Huawei HCIP-Cloud Service Solutions Architect flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Region

A physically isolated geographic area containing multiple availability zones; resources are not replicated between regions automatically, so cross-region disaster recovery must be designed explicitly.

Availability Zone

One or more isolated data centers inside a region with independent power, cooling and network; zones are linked by low-latency fiber, enabling cross-AZ redundancy for high availability.

Shared responsibility model

The provider secures physical facilities, the hypervisor and managed-service platforms; the customer secures the guest OS, applications, data, identities and configuration above the service boundary.

Pay-per-use vs subscription

Pay-per-use bills resources by actual usage with no commitment and suits spiky workloads; yearly or monthly subscription gives a discount for a fixed term and suits steady, predictable workloads.

Elastic Cloud Server (ECS)

A virtual machine defined by flavor (vCPU and memory), image, disks, VPC subnet and security group, with an optional EIP; it can be resized vertically and scaled horizontally.

Auto Scaling

A service that adds or removes instances automatically using a scaling configuration template, a scaling group with minimum, maximum and desired counts, and alarm, scheduled or periodic policies.

Bare Metal Server

A physical server without a hypervisor that provides full hardware performance and isolation; it suits HPC, high-performance databases and workloads with strict licensing or latency requirements.

Cloud Container Engine (CCE)

A managed Kubernetes service in which Huawei Cloud operates the control plane while worker nodes, built from ECS or BMS, are organized into node pools for scheduling.

OBS storage classes

Standard, Infrequent Access, Archive and Deep Archive; moving objects to colder classes reduces storage cost but increases retrieval time and imposes minimum storage durations.

EVS disk tiers

Extreme SSD delivers the highest IOPS and lowest latency, general purpose SSD balances price and performance, and SAS or SATA disks serve low-IO, cost-sensitive workloads.

Scalable File Service (SFS)

A shared file service providing NFS or CIFS file systems that many ECS instances can mount concurrently; the SFS Turbo variant adds higher throughput and lower latency.

Cloud Backup and Recovery

A backup service that applies backup policies to EVS disks and ECSs, storing recovery points in vaults; it supports restoration and cross-region replication of backups.

Security group

A stateful, instance-level virtual firewall that contains only permit rules; inbound traffic is denied by default, outbound is permitted, and response traffic is automatically allowed.

VPC peering connection

A connection between two VPCs that lets instances communicate using private IP addresses without a NAT gateway, VPN or public IP; it is created within a single region.

Direct Connect

A dedicated private physical link from an on-premises data center to Huawei Cloud that provides predictable bandwidth and latency and never traverses the public internet.

Elastic Load Balance

A load balancing service distributing traffic across backend server groups with health checks; dedicated load balancers handle layer 4 TCP/UDP and layer 7 HTTP/HTTPS traffic.

IAM agency

A delegation mechanism that lets a cloud service or another account obtain temporary credentials to act on resources, avoiding the sharing of permanent access keys.

KMS envelope encryption

A scheme in which a customer master key encrypts data keys; data is encrypted with a data key, and the master key never leaves the key management service.

Web Application Firewall

A layer 7 firewall that inspects HTTP and HTTPS requests for SQL injection, cross-site scripting and CC floods before traffic reaches the application behind an ELB or EIP.

Host Security Service

A host security service for ECS and BMS that provides intrusion detection, antivirus, vulnerability and baseline checks, and critical file integrity monitoring.

RDS read replica

A read-only copy of a primary database instance that offloads read queries; replication is asynchronous, so the replica may lag and is not a zero-data-loss failover target.

Distributed Cache Service

A managed Redis or Memcached service providing in-memory, sub-millisecond access, commonly used for caching, session state, counters and hot-spot data in front of a database.

Data Warehouse Service (DWS)

GaussDB(DWS) is a massively parallel, column-oriented data warehouse for large-scale OLAP analytics, distributing and executing queries across many nodes for high throughput.

Distributed Database Middleware

A middleware that shards MySQL data horizontally across multiple RDS instances, presenting a single logical database and scaling write capacity beyond one instance.

Recovery Time Objective

Recovery Time Objective is the maximum acceptable service downtime after a failure, which drives the choice between backup-restore, warm standby and active-active designs.

Recovery Point Objective

Recovery Point Objective is the maximum tolerable data loss expressed in time, which determines backup frequency and whether synchronous or asynchronous replication is needed.

Storage Disaster Recovery Service

A disaster recovery service that replicates ECS instances and EVS disks across availability zones or regions in protection groups and supports one-click failover and failback.

Server Migration Service

A service that moves physical, virtual or other-cloud servers to Huawei Cloud ECS by synchronizing the source operating system and data incrementally before cutover.

Cloud Eye

A monitoring service that collects metrics and events, renders dashboards and raises alarms which can notify SMN or trigger Auto Scaling actions.

Log Tank Service

A log service that collects, stores and searches logs from hosts, containers and cloud services, supporting troubleshooting, analysis and audit retention.

Cost allocation tags

Key-value labels attached to resources so billing reports can group and allocate costs by project, department, environment or application owner.

Right-sizing

Selecting instance flavors and storage capacities that match measured utilization; over-provisioned compute and idle disks are the most common sources of avoidable cloud spend.