This preview shows page 1. Sign up to view the full content.
Unformatted text preview: te pools of storage by consolidating storage
nodes on the network into clusters. Storage Clustering provides online scalability, both within a
volume and across the entire storage pool. All available physical capacity is aggregated and
available to the volumes created on the SAN.
In order to scale capacity and/or performance, the IT administrator simply adds nodes to the
storage cluster. SAN/iQ software automatically redistributes the data for optimal data availability
and performance. All the capacity, processing power, and bandwidth included in each node are
aggregated into the entire SAN, ensuring an increase in performance as the SAN grows. To make
the process even easier, HP P4000 SANs let IT administrators expand volumes and add storage
nodes online, without taking the volumes offline or causing application downtime.
Customers can use SAN/iQ Storage Clustering to implement different tiers of storage in their SAN.
For instance a storage cluster of SAS-based storage nodes can be implemented for performance
while a storage cluster of MDL SAS-based storage nodes is implemented for higher density, all
managed from a single interface.
For increased access to the SAN, customers can install storage nodes anywhere on the IP network.
P4000 nodes in a cluster can be spread out between several racks on a floor, across several
floors in a building, or across buildings on a campus site or a metropolitan area. Typical round
trip latency should be 2 milliseconds or less. A single P4000 SAN can thus eliminate the risk of
data loss in the event of a disk, controller, storage node, power, network, or site failure.
SAN/iQ Network RAID stripes and protects multiple copies of data across a cluster of storage
nodes, eliminating any single point of failure in the HP P4000 G2 SAN. Applications have
continuous data availability in the event of a power, network, disk, controller, or entire storage
SAN administrators can manage redundancy on a per-volume basis to optimize storage utilization
and match the data protection of the volume to the application data on that volume. Customers
choose from Network RAID level 0, 5, 6, 10, 10+1, or 10+2 to protect data across the storage
nodes, only allocating additional storage space for data that warrants additional protection. For
increased protection, Network RAID can also be integrated into environments where application
servers are clustered, enabling true, seamless, geo-cluster solutions that provide both application
and storage clustering across geographies.
Most SAN vendors place the provisioning burden on SAN administrators, asking them to predict
how much space will be needed in the future for volumes, snapshots and remote copies, and what
the SAN's expected growth rate will be. That information is required because most storage
provisioning models call for pre-allocation of storage space on the SAN. Errors in estimation are
inevitable, and can be expensive or cause snapshots and backups to fail. Worse yet, if
View Full Document
- Fall '11