Picking up where we left off in Part 1 of our look at the state of true integration in the storage sector, Part 2 addresses whether storage technology has reached the point where true integration is even achievable, discusses the possibility and likelihood of cooperation between storage vendors advancing the promise of true integration, and reveals whether a common set of management protocols such as SNIA's CIM-SAN-1 could change the future of integration.
To some folks in the storage industry the words ‘true integration’ are nothing more than an oxymoron. After all, as storage vendors continue to boast about true integration, storage customers continue to become more and more perplexed by it. And some are even beginning to wonder if ‘true integration’ should be taken off this year’s holiday list wish and placed on next year’s list instead.
Are Proprietary Protocols the Culprit?
Much of the talk concerning the challenges facing true integration seems to circle around the contention that perhaps the only reason these challenges even exist at all is because storage vendors are not always aware of the latest arrays offered by their competitors and may not have the drivers to support the various hardware available.
The other side of this conversation is that maybe these challenges exist because many storage vendors manage their storage arrays using proprietary protocols that make interoperability a challenge, and require either continual cross-licensing or reverse engineering by competitors.
Paul Ross, director of storage network marketing at EMC, believes that the issues with proprietary protocols are partially true when it comes to true integration and management systems. “The deeper one wants to go in terms of managing heterogeneous storage arrays, the more reliance there is on complying with management standards such as SMI-S and CIM,” says Ross.
However, Ross also points out that in the early days of Fibre Channel SANs, server I/O device drivers that could operate in a switched environment were brand new, and storage vendors had to adapt their then current ‘arrays’ to talk to these device drivers. “But,” he says, “today’s crop of storage arrays benefit from 5 years of development in the fabric device driver space, and there is surprisingly little difference in how switch fabric is implemented in the arrays.”
True Integration: A Two-sided Sword
Integration is a two-sided sword. The only way for vendors to achieve ‘differentiation’ is through proprietary protocols, and the need to differentiate inevitably leads to incompatibility. “Every vendor has its own solution to address storage issues, and that’s where solutions start to vary and diverge,” says Ross.
“That’s the nature of the business, since as competitors, companies do not share that much information with their rivals,” he continues. He also believes that there’s a slim chance that vendors will get together on their own to sort out the parameters/standards in such a way as to fit the customer’s application environment in a unified manner.
And the issue may not simply go away because of standards such as CIM. This is because, says Wayne Lam, vice president of professional services for FalconStor, from a management point-of-view, “you may have a single platform, but all the specific timings, error handling, etc. and all things not part of CIM dictate the behavior of each of the storage devices.”
“Each storage vendor has their own creative way of providing functionality, causing differences and obstacles to integration,” Lam continues. “For true integration to become a reality, you always need something that’s a neutral platform to talk to all of the storage and shield the host (consumer of storage) from any differences represented by the storage.”