Distributed fusion architecture
The H3C UniStor X10000 adopts a fully distributed converged architecture to ensure high performance, high scalability, security, reliability, and automated operation and maintenance of the storage system.
It also supports block, file, object and big data storage
Supports big data storage, Hadoop, Hbase, Hive, Spark, Impala, and other components, and supports append semantics
Distributed symmetric architecture without the need for a separate metadata server
Distributed global cache improves cluster read and write performance
Supports Intel chip platform and Haiguang chip platform
In the open VM architecture, Vendor-developed or third-party streaming media software can be deployed on nodes.
Strong linear expansion capability
It can be scaled up to 8192 nodes to support flexible business development
Performance and capacity increase linearly with the number of nodes and cluster size
High reliability
Multiple copies ensure data reliability and support flexible configuration of 2-8 copies. For example, three copies only when three copies of data are lost at the same time, user data is really lost
Different erasure code levels from N+1 to N+4 can be set based on resource pools and directories
Comprehensive data protection and disaster recovery capabilities, including replication, snapshot, clone, copy, fault domain, and protected domain technologies
Data persistence ≧99.99999999999% (13 9s)
High availability
The fully redundant network ensures no single point of failure. When disks or nodes are damaged, services are not interrupted and applications are not aware
When a disk or node fails, the system automatically reconstructs. You can adjust the reconstruction priority
Supports online uninterrupted capacity expansion. Adding or deleting nodes or disks does not affect service running. Supports automatic cluster data balancing
Data availability ≧99.99995%
Easy to maintain
Rapid cluster deployment, supporting batch deployment, configuration boot, and one-click upgrade
Relying on the monitoring and alarm system, quickly define problems and troubleshoot faults, allowing users to customize data distribution strategies and support fault domain isolation
manageability
Chinese visual operation management platform, supports both IPv4 and IPv6 access
You can intelligently monitor cluster performance and health indicators and set alarms to learn about system status
Supports log and user statistics, file system and resource statistics, and system capacity and performance prediction and analysis for hardware read and write
