Introduction to Data Storage Virtualization Technology (Part 2)

The article "收藏：数据存储虚拟化技术介绍(上篇)" continues the discussion on storage virtualization, describing how heterogeneous storage devices are abstracted into a single transparent resource pool for users.

Storage virtualization maps various physical storage devices to a unified storage pool, providing interoperability, protecting existing investments, and lowering total cost of ownership (TCO). Techniques such as striping, LUN masking, and zoning enable flexible partitioning and allocation of the pool.

Virtualization also allows dynamic, transparent scaling of storage pools to servers, and can extend to SAN‑to‑SAN or global virtualization scenarios.

The evolution of storage technology is traced from single disks and tapes to RAID, DAS, and finally network storage. Traditional Direct‑Attached Storage (DAS) ties storage to a specific server, which is insufficient for distributed enterprise workloads, prompting the development of network storage.

Two main network storage models are introduced:

NAS (Network Attached Storage) : storage devices attached directly to the network, offering good sharing and openness but sharing the LAN can cause congestion and reduced performance, especially during backups.

SAN (Storage Area Network) : dedicated storage networks, typically based on Fibre Channel, providing higher performance, easier system expansion, data migration, backup, disaster recovery, and unified management as a single storage pool.

Although NAS and SAN represent significant advances, they still require administrators to understand the physical characteristics of each device for effective management.

The article concludes by summarizing various forms of storage virtualization (block, disk, file‑system, tape‑library) and provides references to related technical articles.

Additional promotional content mentions a collection of architecture‑related e‑books and a limited‑time discount for a bundled “architect technical store” package.