Huawei HCIA-Storage flashcards

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

RAID 5 usable capacity

Usable capacity equals (number of drives minus one) multiplied by the smallest drive capacity. It requires at least three drives and survives one drive failure.

RAID 6 fault tolerance

RAID 6 uses two independent parity blocks per stripe, tolerating two simultaneous drive failures. It needs at least four drives and usable capacity is (N-2) times the smallest drive.

RAID 10 capacity

RAID 10 mirrors and stripes data, so usable capacity is 50 percent of raw capacity. It tolerates one failure per mirrored pair and requires at least four drives.

Write-back cache

The controller acknowledges a write after storing it in cache but before writing to disk. A battery or supercapacitor protects the cache during power loss.

Cache mirroring

Each controller copies its write cache to the peer controller. If one controller fails, the peer retains the data and can take over without losing acknowledged writes.

LUN masking

A mapping that restricts a LUN to specific host initiators identified by WWN or IQN. It prevents unauthorized hosts from accessing and corrupting the LUN.

Thin provisioning over-subscription

The ratio of total provisioned capacity to physical capacity. A ratio above one is allowed because not all capacity is used immediately, but monitoring prevents out-of-space events.

ALUA

Asymmetric Logical Unit Access defines target port states as active/optimized or active/non-optimized. Hosts send I/O through optimized paths and use non-optimized paths only on failover.

FusionCompute VRM

The Virtual Resource Manager is the management component of FusionCompute. It manages clusters, hosts, and virtual machines and provides the web-based management interface.

FusionCompute CNA

The Computing Node Agent runs on each compute node and executes virtualization operations. It works with the UVP hypervisor to run virtual machines on the host.

FusionStorage OSD

An Object Storage Device is a FusionStorage process that manages a disk and stores data fragments. Multiple OSDs across nodes form the distributed storage pool.

FusionStorage VBS

The Virtual Block Store handles I/O requests on compute nodes. It uses a distributed hash table to locate data and can access any OSD in the cluster.

Huawei EVS

Elastic Volume Service provides block storage volumes that attach to ECS instances in Huawei Cloud. Volumes can be created from snapshots and resized online.

Huawei OBS

Object Storage Service stores unstructured data as objects in buckets. It is accessed through RESTful APIs and is suitable for backup, archiving, and static content.

Security group vs ACL

A security group is stateful and filters traffic at the instance level. A network ACL is stateless and filters traffic at the subnet level; both control inbound and outbound flows.

Critical path

The longest-duration sequence of dependent activities in a project network. It determines the shortest possible project duration and has zero total float.

Cost variance formula

CV equals earned value minus actual cost. A positive CV means the project is under budget; a negative CV means it is over budget at the measurement date.

Schedule performance index

SPI equals earned value divided by planned value. An SPI below one indicates the project is behind schedule; above one indicates it is ahead of schedule.

Estimate at completion

EAC forecasts total project cost. When current cost variance is typical, EAC equals budget at completion divided by the cost performance index.

Risk transfer

A risk response that shifts the impact and ownership of a threat to a third party, such as through insurance, warranties, or fixed-price contracts, without eliminating the risk.

Change control board

A group that reviews change requests, evaluates impact on scope, schedule, and cost, and approves or rejects changes. Approved changes update the project baselines.

Work breakdown structure

A hierarchical decomposition of project deliverables into smaller, manageable work packages. The lowest level supports estimating, scheduling, and assignment of responsibility.

Erasure coding overhead

With k data fragments and m parity fragments, storage overhead is (k+m) divided by k. Any k fragments can reconstruct the original data after up to m losses.

FC SAN vs iSCSI

Fibre Channel is a lossless network with in-order delivery and buffer credits. iSCSI carries SCSI over TCP/IP and relies on TCP retransmission, adding latency under congestion.

SSD over-provisioning

Reserved flash capacity beyond the logical address space, typically 7 to 28 percent. It supports garbage collection and wear leveling, improving endurance and steady-state write performance.

Storage tiering

Placement of data across performance tiers such as SSD and HDD. Hot extents move to fast media and cold extents to capacity media, usually at sub-LUN granularity.

Snapshot copy-on-write

A snapshot technique that stores only blocks changed after the snapshot is taken. It conserves capacity but depends on the source volume remaining intact.

Dual-controller active-active

Both storage controllers process I/O simultaneously and mirror cache to each other. If one fails, the peer takes over its LUNs with no service interruption.

Multipathing

Software that presents multiple physical paths between a host and a LUN as a single logical device. It provides failover and can balance I/O across paths.

Huawei FusionSphere

A cloud operating system comprising FusionCompute for virtualization, FusionManager for management, and FusionStorage for distributed storage. It supports private and public cloud deployments.

Huawei VBS service

Volume Backup Service creates policy-based backups of cloud disks in Huawei Cloud. Backups are stored independently and can be used to restore or create new volumes.

Earned value management

A technique integrating scope, schedule, and cost. It compares earned value with planned value and actual cost to compute variance and performance indices.