Compute practice questions

Direct attachment to VCN Internet Gateways without any intermediate services

Why wrong: Internet gateways route VCN traffic, they do not invoke functions directly.

Scheduled cron jobs running inside block volume snapshots

Why wrong: Snapshots do not execute cron jobs or invoke functions.

VM.Standard1.16

Why wrong: This is a legacy fixed shape with too many resources.

BM.Standard.E3.128

Why wrong: A bare metal shape is vastly overpowered for a simple VM workload.

VM.Standard2.8

Why wrong: Fixed shapes have a preset ratio of OCPUs to memory, which would result in either fewer OCPUs or excessive resource allocation.

Inside the VCN Security List configurations

Why wrong: Security lists govern packet filtering at the subnet level, not VM launch options.

Via the compute instance's operating system kernel parameters

Why wrong: Firmware and NIC virtualization mode are hypervisor-level configurations, not OS kernel parameters.

In the Instance Configuration template parameters

Why wrong: Instance configurations reference images; launch options are properties of the custom image itself.

The instance is terminated, and a new instance is provisioned to replace it, temporarily exceeding the maximum limit if necessary until scaling rules scale it back down.

Why wrong: The pool will terminate the unhealthy instance and provision a replacement within the defined bounds.

The instance is terminated and not replaced because the pool is already at its maximum limit.

Why wrong: Instance pools maintain desired state and will replace terminated/retired instances.

The autoscaling configuration must be manually edited to increase the max limit before the instance can be replaced.

Why wrong: Manual intervention is not required for infrastructure maintenance replacements.

Bare Metal Instances

Why wrong: While bare metal gives you the entire physical server, dedicated virtual machine hosts allow multiple VMs to run on a dedicated host.

Instance Pools

Why wrong: Instance pools are groups of identical instances managed as a single entity, not a multi-tenant isolation feature.

Protected Compute Shapes

Why wrong: Protected compute shapes is not a valid OCI feature name.

Direct polling of Object Storage buckets by the function container

Why wrong: Functions are event-driven and should not continuously poll storage.

Scheduled Cron Jobs in Compute

Why wrong: Cron jobs execute on compute instances, not natively triggering serverless functions via storage events.

API Gateway direct integration

Why wrong: API Gateway triggers functions via HTTP requests, not direct storage events.

VCN Route Tables

Why wrong: Route tables control network traffic routing, not physical hardware placement.

Kubernetes Horizontal Pod Autoscaler

Why wrong: HPA scales the number of pods based on metrics, not physical placement of nodes.

Pod Security Policies

Why wrong: Pod Security Policies govern security settings for pods, not node placement.

Block Volume Backup

Why wrong: Block volume backups are for data protection and disaster recovery, not directly deployable as standardized templates across multiple regions easily without cloning.

Boot Volume Clone

Why wrong: Cloning creates a separate point-in-time copy of a single volume rather than a reusable image template.

Instance Configuration

Why wrong: Instance configurations define the settings to create instances but require an existing image source.

BM.Standard3.64

Why wrong: Standard shapes use network-attached block volumes and do not include local NVMe storage.

VM.Standard.E4.Flex

Why wrong: This is a flexible virtual machine shape and does not include local NVMe storage.

BM.GPU3.8

Why wrong: GPU shapes are optimized for parallel processing and machine learning workloads, not general database logging.

The cluster becomes completely inaccessible, and all pods are evicted permanently.

Why wrong: Control plane upgrades in managed Kubernetes services maintain control plane availability and do not destroy pods.

Control plane upgrades are blocked until all worker nodes are upgraded first.

Why wrong: Control plane is upgraded independently before worker nodes.

All worker nodes are automatically terminated and recreated with the new version.

Why wrong: Worker nodes are not automatically terminated during a control plane-only upgrade.

VM.Standard.E4.Flex

Why wrong: E4 shapes use AMD EPYC processors, not ARM.

BM.DenseIO2.52

Why wrong: Dense I/O bare metal shapes are specialized for high-performance storage and cost significantly more.

VM.Standard3.Flex

Why wrong: Standard3 shapes use Intel processors.

terraform

Why wrong: Terraform is an Infrastructure as Code (IaC) tool.

oci-cli

Why wrong: The standard OCI CLI manages core infrastructure services (compute, VCN, IAM), while the Fn CLI is specialized for functions.

kubectl

Why wrong: kubectl is used to manage Kubernetes clusters.

The Instance Pool capacity limits

Why wrong: Capacity limits define min and max instances, not the time interval between scaling actions.

The VCN DHCP Options timeout value

Why wrong: DHCP options manage IP configurations for instances, unrelated to autoscaling timing.

The Monitoring service metric polling interval

Why wrong: Metric emission intervals are fixed by OCI Monitoring, whereas autoscaling uses threshold evaluation windows and cooldowns.

Event-driven autoscaling policy

Why wrong: Event-driven autoscaling is not a standard autoscaling policy type in OCI Instance Pools.

Schedule-based autoscaling policy

Why wrong: Schedule-based policies scale based on specific times and dates, not metric thresholds.

Manual capacity override policy

Why wrong: Manual overrides are administrative adjustments, not an automated scaling policy type.

Use an API Gateway with a public endpoint and route database queries through HTTP headers.

Why wrong: Routing database credentials and raw queries through HTTP headers is insecure and architecturally incorrect.

Configure a public IP address on the function execution environment.

Why wrong: Functions execute in managed serverless environments and do not support attaching public IP addresses directly.

Functions cannot access private subnets; you must use an internet gateway.

Why wrong: OCI Functions fully support VCN integration for private subnet access.

Use the OCI Database Migration service to copy the operating system boot sectors.

Why wrong: Database Migration service is for databases, not OS boot volumes.

Detach the boot volume, attach it to a secondary compute instance, and use rsync over a VCN remote peering connection.

Why wrong: Boot volumes cannot be attached as secondary data volumes to arbitrary running instances easily, and OS volumes require image export/import mechanisms for cross-region migration.

Create a block volume backup, copy the backup directly to the destination region using the native backup copy feature, and restore it as a boot volume.

Why wrong: Block volume backups cannot be directly copied across regions without exporting them or using cross-region image copying.

The custom OS username and SSH password

Why wrong: SSH keys are defined in the instance configuration, not during instance pool creation.

The external DNS provider settings

Why wrong: DNS configuration is handled at the VCN level, not during instance pool creation.

The new root password for the boot volumes

Why wrong: OCI Linux instances use SSH key-based authentication, not root passwords.

OCI Functions

Why wrong: OCI Functions is a serverless FaaS platform, not a Kubernetes management service.

Compute Instance Pools

Why wrong: Instance pools manage groups of virtual machines, not Kubernetes pods.

Container Instances

Why wrong: Container Instances run containers without managing servers or Kubernetes, but it is not a Kubernetes cluster service.

Temporarily attach a public IP address to the worker node and connect via SSH.

Why wrong: Attaching public IPs to managed OKE worker nodes post-creation is insecure and not recommended best practice.

Modify the VCN security list to allow inbound SSH from 0.0.0.0/0 to the private subnet.

Why wrong: Opening SSH from the entire internet to private subnets violates security compliance standards.

Run `kubectl exec` with privileged mode to gain root access to the underlying node's operating system.

Why wrong: `kubectl exec` runs commands inside a container namespace, not the host operating system.

Terminate the instance, detach the boot volume, attach it to another instance, expand it, and reattach it.

Why wrong: Boot volumes can be resized online in OCI without termination or detaching.

Reboot the instance; OCI automatically detects the new size and extends the file system partitions automatically.

Why wrong: OCI resizes the block storage device at the hypervisor level, but the OS partition and file system must be extended manually or via OCI cloud-init features.

Create a new block volume, mount it as the root directory, and delete the old boot volume.

Why wrong: Root directory cannot be dynamically migrated to a new block volume in this manner.