iSCSI accelerates the transition to network storage

An examination of the potential benefits of iSCSI shows why it may be a worthy alternative (or complement) to DAS, NAS, and Fibre Channel SANs.

By David Dale

Exploding data growth is placing tremendous pressure on enterprises to store, protect, distribute, and derive value from all that data. IT executives are increasingly looking for new storage solutions to manage their growing storage needs. iSCSI is a storage networking standard that promises to deliver the benefits of networked storage to a much broader range of companies.

iSCSI is a standard protocol for encapsulating SCSI commands into TCP/IP packets, enabling block data transport over IP networks.

The Internet Engineering Task Force (IETF) ratified the iSCSI standard a year ago. A variety of iSCSI products, including software drivers, disk arrays, routers, bridges, and host cards that offload both iSCSI and TCP/IP protocol processing from host servers, are shipping today. iSCSI can be used to build SANs using standard Ethernet infrastructure and components.

On the server side, you'll need an "iSCSI initiator." The initiator can be implemented in software (Microsoft and other vendors offer iSCSI software drivers for a variety of operating systems). In this case, a standard Ethernet network interface card (NIC) in the server can provide connectivity to the iSCSI SAN. This is an acceptable solution for relatively lightly loaded servers.

Alternatively, you can use a hardware-assisted iSCSI initiator. In this case, an iSCSI host bus adapter (HBA) in the server enables the server to connect to the iSCSI storage network over Gigabit Ethernet. First-generation iSCSI HBA performance is well-suited for workgroups and departmental storage requirements. The availability of TCP/IP offload engines (TOEs) further improves the performance of

iSCSI SANs, sometimes referred to as iSANs or IP SANs, require initiators (device drivers or HBAs) and targets (disk subsystems).
Click here to enlarge image

iSCSI at Gigabit Ethernet speeds and will allow vendors to scale to 10Gbps Ethernet. An iSCSI SAN uses standard Gigabit Ethernet switches and cabling. On the storage side, you need iSCSI targets (storage subsystems).

The major advantages of iSCSI SANs over Fibre Channel SANs are lower cost and reduced complexity. As a result of using Ethernet, core transport layers can be managed using existing network management applications. High-level management activities of the iSCSI protocol (i.e., permissions, device information, and configuration) are layered over or built into these applications. Some iSCSI HBAs and Gigabit Ethernet switches are less than half the cost per port of Fibre Channel HBAs and switches.

For these reasons, as well as the support of major vendors, deployment of iSCSI could occur relatively quickly.

iSCSI benefits

By combining SCSI, Gigabit Ethernet and TCP/IP, iSCSI can deliver these key advantages:

  • Builds on stable and familiar standards, infrastructure, and components;
  • Creates a SAN with reduced TCO and installation/maintenance costs;
  • Provides a high degree of interoperability; and
  • IP datagrams can travel over the global IP network with practically no distance limitations.

iSCSI SANs will initially be most attractive to organizations with the following environment:

  • Distributed IT environment;
  • Proliferation of Intel architecture servers in divisional, departmental, and workgroup environments;
  • Business requirement to consolidate data storage and management to improve operational efficiency, data availability, and storage resource management; and
  • Budget and staffing limitations, which preclude a Fibre Channel SAN.

Initial iSCSI deployment is likely to occur in the following environments: storage consolidation for Windows servers; networked storage for rack-mounted Linux servers; and centralized data management for departmental and divisional data centers.

Server proliferation is causing storage management in many distributed environments to become complex and expensive, particularly for applications experiencing significant data growth, such as Microsoft Exchange. iSCSI enables centralized network storage, potentially delivering significant cost savings and improved data availability and recoverability.

Many organizations are considering "compute farms" consisting of dozens of rack-mounted servers running Linux to significantly reduce the cost of running I/O-intensive applications.

In this type of environment, DAS often makes data management prohibitively complex, and traditional SANs are too costly. However, iSCSI-based SANs can solve the complexity problem at an affordable price.

Many companies need to support their departmental and regional data centers with minimal staff. However, the data management needs of the servers often make this impossible. iSCSI enables companies to solve this problem by using standard Ethernet infrastructure to connect servers to storage devices.

Organizations with server proliferation and data growth problems in departmental and workgroup data centers will be the first to benefit from the introduction of IP storage and iSCSI, which will accelerate the transition to networked storage.

As the technology matures and performance increases, iSCSI storage networks will gradually displace DAS in high-end data centers and mission-critical environments.

David Dale is the Industry Evangelist at Network Appliance (www.netapp.com) in Sunnyvale, CA.

This article was originally published on January 01, 2004