# Free 1Z0-1072-26 practice questions

> 10 free 1Z0-1072-26 practice questions with the correct answer and an explanation.

Page: https://devopscodelab.com/exams/oracle-1z0-1072-26/free-practice-questions/
Practice exams: https://devopscodelab.com/exams/oracle-1z0-1072-26/

### Question 1 of 10

Project teams each own a compartment under Projects. The central team wants each team's admin group to manage everything in its own compartment, including writing policies there, but never affect other compartments. Which statement should be written for the team Alpha?

- A. Allow group AlphaAdmins to manage all-resources in compartment Projects
- B. Allow group AlphaAdmins to manage all-resources in compartment Projects:Alpha
- C. Allow group AlphaAdmins to manage policies in tenancy where target.compartment.name = 'Alpha'
- D. Allow group AlphaAdmins to use all-resources in tenancy where request.region = 'Alpha'

**Answer:** B

**Explanation:** Granting manage all-resources on the Alpha compartment, referenced by its path, includes the ability to manage policies that are attached in that compartment and applies only to Alpha and its children. Granting it on Projects would cover every team, a tenancy-level policies statement with a name condition is not how compartment administration is normally scoped and leaves other resources unmanaged, and request.region checks a region rather than a compartment.

### Question 2 of 10

A policy attached to the root compartment grants the group Ops manage all-resources in compartment Shared. The owner of the child compartment Shared:Payments wants to block Ops from Payments. What can the owner do in Payments?

- A. Enable a Security Zone on Payments, which removes all inherited permissions from groups
- B. Attach a deny statement to Payments, which overrides the permissions that are inherited from Shared
- C. Add a compartment quota in Payments that sets the Ops group's allowed actions to zero
- D. Nothing in Payments, because policies only allow and inherited access cannot be denied below

**Answer:** D

**Explanation:** OCI IAM policies are allow-only, so access inherited from a parent compartment cannot be revoked by adding something in the child; the parent policy must be changed (for example with a condition excluding Payments) or Payments moved. Deny statements are not supported, quotas limit resource consumption rather than group actions, and Security Zones enforce configuration rules rather than removing permissions.

### Question 3 of 10

A company has three spoke VCNs and an on-premises network connected by FastConnect. All spoke-to-spoke and spoke-to-on-premises traffic must pass through a network firewall in a hub VCN. Which design meets this requirement with the fewest gateways to manage?

- A. Create a local peering gateway pair between every spoke VCN and the hub VCN, and route the FastConnect virtual circuit to a second DRG that is attached only to the hub VCN, so the hub sees every flow
- B. Attach all VCNs and the virtual circuit to one DRG, use DRG route tables that send spoke and on-premises traffic to the hub VCN attachment, and steer it to the firewall with VCN ingress routing
- C. Attach each spoke VCN to its own DRG, create remote peering connections between these DRGs, and terminate the FastConnect circuit on the DRG of the first spoke
- D. Attach all VCNs and the virtual circuit to one DRG and rely on its default route table, which already sends every attachment's traffic to the hub VCN first

**Answer:** B

**Explanation:** A single upgraded DRG can attach multiple VCNs and virtual circuits, and custom DRG route tables plus a route table on the hub VCN attachment steer traffic through the firewall, giving a hub-and-spoke transit design. LPG meshes add many gateways and cannot carry on-premises transit cleanly, remote peering connections are for cross-region DRGs, and the DRG default route tables import all routes and send traffic directly between attachments, bypassing the firewall.

### Question 4 of 10 (choose 2)

An on-premises DNS server must resolve names in a private DNS zone attached to a VCN, and VCN instances must resolve names in the corporate domain corp.example.local. Which configuration of the VCN's private DNS resolver is required? (Choose two.)

- A. A forwarding endpoint plus a resolver rule that sends corp.example.local queries to the on-premises DNS servers
- B. A traffic management steering policy of type failover that answers for the private zone names
- C. An internet gateway route so the VCN resolver can reach the on-premises DNS servers by public IP
- D. A public DNS zone named corp.example.local with A records copied from the on-premises servers
- E. A listening endpoint on the resolver that on-premises DNS servers forward queries to

**Answer:** A, E

**Explanation:** A listening endpoint receives queries forwarded from on-premises DNS, and a forwarding endpoint with a conditional resolver rule sends corp.example.local queries out to on-premises servers over FastConnect or VPN. A public zone would expose internal records and duplicate data, steering policies answer public queries rather than bridging private resolution, and resolver traffic to on-premises should go over private connectivity, not the internet.

### Question 5 of 10

On-premises servers connected by FastConnect private peering must upload backups to Object Storage without using the internet. Which design enables this?

- A. Transit routing: the DRG sends the traffic to a hub VCN, whose DRG-attachment route table sends Oracle Services Network traffic to a service gateway
- B. Add a route on the on-premises router that sends the Object Storage public CIDRs to the FastConnect private virtual circuit with no OCI-side changes
- C. Use a local peering gateway between the DRG and the service gateway so that on-premises traffic reaches Oracle services directly
- D. Create a NAT gateway in a VCN attached to the DRG, and route all on-premises traffic for Object Storage through it so it reaches the public regional endpoint

**Answer:** A

**Explanation:** Private access from on-premises to Oracle services uses transit routing: traffic enters the VCN through the DRG attachment and a route table associated with that attachment sends it to the service gateway (FastConnect public peering is the alternative). Without OCI-side routing the private circuit does not reach Object Storage, a NAT gateway leads to public endpoints, and LPGs peer VCNs rather than connecting a DRG to a service gateway.

### Question 6 of 10

A security review found that applications on some instances can query the instance metadata service using the legacy endpoint that does not require request headers. How should the team harden these instances?

- A. Add a security list egress rule that blocks traffic to 169.254.169.254 from every instance in the subnet
- B. Remove the default route to the internet gateway so that the metadata service is no longer reachable from any VNIC
- C. Rotate the instance principal certificates every day so stolen metadata responses expire much faster
- D. Disable the legacy IMDS v1 endpoints so that only IMDS v2 requests with the authorization header are accepted

**Answer:** D

**Explanation:** OCI lets you disable the legacy IMDS v1 endpoints so only IMDS v2, which requires an Authorization header, is served, reducing SSRF-style abuse. The metadata service is link-local and does not use the internet gateway, blocking 169.254.169.254 would also break the OS and agents that need metadata, and certificate rotation is automatic and does not address the legacy endpoint.

### Question 7 of 10

Which Object Storage tier should be used for compliance records that are rarely accessed, must be kept for years, and can tolerate waiting about an hour before data becomes readable?

- A. Infrequent Access Storage
- B. Block Volume Lower Cost
- C. Archive Storage
- D. Standard Storage

**Answer:** C

**Explanation:** Archive Storage is the lowest-cost tier for long-term retention and requires a restore before objects can be read, which takes on the order of an hour. Standard is for frequently accessed data, Infrequent Access costs more per GB than Archive and has retrieval fees but immediate access, and the Block Volume lower cost performance level is for volumes attached to instances, not object archives.

### Question 8 of 10

Several Linux instances in different availability domains must mount the same POSIX file system over NFS for shared configuration files. Which OCI service provides this with the LEAST administration?

- A. An Object Storage bucket mounted on each instance using a pre-authenticated request URL
- B. A local NVMe drive on one instance shared to the others through a self-managed NFS server
- C. A block volume attached as shareable read/write to every instance and formatted with an ext4 file system on each one
- D. File Storage with a mount target in a regional subnet and an export that the instances mount

**Answer:** D

**Explanation:** File Storage is a managed NFS service; instances in any AD in the region can mount an export through a mount target. A multi-attach block volume requires a cluster-aware file system, since ext4 would corrupt with concurrent writers, and block volumes are AD-local. Buckets are not POSIX file systems, and a self-managed NFS server on local NVMe adds administration and a single point of failure.

### Question 9 of 10

An application team needs a MySQL database that survives the loss of an availability domain with automatic failover and no data loss. Which MySQL HeatWave configuration should be chosen?

- A. A standalone DB system plus a read replica in the same AD that the application promotes manually after failure
- B. A standalone DB system with automatic backups enabled and a backup retention period of 35 days in the region
- C. A standalone DB system with an attached HeatWave cluster that processes queries across multiple nodes in memory
- D. A high-availability DB system, which places three instances across ADs or fault domains with automatic failover

**Answer:** D

**Explanation:** A high-availability MySQL HeatWave DB system uses three instances with group replication across availability domains (or fault domains in single-AD regions) and fails over automatically without data loss. Backups require a restore, a HeatWave cluster accelerates analytics but is not HA for the primary, and manual promotion of a same-AD replica is neither automatic nor AD-resilient.

### Question 10 of 10 (choose 2)

Operations wants an email and a PagerDuty page whenever average CPU on any instance in the Prod compartment stays above 85 percent for 5 minutes. Which combination of services should be used? (Choose two.)

- A. An Events rule that matches the com.oraclecloud.computeapi.updateinstance event type
- B. A Notifications topic with email and PagerDuty subscriptions set as the alarm destination
- C. A Monitoring alarm with an MQL query on the CpuUtilization metric and a trigger delay of 5 minutes
- D. A Logging custom log that captures the top command output from each instance agent
- E. A Connector Hub connector from the Audit log to an Object Storage bucket for review

**Answer:** B, C

**Explanation:** A Monitoring alarm evaluates the CpuUtilization metric with a 5-minute pending duration, and its destination is a Notifications topic whose subscriptions deliver email and PagerDuty messages. Events reacts to resource state changes rather than metric thresholds, a custom log of top output would need extra processing to alert, and archiving Audit logs does not evaluate CPU.

Ready for a timed run? Take the free 15-question 1Z0-1072-26 mini exam: https://devopscodelab.com/app/exams/oracle-1z0-1072-26/mini/

DevOpsCode Lab is not affiliated with Oracle. 1Z0-1072-26 and related names are trademarks of their owners. We write every question ourselves; these are not real exam questions.
