No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search


To have a better experience, please upgrade your IE browser.


FusionStorage 8.0.0 Product Description 02

Rate and give feedback:
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).
Software Architecture

Software Architecture

Functional Architecture

Figure 3-2 shows the functional architecture of FusionStorage.

Figure 3-2 Functional architecture

Table 3-1 describes the functions of software modules.

Table 3-1 Functions of software modules



Storage interface layer


Provides volumes for operating systems and databases by locally mapping the volumes using a SCSI standard driver and interacting with the storage system through a private communication protocol.


Works with multipathing software to support iSCSI interfaces and provides volumes for operating systems and databases.

Storage service layer


Provides the snapshot function. HyperSnap does not physically copy data. Instead, it quickly provides complete data copies by locating data through mapping tables.


Provides the clone function. Created based on volume snapshots, linked clone volumes have the same data as the snapshots and are used for data backup or other purposes.


Provides the remote replication function. HyperReplication creates usable data copies of a local storage system in near real time on a remote storage system. The copies are instantly available without data restoration, maximizing service continuity and data availability. After HyperReplication pairs are added to a consistency group, any operation to the consistency group also takes effect on the HyperReplication pairs in the group. This ensures that data in multiple HyperReplication pairs is consistent in time.


Provides the active-active storage function. HyperMetro enables real-time data synchronization and access between two storage systems. If data access fails in any storage system, HyperMetro implements seamless service switchover, ensuring data security and service continuity.


Implements intelligent service quality control. SmartQoS controls the performance of volumes to ensure service quality of high-priority applications.


Uses high-speed and small-capacity disks as the buffer for low-speed and large-capacity disks. SmartCache caches data and improves the read and write performance of the storage system.

SmartDedupe and SmartCompression

Provides deduplication and compression functions. SmartDedupe analyzes and deletes duplicate data in the storage system, and SmartCompression compresses written data blocks by changing data storage formats, greatly reducing storage space occupation.

Storage engine layer

Cluster status control

Controls the status of the distributed cluster.

Data redundancy management

Calculates EC and copies to ensure high data reliability.

Distributed data routing

Evenly distributes data and metadata to storage nodes according to preset rules.

Data rebuilding and balancing

Rebuilds and balances data.

Strong-consistency replication protocol

Ensures data consistency for HyperReplication pairs and HyperMetro pairs.

Storage management

Operates, manages, and maintains the system, and provides functions such as system installation, deployment, service configuration, device management, monitoring, alarm reporting, upgrade, and capacity expansion.


For details about key service processes of FusionStorage, see FusionStorage Technical White Paper of the desired version.

Component Architecture

Figure 3-3 and Figure 3-4 show the component architecture, depending on whether Virtual Block Service (VBS) is deployed on storage or compute nodes. Table 3-2 describes key components and processes.


VBS: a process that provides the distributed storage access point service through SCSI or iSCSI interfaces and enables application servers to access distributed storage resources

Management node: a server that runs the FusionStorage Manager (FSM) process and provides operation and maintenance (O&M) functions, including alarm reporting, monitoring, logging, and configuration. The FSM process can be deployed in a virtual environment, an external physical server, or a storage node. Deploying the FSM process in active/standby mode is recommended. Figure 3-3 and Figure 3-4 use deploying the FSM process on external physical servers as an example.

Figure 3-3 Component architecture, with VBS deployed on storage nodes
Figure 3-4 Component architecture, with VBS deployed on compute nodes
Table 3-2 Key processes and components of FusionStorage

Process or Component



Management process of FusionStorage. You are advised to deploy FSM on two management nodes in active/standby mode.

FSM automatically starts during management node installation.

FusionStorage Agent (FSA)

Management agent process of FusionStorage deployed on storage nodes and compute nodes to implement communication between the nodes and FSM.

FSA automatically starts during storage or compute node installation.

Metadata Controller (MDC)

Metadata control component that controls the status of the distributed cluster as well as data distribution and rebuilding rules.

When you create a control cluster using three, five, or seven nodes, MDC automatically starts on these nodes. The ZooKeeper (ZK) cluster will elect a primary MDC to monitor other MDCs. When the primary MDC fails, the ZK cluster elects a new one. Each resource pool has one MDC. When a resource pool is added, its MDC automatically starts. A maximum of 96 MDCs can be started in one system.


Distributed coordination service process for applications. ZK is deployed on each node in the control cluster to form a ZK cluster. The ZK cluster provides primary arbitration for the MDC cluster. More than half of the ZKs must be active and accessible.

ZK automatically starts on the nodes that form the control cluster during control cluster creation.


Process that provides the distributed storage access point service through SCSI or iSCSI interfaces and enables application servers to access distributed storage resources.

VBS needs to be manually started on servers, and the servers running VBS form a VBS cluster.

Enterprise Data Service (EDS)

Process that provides enterprise-class features, including HyperSnap, HyperReplication, and SmartDedupe.

EDS automatically starts on the servers that have been added to storage pools.

Object Storage Device (OSD)

Process that handles I/O messages from VBS, implements data redundancy protection, and stores data onto storage media.

OSD automatically starts on the servers that have been added to storage pools. Each server runs only one OSD.

Cluster Manager (CM)

Cluster management process that manages and controls cluster information.

CM automatically starts during control cluster creation.

Cluster Configuration Database (CCDB)

Cluster configuration database used to store cluster configurations, such as configurations of HyperMetro and HyperReplication pairs and consistency groups.

CCDB automatically starts during control cluster creation.

Updated: 2019-07-12

Document ID: EDOC1100081421

Views: 1617

Downloads: 31

Average rating:
This Document Applies to these Products
Related Documents
Related Version
Previous Next