1Z0-997-26
Free 1Z0-997-26 practice questions
10 questions from our 1Z0-997-26 practice tests, each with the correct answer and a written explanation. Read them with the answers shown, or switch to quiz mode and answer first.
Question 1 of 10, Architect High Availability and Disaster Recovery Solutions
An insurance company uses OCI Full Stack Disaster Recovery to protect an application between Frankfurt and Amsterdam. Before failover, the application tier must quiesce a message queue by running a shell script on an instance, and after failover a serverless routine must update a third-party DNS provider. How should the architect build this into the DR plan?
- AAdd user-defined plan groups that run the script on the instance through Run Command and invoke an OCI Function for the DNS update Correct answer
- BStore the script and DNS logic in a Resource Manager stack and run an apply job manually after Full Stack Disaster Recovery finishes
- CCreate an Events rule on the DR protection group that triggers the script and DNS update when the plan execution starts
- DEdit the built-in plan groups so they call the script and the DNS update, because built-in steps can be changed to run any command
Answer and explanation: Full Stack Disaster Recovery lets architects add user-defined plan groups whose steps run scripts on Compute instances (through Run Command) or invoke OCI Functions, and these steps run in order with the built-in steps. Built-in plan groups are generated by the service and cannot be rewritten to run arbitrary commands. An Events rule runs outside the plan, so its timing and success are not tracked as plan steps. A manual Resource Manager job after failover breaks the single automated, ordered runbook that the plan provides.
Question 2 of 10, Architect High Availability and Disaster Recovery Solutions
A media company stores customer-uploaded images in an Object Storage bucket in Ashburn and needs a copy in Phoenix for regional disaster recovery. The copy must stay up to date automatically and must not be changed by applications in Phoenix during normal operation. What should the architect configure?
- AA pre-authenticated request in Phoenix that copies objects from Ashburn hourly
- BA replication policy on the Ashburn bucket that targets a bucket in Phoenix Correct answer
- CA lifecycle policy rule on the Ashburn bucket that archives objects to Phoenix
- DA Connector Hub connector that moves objects from Ashburn into Phoenix nightly
Answer and explanation: An Object Storage replication policy asynchronously copies new and updated objects to a destination bucket in another region, and the destination bucket is read-only while it is a replication target. Lifecycle policies archive or delete objects within the same bucket and do not copy them to another region. A pre-authenticated request only grants access to objects; it does not copy anything on a schedule. Connector Hub moves logs, metrics and stream data, not existing objects between buckets.
Question 3 of 10, Architect Cloud-Native Solutions
An order-processing system must decouple a front end from a back end. Each message must be processed independently, failed messages must be retried and eventually moved aside for analysis, and consumers should not need to track offsets. Which service should the architect use?
- AOCI Queue, with a dead letter queue and a maximum delivery attempts value Correct answer
- BOCI Streaming, where consumers commit offsets in a consumer group after processing
- COCI Notifications with an HTTPS subscription to the back end
- DOCI Events, with a rule that forwards each order to a function and retries on failure
Answer and explanation: OCI Queue delivers messages to consumers individually, handles visibility timeouts and retries, and moves messages that exceed maximum deliveries to a dead letter queue, with no offsets to manage. Streaming is a partitioned log in which consumers track offsets and messages are not deleted individually. Notifications pushes messages but has no dead letter handling suited to order processing. Events reacts to OCI resource changes rather than carrying application messages.
Question 4 of 10, Architect Cloud-Native Solutions
A company uses OCI DevOps to deploy a web application to OKE. The release team wants new versions to receive a small percentage of production traffic first, with a manual approval before full rollout. Which deployment strategy should the pipeline use?
- AA deployment stage that applies the manifest and then runs a shell stage for tests
- BA blue-green deployment stage that switches all traffic at once after deployment
- CA rolling deployment stage that replaces pods in batches without any approval
- DA canary deployment stage with traffic shifting and a manual approval stage Correct answer
Answer and explanation: The canary strategy in OCI DevOps deploys the new version alongside the old one, shifts a portion of traffic to it, and supports an approval before promoting it fully. Rolling updates replace pods in batches but do not hold a defined percentage of traffic for validation. Blue-green moves all traffic at once after the switch, rather than a small percentage first. Applying a manifest and running tests does not control the share of live traffic.
Question 5 of 10, Architect Security Solutions
A company wants to inspect all traffic between spoke VCNs and the internet for intrusion attempts and to block access to certain URL categories. Which design should the architect use?
- AUse Cloud Guard with the threat detector recipe on each spoke compartment to block intrusion attempts
- BEnable VCN flow logs on all spoke subnets and alert through Logging when suspicious flows appear
- COCI Network Firewall in a hub VCN, with DRG and VCN route tables steering spoke traffic through it Correct answer
- DAttach a WAF policy to every load balancer in the spoke VCNs and enable the protection rule sets
Answer and explanation: OCI Network Firewall provides intrusion detection and prevention and URL filtering, and in a hub-and-spoke design the DRG and route tables steer spoke traffic through it. WAF protects inbound HTTP applications behind load balancers and does not filter outbound URL categories. Flow logs only record traffic after the fact and do not block anything. Cloud Guard finds risky configurations and activity but is not an inline traffic inspection device.
Question 6 of 10, Architect Security Solutions
An architect is designing compartments for a tenancy that hosts several applications, each with development and production environments. Network teams manage shared VCNs, and security teams manage Vaults centrally. Which structure follows OCI landing zone best practice?
- AOne compartment per application that holds that application's network, vaults and both of its environments
- BA compartment per individual user, so each person's resources are isolated from everyone else
- CSeparate network, security and per-environment workload compartments under one parent Correct answer
- DAll resources in the root compartment, separated by tags that identify the application and environment
Answer and explanation: OCI landing zone guidance separates shared network and security resources from workload compartments, under an enclosing compartment, so each team gets permissions only where it needs them. Mixing networks and vaults into each application compartment prevents central network and security management. Putting everything in the root compartment makes it impossible to scope policies cleanly, since tags alone are a weaker control. Per-user compartments do not reflect ownership by teams or environments.
Question 7 of 10, Architecting, Implementing, and Operating Databases in OCI
An application on Autonomous AI Database Serverless must survive the loss of an availability domain with automatic failover and minimal data loss. Cost should stay low, and protection against regional disasters is not required. What should the architect enable?
- AAutonomous Data Guard with a local standby database Correct answer
- BAutonomous Data Guard with a cross-region standby only, in a paired region
- CA refreshable clone in another compartment that refreshes automatically on a schedule
- DBackup-based disaster recovery to another region
Answer and explanation: A local Autonomous Data Guard standby runs in a different availability domain (or separate infrastructure in single-AD regions) and supports automatic failover with minimal data loss. A cross-region standby addresses regional disasters, which is not required, and costs more. Backup-based DR restores from backups and has a longer recovery time with higher data loss. A refreshable clone is a read-only copy for reporting that is refreshed periodically and is not a failover target.
Question 8 of 10, Architecting, Implementing, and Operating Databases in OCI
A company runs Oracle databases on Base Database Service and must protect them with backups that capture changes in real time so that data loss after a failure is close to zero, without building a separate standby database. What should the architect enable?
- AAutomatic backups to Object Storage with a short weekly full backup window
- BA Data Guard association with a standby database system in a fault domain
- CBlock volume backups of the database storage taken every few minutes by policy
- DAutonomous Recovery Service with real-time data protection for the databases Correct answer
Answer and explanation: Autonomous Recovery Service with real-time data protection continuously ships redo to the recovery service, giving near-zero data loss on restore without a standby database. Object Storage backups depend on archive log backup intervals and lose more data. Block volume backups of database storage are not database-aware and are not a supported way to back up Base Database systems. Data Guard achieves low data loss but it is a standby database, which the requirement excludes.
Question 9 of 10, Implementing Observability Solutions
Operations wants to be paged when average CPU utilization of any instance in a production compartment stays above 85% for 10 minutes, and they want one notification per affected instance. How should the architect configure the alarm?
- ACreate one alarm per instance by tagging each instance and choose the tag as the alarm dimension
- BA CpuUtilization mean query with a 10-minute pending duration and split notifications per metric stream Correct answer
- CUse a CpuUtilization max query with no pending duration and add one Notifications topic per instance
- DUse a CpuUtilization mean query grouped across the compartment and set a repeat notification of 10 minutes
Answer and explanation: An alarm on mean CPU with a pending duration of 10 minutes fires only when the condition persists, and splitting notifications sends a separate message for each metric stream, which is each instance. A max query with no pending duration fires on short spikes. Aggregating across the compartment hides which instance is affected, and repeat notification only resends messages. Creating an alarm per instance is hard to maintain and tags are not alarm metric dimensions.
Question 10 of 10, Implementing Observability Solutions
Security must be alerted within minutes whenever someone changes an IAM policy in the tenancy. Which design meets this requirement with the LEAST custom code?
- AA Cloud Guard target on the root compartment that blocks every IAM policy change
- BA daily Logging search for policy changes, exported manually to the security team
- CA Function that polls the Audit API every five minutes for policy changes and sends mail
- DAn Events rule matching IAM policy update events, with a Notifications topic action Correct answer
Answer and explanation: IAM operations emit events, and an Events rule filtering on policy create, update and delete events can send them to a Notifications topic without code. Polling the Audit API requires custom code and scheduling. A daily manual search does not deliver alerts within minutes. Cloud Guard detects issues and can remediate some, but it is not meant to block every policy change and that would stop legitimate administration.