HP StorageWorks 1000i Virtual Library System - Manual (Page 16)

Important concepts
To understand the configuration of the backup network and how it fits into the local-area network (LAN),
review the following sections.
Internet SCSI (iSCSI) protocol
Internet SCSI (iSCSI) is a standard protocol for universal access to shared storage devices over standard,
Ethernet-based transmission control protocol/Internet protocol (TCP/IP) networks. The connection-oriented
protocol transports SCSI commands, data, and status across an IP network.
The iSCSI architecture is based on a client-server model. The client is a host system that issues requests
to read or write data. iSCSI refers to a client as an initiator. The server is a resource that receives and
executes client requests. iSCSI refers to a server as a target.
File servers, which store the programs and data files shared by users, normally play the role of server.
With the VLS, the application and backup servers within your network act as clients or initiators and
the VLS acts as a server or target. The initiators can either be iSCSI software simulation or host bus
adapters (HBAs) on the server that is being backed up.
Disk-to-Disk-to-Tape (D2D2T) backup capabilities
The VLS is a storage resource used by a single backup server or shared by multiple backup servers using
an Ethernet network. By using standard backup software, you can copy backup data that resides on the
VLS to physical tape for long-term data retention.
The following illustration shows application servers sending backup data over a Gigabit Ethernet (GbE)
LAN to backup servers sharing VLS D2D storage over GbE.
NOTE:
The connection from the Client — Tape can be either FC or direct attached SCSI.
Clients
Ethernet
VLS 1000i Client Client - Tape
Figure 1 Disk-to-Disk-to-Tape backup
In addition to being part of the LAN, the backup servers, and the VLS are part of the GbE backup LAN.
Redundant Array of Independent Disks (RAID)
RAID provides convenient, low-cost, reliable storage by saving data on more than one disk drive
simultaneously. If one disk drive in a RAID 5 configuration becomes unavailable, the others continue to
work in a degraded state, thus avoiding downtime for users.
16 Introduction