块存储_分布式软件定义存储_XEBS块存储

Get What You Need with XEBS

Traditional distributed storage solves the problem of capacity limits and scale-out capability of dedicated storage devices under large capacity and cold data scenarios. However, it cannot smoothly replace the traditional SAN storage in terms of the adaptability breadth to application scenarios, the complexity of operation and maintenance management and TCO. Users need innovativeness as well as enterprise-level block storage features to support VMware, Citrix, Oracle, Microsoft, K8s，etc.

Product Value

XEBS（XSKY Enterprise Block Storage）是 XSKY 星辰天合提供的专业的分布式块存储产品。在“去中心化”的分布式架构之上支持多活路径的 FC、iSCSI 等传统 SAN 接口，与 AIX、Linux、Windows 下的应用或 VMware、Citrix 等企业级虚拟化技术充分认证对接。

Advantages

Professional APIs and Enhanced Reliability

Support a full range of IP-based iSCSI command words and the most widely used FC by professional storage devices
Redundant design of key IO paths and management control modules, supporting multi-path MPIO or VIP, active/active access, ensuring high system availability and fast switch
Adapt to various complex scenarios and processes of enterprise storage management with enterprise-level advanced features such as ROW instant snapshot, cloning, replication, automatic thin provisioning, online volume expansion and shrinkage, stretched cluster, cloud backup, and multiple QoS control
Volume-level QoS settings allow users to set IOPS threshold limits on volumes and performance limits in case of emergency to ensure consistent experience in storage performance. And real-time changes to QoS of volumes better help administrators and application developers dynamically allocate resources for performance to deal with tidal scenarios

基于 XSKY 星辰天合最佳实践的性能优化

Combined with mainstream NVMe flash memory technology and optimized IO path, tens of thousands of IOPS can be measured at a single node and a high-performance multi-level Cache acceleration engine can be provided, including two-level RM and SSD Cache, among which RAM Read Cache adopts a patented enhanced read-ahead algorithm to improve system read performance
SSD R/W Cache offers smart IO algorithm to greatly improve system R/W performance for OLTP
Innovative IO aggregation enables small blocks to be randomly aggregated into large sequential IO blocks, provides stable performance under high pressure and high concurrency, and perfectly supports EC
Best interaction practices under enterprise-level virtualization scenarios: Seamless connection to OpenStack ecosystem, VMware VAAI Ready, VMware Ready and Citrix Ready Certification
基于容器场景的 XSKY 星辰天合最佳实践，提供官方认证的 CSI-iSCSI 驱动

Enhanced Easy-of-use and Manageability

Rich visual management functions and comprehensive management processes help users achieve rapid resource creation, installation, deployment and expansion, and enable users to easily control the complexity of distributed storage hardware
The tiered design of the fault domain and storage pool allows users to easily cope with resource planning and critical application partitions
Treatment of hard disk and network sub-health, and automatic identification of bad disk, slow disk and network failure with automatic isolation
Rich RESTful APIs, command line, SSD life prediction, S.M.A.R.T., operation log, system log and multi-level alert function help users quickly locate and solve emergencies in cluster

Typical Scenarios

Multi-virtualization

Meet the unified storage demand of enterprise-level virtualization platforms including VMware ESXi, Microsoft Hyper-V and Citrix to simplify architecture and eliminate data islands.

Multi-database

Support the performance and capacity requirement of OLTP and OLAP database, reducing uneven distribution and building a flexible and in-control data resource pool.

Mainstream Container

Support of latest container ecosystem with certified CSI-iSCSI driver.

Hybrid Cloud

Connect private and public cloud to construct a hybrid cloud and support data backup from private cloud to public cloud on demand.

THE Distributed Block Storage for Enterprise Applications