Emerging iSCSI viability depends on startups

Feature

Emerging iSCSI viability depends on startups

An emerging storage protocol, supported by industry heavyweights such as IBM, Cisco Systems and EMC, promises to make storage area networks available to small and medium-sized businesses, and create a market forecast to be worth $5bn over the next four years.

Because they can use inexpensive Ethernet adapters and switches to transfer data, Sans built using the Internet Small Computer Systems Interface (iSCSI) could be cheaper and easier to manage than those based on the dominant San interconnect, Fibre Channel.

But analysts warn that until smaller startup suppliers prove that there really is a market for iSCSI, the likes of Hewlett-Packard, IBM, and EMC will be reluctant to support it fully in their product line.

When it was first proposed three years ago, the iSCSI protocol looked good on paper. It was based on a simple idea: Take the SCSI interface that most servers use to send block-level data to their hard drives, bolt it onto the TCP/IP networking protocol, and suddenly you have a storage protocol that can run over Ethernet.

Approved by the Internet Engineering Task Force in February, iSCSI arrived a few years behind Fibre Channel. And, while Fibre Channel components are much more expensive than their iSCSI counterparts, and require special networking expertise, there are products shipping that use it.

Some companies have dabbled with iSCSI over the past few years. IBM was first out of the gate, releasing its TotalStorage IP Storage 200i array in June 2001, and HP released an iSCSI router called the StorageWorks SR2122 router this February.

But the bulk of iSCSI products have come from smaller companies such as Network Appliance and Sanrad, as well as Cisco, which has released a number of products designed to connect iSCSI to Fibre Channel and SCSI arrays.

Though the big companies remain tight-lipped about when, if ever, they plan to integrate iSCSI into their existing storage product lines, observers believe that this will happen only after smaller startup companies have proved that there is a viable market for iSCSI devices.

"The startups are the leaders because the startups are the ones that are out there today," said Larry Boucher, the chief executive officer of Alacratech. "They are the ones who are going to pull the big guys into this business."

The market for iSCSI arrays was a miniscule $12m last year, according to IDC,  but the research firm predicted that it would rapidly grow to the point where the HPs and EMCs of the world will be forced to take notice, hitting $4.9bn by 2007.

iSCSI's growth is expected to far surpass that of Fibre Channel over the next four years, but its total market size is not. IDC predicted that Fibre Channel will remain entrenched at the high end, with its broader vendor support and superior speed, while iSCSI will make inroads in small and medium-sized businesses because of its lower price and ease of use.

Gartner analyst James Opfer agreed. "iSCSI is not going to displace Fibre Channel in the current state of affairs," he said. With Fibre Channel running at a data rate of 1.6Gbps (bits per second) compared with iSCSI's 1Gbits/second, it has enough of a performance advantage to remain viable.

But for those who are happy with iSCSI's 1Gbits/second transfer rate, and who may not have the money to purchase the more expensive Fibre Channel network cards and switches, iSCSI can be compelling.

Today, the iSCSI market is dominated by switching devices, like Sanrad's, that convert iSCSI traffic to Fibre Channel or SCSI, but iSCSI storage systems are beginning to emerge.

Network Appliance has been shipping two arrays - the NetApp F800 and FAS900 - since February, and next month startup Intransa will begin shipping what it bills as the first highly scalable iSCSI array, capable of up to 10Tbytes of storage. 

IBM said its 200i product was a "primer for the market", rather than the basis of its iSCSI product line. "What we wanted was to have a platform that people could play around with,"  said IBM director of storage Strategies Clodoaldo Barrera. "But we did not think then, nor do we think now, that the 200i is the final optimised ... product."

IBM expected to introduce more iSCSI products by early 2004, Barrera added.

EMC said it would support iSCSI in its Clariion and Symmetrix product lines, and HP said that it could ship iSCSI arrays as early as 2004.

Sun Microsystems storage chief technology officer Balint Fleischer said that while iSCSI did "hold promise" in the small workgroup space, he believed that the "the reality is iSCSI is not a mature technology, and we have to see how it matures".

While Sun was an early investor in iSCSI technology, "the more time progressed, the more sceptical we became about how fast and how soon iSCSI is going to be adopted," Fleischer said.

IDC predicted that iSCSI will first catch on as a direct attached storage technology, with 85% of iSCSI sales being direct-attach in 2004. But after that, SAN deployments will begin to dominate the iSCSI market, making up 65% of total sales by 2007.

But iSCSI was unlikely replace Fibre Channel for the foreseeable future, and. IDC predicted Fibre Channel SAN sales would remain ahead of iSCSI for the foreseeable future, growing to more than $6bn by 2007.

Robert McMillan writes for IDG News Service


Email Alerts

Register now to receive ComputerWeekly.com IT-related news, guides and more, delivered to your inbox.
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

This was first published in May 2003

 

COMMENTS powered by Disqus  //  Commenting policy