Many organizations know the benefits of centralized file storage on shared drives, and more latterly the convenience of cloud storage like Dropbox and Google Drive. But did you know that the benefits of consolidating and managing information centrally also extend to servers? And to the websites and applications they serve? They do. And they are critical in virtualized environments where computing and storage are separated.
Centralized storage is the storage of files, data and databases shared between computing servers over a network. Indeed, it is also known as networked storage. This is different from storing data locally in the same chassis or directly attached in a neighboring server (known as Direct-Attached Storage).
Centralized data becomes useful when two or more servers need access to the same data. This is commonly the case in high-availability systems, where redundancy and replication play a role, and in virtualized environments, where virtual servers are created frequently or automatically as systems scale (and need to access existing data).
The benefits of centralized storage are several-fold. By keeping data in one place, it’s easier to manage both the hardware and the data itself. That means closer control on data protection, version control and security. It means a single, consistent set of data. It means better control of hardware configuration, capacity and performance. And, by focusing your efforts in one place, it means reduced expenditure and risk.
Even though the additional features of some centralized storage solutions – notably redundancy, disk performance and reliability – come at a premium, the centralization of the investment and the improved reliability often means that the lifetime cost is lower.
And by hosting servers and storage in a third party data center, capital requirements, risk and the overhead of managing hardware and networks are further mitigated.
Popular Storage Configurations
Network-Attached Storage (NAS) - A NAS architecture allows you to connect multiple app servers to centralized storage servers with replication and failover. Each app server accesses the same data.
Storage Area Network (SAN) - A Storage Area Network (SAN) provides a redundant array of disks seen as a local volume by the servers attached to it, for high availability and faster disk I/O performance than NAS. SAN’s are common virtualization clusters and for providing centralized access to I/O intensive databases.
High-Availability Cluster – An HA Cluster contains two or more of each service, not just file storage and database servers but also application or web servers. Traffic is distributed by a load balancer that also ensures failover in the event of hardware or application failure.
