Retries repeated at every layer of the call chain
d) Exponential backoff with random jitter
 
02. Customers of an online grocery report that order-status pages load slowly. An APM trace for one slow request contains these spans:
api-gateway GET /orders/{id} — 2,450 ms, root span
order-service GET /orders/{id} — 2,410 ms, child of the gateway span
inventory-service GET /stock — 90 ms, child of order-service
pricing-service POST /quote — 2,230 ms, child of order-service
pricing-db SELECT quote_rules — 45 ms, child of pricing-service
The database team also reports high CPU on the pricing database.
Where should the developer look first for the cause of the delay?
a) The order-service span, the longest span below the root
b) The quote_rules query against the pricing database
c) Inside pricing-service's own code
d) The gateway, before the request reaches order-service
 
03. A developer at a pet-grooming booking startup created an APM domain in a shared observability compartment. The developer enabled tracing on the OCI Functions application with that domain selected and enabled tracing on each function. The functions run successfully, but no traces from them ever appear in the domain.
What is the most likely cause?
a) Invocation logs are not enabled on the application, so no spans are recorded
b) No policy lets OCI Functions access the APM domain
c) The functions do not propagate X-B3 headers on their outgoing calls
d) The function code creates no spans of its own with a Zipkin or OpenTelemetry library
 
04. Two business units at a travel company keep container images in separate compartments, bookings and loyalty. The loyalty group has manage repos in the loyalty compartment and no access to bookings. A loyalty developer tries to create a repository named web/profile-ui in loyalty, and the request fails. The bookings unit already has a repository called web/profile-ui in its own compartment.
What is the most likely cause of the failure?
a) A repository name that contains a slash can be created only in the root compartment
b) Creating any repository requires manage repos granted at the tenancy level
c) Repository names must be unique across all compartments in the tenancy
d) The loyalty group also needs read repos in bookings
 
05. An online grocery retailer scaled its web application from one container instance to six behind a load balancer. Customers now report that items disappear from their baskets between page loads, and receipt PDFs that the application writes to its container file system are sometimes missing after an instance restarts.
Which two changes make the application instances stateless?
(Choose two.)
a) Enable session affinity on the load balancer so each customer returns to one instance
b) Mount a host path volume from the worker node for the receipt files
c) Keep basket contents in an external cache or database shared by all instances
d) Write the receipt files to Object Storage in place of the local file system
e) Give each instance more memory for holding customer baskets
 
06. After a long outage of a downstream warehouse API, a media company decides to abandon the backlog of stale playback events in its stream. The consumer code now calls CreateGroupCursor with the LATEST type, keeping the existing group name playback-agg, and every instance is restarted.
The instances still work through the old backlog from where they stopped. What is the most likely cause?
a) For a group cursor, LATEST starts at the oldest retained message, and it means the newest message only for a plain cursor
b) For a group that already exists, the cursor type is ignored and consumption resumes from the group's committed offsets
c) Unread messages must be drained by every group before a LATEST cursor can move past them
d) LATEST applies only to partitions added to the stream after the cursor was created
 
07. A car rental company's Reservations service commits each booking to its own database and then publishes a BookingCreated event to a message stream in the same request handler. Occasionally the process terminates between the commit and the publish, and the Billing and Loyalty services never learn of the booking.
Publishing before the commit is not acceptable, because consumers must not receive events for bookings that were never saved.
Which design ensures that every saved booking eventually produces its event?
a) Write the event to an outbox table in the same local transaction as the booking, and have a separate relay publish it
b) Catch publishing errors inside the request handler, back off and retry the publish several times before returning a response to the caller
c) Use a saga that deletes the booking as a compensating action whenever the publish step fails
d) Have Billing and Loyalty read the stream more often so they receive events sooner
 
08. An e-learning platform's course service on OKE reads its database password from an OCI Vault secret once, at container start-up, and keeps it in memory. The database team rotated the password by adding a new secret version and immediately revoked the old password on the database. Every running pod then failed to open new database connections until it was restarted.
The team wants future rotations to cause no connection failures. Which two changes should be made?
(Choose two.)
a) Re-read the current secret version when the database rejects the password, and on a regular refresh interval
b) Keep the old password valid until every pod has loaded the new version
c) Pin the service's configuration to a specific secret version number so every pod uses the same value
d) Rotate the vault master encryption key that protects the secret at the same time as the password
e) Add a secret expiry rule that blocks retrieval of the secret contents once they expire
 
09. An online learning platform's grading API runs on an OCI API Gateway deployment with an HTTP back end. The deployment's token authentication policy validates JWTs against a remote JWKS URI and a list of allowed issuers, with the default validation failure behavior. Over the weekend the identity team moved all users to a new identity domain and carried the API's audience value over unchanged. Since then every call returns HTTP 401, the back-end servers receive no requests, and the tokens clients send are unexpired.
What is the most likely cause?
a) The gateway stopped trusting the back end's TLS certificate when users moved to the new identity domain
b) The allowed issuers and remote JWKS URI in the policy still reference the old identity domain, so new tokens fail validation
c) The CLIENT_IP rate limit is exceeded, because users now sign in through the new identity domain
d) The CORS policy does not list the new identity domain as an allowed origin, so the gateway refuses each call
 
10. A ride-hailing company added timeouts, retries and a circuit breaker to its dispatch service's calls to a pricing microservice on OCI Kubernetes Engine (OKE). Unit tests of each pattern pass. The team now needs evidence that dispatch keeps accepting ride requests when the pricing service becomes slow or its pods terminate under realistic traffic.
Which testing approach provides this evidence?
a) Run controlled chaos experiments that inject latency and terminate pricing pods while measuring dispatch
b) Add contract tests that check the pricing service's responses against the format dispatch expects
c) Scale the pricing deployment to more replicas before each release
d) Run a load test that raises ride request volume step by step until dispatch response times exceed the target

Answers:

Question: 01

Answer: d

Question: 02

Answer: c

Question: 03

Answer: b

Question: 04

Answer: c

Question: 05

Answer: c, d

Question: 06

Answer: b

Question: 07

Answer: a

Question: 08

Answer: a, b

Question: 09

Answer: b

Question: 10

Answer: a

 

Rating: 4.9 / 5 (77 votes)