I85PREP
⌘ K
Huawei HCIP-Cloud Service Solutions Architect study materials
Huawei HCIP-Cloud Service Solutions Architect study resources

Huawei HCIP-Cloud Service Solutions Architect practice questions

One question at a time. Build your Huawei HCIP-Cloud Service Solutions Architect understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which Huawei Cloud service provides resizable virtual computing resources on which you install and run your own operating systems and applications?

  1. Bare Metal Server (BMS)
  2. Elastic Cloud Server (ECS) Correct answer.
  3. Image Management Service (IMS)
  4. Cloud Container Engine (CCE)

Explanation

ECS delivers virtual machines whose vCPU and memory can be selected and changed, and which run a customer-chosen image with full OS-level control. It is the general-purpose compute building block used when workloads must run on a conventional server. BMS, CCE and IMS play different roles in the compute stack.

Question 2Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A financial analytics workload must run directly on physical hardware with no virtualization layer, and its operating system cannot be modified. Which Huawei Cloud service fits?

  1. Elastic Cloud Server
  2. Cloud Container Engine
  3. Bare Metal Server Correct answer.
  4. FunctionGraph

Explanation

Bare Metal Server gives a dedicated physical server with no hypervisor in the path, so the workload sees hardware directly and can keep an unmodified OS. This suits performance-sensitive, license-bound or non-virtualizable applications. ECS, CCE and FunctionGraph all introduce abstraction layers the scenario excludes.

Question 3Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A media company must keep petabytes of images and video, serve them over HTTP/HTTPS, and pay only for the capacity and traffic consumed. Which service should it use?

  1. Elastic Volume Service (EVS)
  2. Object Storage Service (OBS) Correct answer.
  3. Scalable File Service (SFS)
  4. Relational Database Service (RDS)

Explanation

OBS is massively scalable object storage accessed through RESTful HTTP/HTTPS APIs, with pay-per-use billing on stored capacity, requests and traffic. It is the standard choice for unstructured content such as media files and backups. File, block and relational services are not designed for internet-facing object access at this scale.

Question 4Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Several ECS instances in one VPC must read and write the same directory tree at the same time, exactly as they would use a local file system. Which service provides this?

  1. Scalable File Service (SFS) Correct answer.
  2. Elastic Volume Service (EVS)
  3. Cloud Container Engine (CCE)
  4. Object Storage Service (OBS)

Explanation

SFS offers a shared, POSIX-compliant file system that many compute instances in the same network can mount simultaneously, which suits shared content directories and home directories. OBS and EVS do not present a standard shared POSIX mount to multiple servers. CCE is a container orchestration service, not a storage service.

Question 5Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which Huawei Cloud service lets you build a logically isolated network with your own CIDR blocks, subnets and security groups?

  1. Direct Connect
  2. NAT Gateway
  3. Elastic IP (EIP)
  4. Virtual Private Cloud (VPC) Correct answer.

Explanation

A VPC is the isolated virtual network boundary in which you define address ranges, subnets, route tables and security groups for your cloud resources. The other services attach to or connect networks but do not define the isolated network itself.

Question 6Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A web application runs on ECS instances spread across two AZs and needs a single entry point that spreads requests and stops sending traffic to unhealthy instances. Which service should be used?

  1. VPN Gateway
  2. Elastic Load Balance Correct answer.
  3. Elastic IP
  4. NAT Gateway

Explanation

Elastic Load Balance provides one service address, distributes traffic across backends and removes unhealthy instances through health checks, which is the standard way to expose a multi-AZ web tier. NAT, EIP and VPN services handle address translation or connectivity rather than traffic distribution and health checking.

Question 7Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

An instance needs a static public address that can be detached from a failed ECS and rebound to its replacement within minutes. Which service provides this?

  1. NAT Gateway
  2. Elastic IP Correct answer.
  3. Virtual Private Cloud
  4. Elastic Load Balance

Explanation

An Elastic IP is a static public address that is allocated to your account and can be unbound from one resource and rebound to another, which makes it suitable for quick failover of a known public endpoint. NAT and ELB provide translation or distribution rather than a movable single address, and a VPC is a network boundary.

Question 8Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Demand for a workload changes hour by hour, and the architect wants instances added or removed automatically based on CPU utilization. Which service provides this?

  1. Auto Scaling Correct answer.
  2. Cloud Eye
  3. Cloud Container Engine
  4. Elastic Load Balance

Explanation

Auto Scaling manages a group of instances and adjusts its size according to scheduled or metric-based policies, which is how elastic capacity is delivered. Cloud Eye supplies the metrics that feed such policies, and ELB distributes traffic, but neither adds or removes instances.

Question 9Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which service offers a managed Kubernetes environment for deploying, scaling and orchestrating containerized applications?

  1. Software Repository for Container (SWR)
  2. Cloud Container Instance (CCI)
  3. Bare Metal Server (BMS)
  4. Cloud Container Engine (CCE) Correct answer.

Explanation

CCE is the managed Kubernetes service: the control plane is operated by the provider while customers deploy workloads onto node pools. SWR stores images and CCI runs containers without cluster management, but neither provides a Kubernetes control plane. BMS supplies physical servers.

Question 10Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A team wants small pieces of code to run in response to object upload events, with no servers to manage and billing based on execution time. Which service fits?

  1. Cloud Container Engine
  2. FunctionGraph Correct answer.
  3. Auto Scaling
  4. Elastic Cloud Server

Explanation

FunctionGraph is the serverless function service: code runs on demand triggered by events such as OBS object creation, and billing follows execution count and duration rather than provisioned servers. The other options require you to own and size running capacity.

Question 11Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which service supplies managed MySQL and PostgreSQL instances where the provider handles operating system patching and automated backups?

  1. Document Database Service
  2. Relational Database Service Correct answer.
  3. Elastic Cloud Server
  4. Distributed Cache Service

Explanation

RDS is the managed relational database service and supports engines such as MySQL, PostgreSQL and SQL Server, with automated backup, patching and failover performed by the platform. Running a database on ECS would leave those tasks to the customer, and the other services target non-relational or caching workloads.

Question 12Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

An application stores flexible-schema JSON documents and needs an API that is compatible with MongoDB. Which managed Huawei Cloud service should be used?

  1. Document Database Service Correct answer.
  2. Data Warehouse Service
  3. Relational Database Service
  4. Distributed Cache Service

Explanation

DDS is the managed document database service, offering MongoDB-compatible interfaces for schema-flexible JSON-style data with replication and backup handled by the platform. RDS, DCS and the data warehouse service address relational, caching and analytical workloads respectively.

Question 13Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which service controls who can access which cloud resources by managing users, user groups, policies and federated identities?

  1. Cloud Trace Service
  2. Data Encryption Workshop
  3. Cloud Eye
  4. Identity and Access Management Correct answer.

Explanation

IAM is the identity and authorization service where users, groups, roles and fine-grained policies are defined, and where enterprise identities can be federated. It answers who may perform which action on which resource. Auditing, encryption and monitoring services cover different concerns.

Question 14Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

An online store must block SQL injection and cross-site scripting attacks aimed at its web front end. Which service is designed for this?

  1. Cloud Firewall
  2. Web Application Firewall Correct answer.
  3. Anti-DDoS
  4. Host Security Service

Explanation

A Web Application Firewall inspects HTTP/HTTPS requests and blocks application-layer attacks such as SQL injection and cross-site scripting before they reach the web tier. Anti-DDoS, HSS and network firewalls operate at other layers and do not parse application payloads for these patterns.

Question 15Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Which service collects metrics from resources such as ECS and RDS and raises alarms that can send notifications or trigger scaling actions?

  1. Log Tank Service
  2. Simple Message Notification
  3. Cloud Eye Correct answer.
  4. Cloud Trace Service

Explanation

Cloud Eye is the monitoring service that gathers resource metrics, evaluates alarm rules and can act on them, for example by notifying subscribers or invoking Auto Scaling. Log, audit and messaging services collect different data or deliver notifications rather than evaluating metrics.

Question 16Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

An architect proposes attaching one EVS data disk to ECS instances located in three different AZs of the same region. Which statement describes the actual constraint?

  1. The disk can be attached to instances in any AZ of the same region
  2. The disk is AZ-scoped and attaches only to instances in the same AZ Correct answer.
  3. The disk can span AZs once cross-AZ replication is enabled
  4. Only SSD-type disks may be attached across availability zones

Explanation

EVS is zonal storage: a volume is created in an AZ and can only be attached to instances in that same AZ, although it may be shared by several instances within the zone. Moving data elsewhere requires snapshots or replication, which create separate copies rather than extending one disk across zones.

Question 17Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A front end writes jobs to a back end that becomes slow during peaks and drops requests. Which change best decouples the two tiers?

  1. Increase the session timeout on the load balancer
  2. Buffer jobs in a message queue between the tiers Correct answer.
  3. Attach a larger block storage disk to the back end
  4. Upgrade the front-end instances to a larger flavor

Explanation

Inserting a message queue between tiers lets the front end enqueue work and return quickly while the back end consumes at its own pace, absorbing bursts without failing requests. Larger instances or disks raise capacity temporarily but keep the synchronous coupling that causes the drops.

Question 18Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A company retains twelve months of backups that are almost never read but must stay durable and retrievable on request. Which choice lowers storage cost the most?

  1. Store the backups in a high-throughput SFS Turbo file system
  2. Keep the backups on high-performance EVS volumes attached to ECS
  3. Keep the backups on the system disks of the application servers
  4. Store the backups in an OBS infrequent access tier Correct answer.

Explanation

Object storage offers tiers priced for access frequency, and an infrequent access tier trades a slightly higher retrieval charge for a lower storage rate, which fits rarely read but retained data. Block and file systems provisioned for active workloads cost more per gigabyte for cold data.

Question 19Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

A relational database must survive the loss of an entire AZ with near-zero data loss. Which deployment satisfies this?

  1. An RDS primary/standby pair in two AZs of the region Correct answer.
  2. A single RDS instance with hourly automated backups
  3. An RDS read replica placed in the primary's own AZ
  4. Nightly logical dumps exported manually to object storage

Explanation

A primary/standby deployment across two AZs replicates changes synchronously, so losing one zone does not lose committed transactions and the standby can take over. Backups and same-AZ replicas leave a recovery window or share the same failure domain.

Question 20Cloud Service Solution Design Principles and Huawei Cloud Service Portfolio Overview

Web-tier instances store user sessions in local memory, so adding instances breaks user logins. Which change best enables horizontal scaling?

  1. Spread the tier across more availability zones
  2. Externalize session state to a shared cache Correct answer.
  3. Enable sticky sessions on the elastic load balancer
  4. Replicate the local session files to every running instance

Explanation

Making the tier stateless by moving session data into a shared store lets any instance serve any request, which is the prerequisite for freely adding or removing instances. Sticky sessions and file replication preserve the coupling to a specific instance and add fragility.

Question 21Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

An in-memory database on Huawei Cloud requires a large amount of RAM and moderate CPU. Which ECS flavor type is most appropriate?

  1. Memory-optimized Correct answer.
  2. GPU-accelerated
  3. Disk-intensive
  4. General computing

Explanation

Memory-optimized flavors provide a higher memory-to-vCPU ratio, which suits in-memory databases, caching, and memory-intensive analytics. General computing balances CPU and memory, while disk-intensive and GPU-accelerated flavors target storage throughput and parallel compute respectively.

Question 22Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which statement best describes a key difference between a Bare Metal Server (BMS) and an Elastic Cloud Server (ECS)?

  1. BMS supports only Windows images.
  2. BMS must be deployed outside a VPC.
  3. BMS always shares a physical host with other tenants.
  4. BMS exposes physical hardware directly. Correct answer.

Explanation

A BMS gives the operating system direct access to physical hardware, avoiding a hypervisor layer and its overhead. An ECS runs on virtualized resources, while both BMS and ECS can be placed in a VPC and can run supported Linux or Windows images.

Question 23Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

What is the primary purpose of Auto Scaling in Huawei Cloud?

  1. To distribute traffic across multiple regions
  2. To provide a managed Kubernetes control plane
  3. To encrypt data stored on ECS disks
  4. To scale compute capacity automatically Correct answer.

Explanation

Auto Scaling adjusts the number of computing instances according to configured policies, schedules, or alarms. It helps match capacity to demand without manual intervention and can reduce cost during low-demand periods.

Question 24Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

CCE is Huawei Cloud's managed container service built on which orchestration platform?

  1. Kubernetes Correct answer.
  2. OpenStack
  3. Apache Mesos
  4. Docker Swarm

Explanation

Cloud Container Engine (CCE) is a managed Kubernetes service that provides cluster management, workload scheduling, and integration with other Huawei Cloud services. Kubernetes is the orchestration platform behind CCE.

Question 25Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which EVS disk type is intended for workloads that need the highest IOPS and lowest latency?

  1. High I/O
  2. Object storage
  3. Extreme SSD Correct answer.
  4. Common I/O

Explanation

Extreme SSD is the highest-performance EVS block disk type and is suited to latency-sensitive databases and high-IOPS workloads. Common I/O and High I/O provide lower performance tiers, while object storage is not an EVS block disk type.

Question 26Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

After creating a Linux ECS using a key pair, how does the user authenticate for SSH login?

  1. With an IAM token instead of SSH
  2. With a shared secret sent by SMS
  3. With the matching private key Correct answer.
  4. With the ECS administrator password

Explanation

When a key pair is used, the public key is placed on the ECS and the user authenticates with the corresponding private key. Password login may be disabled or may require separate configuration.

Question 27Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

What is the main role of SWR in a container architecture?

  1. To terminate TLS traffic at the edge
  2. To schedule pods onto cluster nodes
  3. To store and manage container images Correct answer.
  4. To provide persistent block storage for pods

Explanation

SoftWare Repository for Container (SWR) is a container image repository service used to store, manage, and distribute container images. CCE and CCI can pull images from SWR when starting containers.

Question 28Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

In Auto Scaling, what does the cooldown period do?

  1. It defines how long an instance can remain unhealthy
  2. It blocks new scaling actions after a scaling action Correct answer.
  3. It delays instance boot until the image is ready
  4. It pauses billing for scaled-in instances

Explanation

The cooldown period prevents rapid, repeated scaling actions after a scaling event. During cooldown, new scaling requests triggered by policies are typically not executed, which helps avoid oscillation and unnecessary instance launches.

Question 29Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

When an ECS instance is stopped but not deleted, what happens to the data on its system disk?

  1. It is retained on the associated EVS disk. Correct answer.
  2. It is moved to object storage.
  3. It is automatically deleted.
  4. It is converted into an image automatically.

Explanation

Stopping an ECS releases compute resources such as vCPU and memory, but the system disk remains on its associated EVS disk. The data persists until the disk or instance is deleted according to the chosen deletion settings.

Question 30Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which statement about Huawei Cloud security groups is correct?

  1. They can be associated only with BMS instances.
  2. They are stateless, so return traffic must be allowed explicitly.
  3. They are stateful, so return traffic is automatically allowed. Correct answer.
  4. They apply only to outbound traffic.

Explanation

Security groups are stateful: when an inbound or outbound connection is allowed, return traffic for that connection is automatically permitted. This means separate return-traffic rules are not required for established connections.

Question 31Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

During BMS creation, a user configures local disks with RAID 10. What is the main benefit of this configuration?

  1. It boosts performance and adds disk redundancy. Correct answer.
  2. It encrypts all network traffic to the BMS.
  3. It removes the need for an OS image.
  4. It automatically replicates data to another region.

Explanation

RAID 10 combines mirroring and striping, improving read/write performance while providing disk redundancy within the server. It does not encrypt network traffic, replace an OS image, or provide cross-region replication.

Question 32Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

In Auto Scaling, what does a scaling configuration template define?

  1. The schedule for recurring scale-in events
  2. The specifications and image used for new instances Correct answer.
  3. The IAM users allowed to trigger scaling
  4. The monitoring dashboard layout for scaled instances

Explanation

A scaling configuration template specifies how new instances are created, including flavor, image, disks, network, and related settings. Auto Scaling uses this template whenever a scale-out action launches additional instances.

Question 33Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Traffic to an application peaks predictably every weekday at 09:00. Which Auto Scaling policy type is most suitable?

  1. Lifecycle hook
  2. Alarm-based policy
  3. Scheduled policy Correct answer.
  4. Manual scaling

Explanation

A scheduled policy triggers scaling at specified times and is well suited to predictable traffic patterns. Alarm-based policies react to metrics, while manual scaling requires user action and lifecycle hooks support custom actions during instance lifecycle events.

Question 34Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

An Auto Scaling group has a minimum of 2, a maximum of 10, and a current desired capacity of 4. An alarm-based policy triggers a scale-out action that adds 3 instances. What is the resulting desired capacity?

  1. 7 Correct answer.
  2. 3
  3. 4
  4. 10

Explanation

The scale-out action adds 3 instances to the current desired capacity of 4, producing 7. This value is within the minimum of 2 and maximum of 10, so no boundary adjustment is needed.

Question 35Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

In CCE, what is a node pool?

  1. A set of container images stored in SWR
  2. A set of similarly configured nodes managed together Correct answer.
  3. A network policy applied to the cluster
  4. A logical grouping of pods by namespace

Explanation

A node pool groups cluster nodes that share the same configuration, such as flavor, disk, and labels. Node pools simplify scaling, upgrading, and managing groups of worker nodes with similar requirements.

Question 36Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which Huawei Cloud service runs containers without requiring the user to create or manage cluster nodes and charges based on pod resource usage?

  1. ECS
  2. CCI Correct answer.
  3. CCE
  4. BMS

Explanation

Cloud Container Instance (CCI) is a serverless container service: users run containers without managing cluster nodes and are billed based on the resources consumed by the containers. CCE provides managed Kubernetes clusters with worker nodes that users configure.

Question 37Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which CCE capability automatically changes the number of pod replicas based on metrics such as CPU utilization?

  1. Cluster Autoscaler
  2. Node pool upgrade
  3. Horizontal Pod Autoscaler Correct answer.
  4. Image pull policy

Explanation

The Horizontal Pod Autoscaler adjusts the replica count of a workload based on observed metrics such as CPU utilization. The Cluster Autoscaler adjusts the number of nodes, while node pool upgrade and image pull policy affect node maintenance and image retrieval.

Question 38Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A CCE workload must pull an image from a private SWR repository. What must be configured in the cluster?

  1. An image pull secret with repository credentials Correct answer.
  2. A NAT gateway for each pod
  3. A hostPath volume mounted to /images
  4. A VPC peering connection to SWR

Explanation

Private repositories require credentials, which Kubernetes supplies through an image pull secret referenced by the pod specification. VPC peering, NAT gateways, and hostPath volumes do not provide repository authentication.

Question 39Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

What is the primary function of Image Management Service (IMS) in Huawei Cloud?

  1. To create and manage private images for ECS and BMS Correct answer.
  2. To provide DNS resolution for ECS instances
  3. To back up EVS disks on a schedule
  4. To monitor ECS CPU and memory metrics

Explanation

IMS creates, stores, and manages images that can be used to launch ECS and BMS instances. Monitoring, DNS, and disk backup are handled by separate services.

Question 40Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

An AI team needs ECS instances with GPUs for model training. Which ECS flavor family should they choose?

  1. GPU-accelerated Correct answer.
  2. General computing
  3. Disk-intensive
  4. Memory-optimized

Explanation

GPU-accelerated ECS flavors provide GPUs for parallel processing tasks such as model training and inference. General computing, memory-optimized, and disk-intensive flavors do not provide GPU acceleration.

Question 41Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A BMS is used for a high-performance computing workload that requires very low network latency and high packet throughput between nodes. Which BMS networking feature is most relevant?

  1. SR-IOV enhanced networking Correct answer.
  2. VPC route table entries
  3. Elastic IP address binding
  4. Security group stateful filtering

Explanation

SR-IOV enhanced networking gives virtual functions direct access to the physical NIC, reducing virtualization overhead and improving throughput and latency. EIP binding provides public access, security groups filter traffic, and route tables direct traffic but do not provide the same data-path acceleration.

Question 42Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

An Auto Scaling group uses a warm pool that currently contains two stopped pre-initialized instances. A scale-out event requires three additional instances. What is the expected behavior?

  1. It fails the scale-out because stopped instances cannot be used.
  2. It starts only one warm-pool instance and creates two new instances.
  3. It starts the two warm-pool instances and creates one new instance. Correct answer.
  4. It creates three new instances and ignores the warm pool.

Explanation

A warm pool holds pre-initialized instances that can be used to speed up scale-out. If two stopped instances are available and three are needed, Auto Scaling starts the two warm-pool instances and creates one additional instance.

Question 43Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A CCE workload must preserve stable pod names and persistent storage across rescheduling. Which workload type is appropriate?

  1. Job
  2. Deployment
  3. StatefulSet Correct answer.
  4. DaemonSet

Explanation

StatefulSets provide stable pod identities and ordered deployment and scaling, and they are commonly used with persistent storage for stateful applications. Deployments and DaemonSets are better suited to stateless or node-wide workloads, while Jobs run to completion.

Question 44Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A CCE cluster uses HPA to add pods, but the existing nodes have insufficient CPU. Which component should add more nodes?

  1. Cluster Autoscaler Correct answer.
  2. SWR image replication
  3. HPA itself
  4. The CCE dashboard

Explanation

HPA scales pod replicas, but when pods cannot be scheduled because nodes lack resources, the Cluster Autoscaler adds nodes to the node pool. Image replication and the dashboard do not provide compute capacity.

Question 45Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A database running on a BMS stores its data on local disks. The BMS must be rebuilt after a hardware failure. Which design change best ensures the data survives the rebuild?

  1. Assign a static public IP to the BMS.
  2. Enable a larger local RAID array on the BMS.
  3. Store database data on attached EVS disks. Correct answer.
  4. Move the BMS into a different VPC.

Explanation

EVS disks are independent block storage resources that can be detached and reattached, so data on EVS can survive a BMS rebuild. Local RAID improves local disk performance and redundancy but does not protect data if the physical server must be replaced.

Question 46Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A user needs an Elastic Cloud Server to be reachable directly from the internet through a public IP address. Which resource must be associated with the ECS for this purpose?

  1. A VPC peering connection
  2. A security group rule
  3. A VPN gateway
  4. An elastic IP address (EIP) Correct answer.

Explanation

An EIP is a public IP address that can be bound to an ECS (or other resource) so traffic from the internet can reach it directly. Peering and VPN connections build private paths between networks, and security group rules only filter traffic rather than supply a public address.

Question 47Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

In Cloud Container Engine (CCE), which Kubernetes object provides a stable virtual IP address and load balancing in front of a changing set of pods?

  1. PersistentVolumeClaim
  2. Namespace
  3. Service Correct answer.
  4. ConfigMap

Explanation

A Service defines a stable virtual IP and distributes traffic to the pods selected by its label selector, so clients do not need to track individual pod IPs. The other objects deal with configuration data, logical separation, and storage requests rather than traffic distribution.

Question 48Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

An Auto Scaling group is configured so that instances are checked by the load balancer health check. One member ECS repeatedly fails that check. How does Auto Scaling respond?

  1. It reboots the instance's operating system in place without any replacement.
  2. It moves the instance into a warm pool and keeps it there indefinitely.
  3. It suspends all further scale-out actions until the cooldown period ends.
  4. It removes the instance and launches a replacement to maintain the desired capacity. Correct answer.

Explanation

When load balancer health checking is enabled for a scaling group, Auto Scaling treats a failing member as unhealthy, removes it from the group, and creates a new instance so the desired capacity is restored. The feature is about replacing faulty members, not about rebooting them or changing the group's scaling limits.

Question 49Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

Which statement correctly describes the compute resources provided by a Bare Metal Server (BMS)?

  1. The server can be used only as a container node and cannot run general workloads.
  2. CPU and memory are shared with other tenants while local disks are dedicated.
  3. The physical server's CPU, memory, and local disks are dedicated to that single server. Correct answer.
  4. Resources are virtualized and overcommitted in the same way as a general-purpose ECS.

Explanation

A BMS gives a tenant an entire physical server, so processors, memory, and local disks are not shared with other tenants and there is no hypervisor layer competing for resources. This is why BMS suits workloads sensitive to virtualization overhead or needing strong isolation.

Question 50Computing Services Design (Elastic Cloud Server, Bare Metal Server, Auto Scaling, Containers)

A scaling group has a lifecycle hook configured for the terminate action. What does the hook allow during the removal of an instance?

  1. It converts the instance into a standalone pay-per-use ECS outside the group.
  2. It raises the maximum instance count of the group and then continues the scale-in.
  3. It permanently blocks any future scale-in event for the entire scaling group.
  4. It keeps the instance in a waiting state so custom actions such as log collection can run before removal. Correct answer.

Explanation

A lifecycle hook suspends an instance in a waiting state when the associated action is triggered, giving an external system time to perform custom operations such as draining sessions or collecting logs before the instance is removed. If the hook is not completed within its timeout, the action proceeds by default.