Administrator Guide

2
Architecture Fundamentals
The Dell EqualLogic product family provides a unied le and block storage platform.
Block-level storage consists of a sequence of bytes and bits of a certain length, called a block. Each block stores the data (like a hard
drive) and the disk controller reads and writes data to the disks inside the storage array. Block-level access enables storage
administrators to stipulate which block to send reads and writes to for the best performance. In le-level storage, the operating
system keeps track of data in a directory of le and folder names. Access to data by an application is by le name and location within
the le, and then translated into block-level access for physical reading and writing of data. Block-level transfers in a SAN are
typically faster than le-level transfers in a NAS because they do not have any le system or network overhead.
The PS Series SAN provides block-level access to storage data through direct iSCSI access to PS Series arrays. The FS Series NAS
appliance provides le-level access to storage data using NFS (UNIX) or SMB (Windows) protocols and the Dell FluidFS le system.
This architecture uses a scale-out design, which enables you to increase performance as you add capacity, because the software
manages the workload of data writes and the subsequent data reads across all the resources in the storage infrastructure.
PS Series Architecture
The PS Series architecture comprises three fundamental components: the PS Series array, the PS Series group, and the PS Series
storage pool. These components coordinate operations and work together to form a virtual storage system.
The foundation of the PS Series architecture is a PS Series array. Each array is composed of redundant components: disks,
controllers with mirrored write-back caches, network interfaces, power supplies, and cooling fans. When you congure a PS Series
array, you add it to a PS Series group. A group can consist of up to 16 arrays of any family or model, as long as all arrays in the group
are running rmware with the same major and minor release number. The group appears to the client servers as a single entity that
oers network storage access in block mode.
NOTE: Dell recommends that all arrays run the same version of PS Series rmware at all times, except during the
rmware update process.
Each array in a group is called a group member. You add members to the group, and allocate storage space in the group by creating
storage pools. A pool can have from one to eight members, and you can create up to four pools in a group. Each member contributes
its resources to the pool (disk space, processing power, and network bandwidth). As you add or remove members, the pool of
storage space grows or shrinks. Hardware and other details remain hidden.
By default, a group member’s RAID-protected disk space is added to the default storage pool. When you add a new member, it is
automatically assigned to the default pool. The default pool cannot be deleted; however, you can create additional pools as required
for your environment. You can assign members to the pools at any time.
The PS Series array software load balances data and network I/O to the group across all the group members’ resources. When a
group contains more than one member, those members might have dierent capacity levels and use dierent RAID levels. The load
balancers transparently maintain optimal system performance and eliminate downtime to servers, applications, and users.
Next, you create volumes and assign them to the appropriate pool. (You can also organize volumes into folders for easy reference.) A
volume provides the structure for the group. You can create a volume on a single group member or one that spans multiple group
members. You identify a volume by specifying a unique volume name. Hosts on the network see these volumes as iSCSI targets. You
set access controls for each volume so that only computers with an iSCSI initiator and the correct access credentials can access
volume data.
Architecture Fundamentals
25