Such confidence was grounded in the understanding that storage buyers are generally conservative people--which is why they're willing to pay premium prices for quality products.
Thanks to a solid reputation, storage subsystems from companies such as EMC, IBM, Network Appliance, Compaq Computer and Hitachi have commanded large margins--despite the fact that disk drives are among the lowest-margin products in technology.
Both the buying and investment markets have placed an extremely high value on subsystem technologies that configure and manage a lot of disk drives.
The storage industry has maintained a theory that storage purchases would continue apace even if the rest of the industry put on the brakes. Insiders have pointed to several studies and predictions that show storage requirements would continue to accelerate over the next several years.
Now, Network Appliance's report has shaken that sense of security. Network Appliance is a credible company with reliable products in a market many have anointed the fastest-growing in all of storage.
Thanks to a reputation for creating hardware that is easy to install and manage, Network Appliance has established an excellent track record and is the clear market share and technology leader in the network-attached storage (NAS) industry. Although Network Appliance gained EMC as a new competitor, it's ludicrous to think EMC could have spoiled Network Appliance's status so quickly with a "me-too" product in sheep's clothing.
Instead, this reversal teaches a different lesson: The market is unexpectedly conservative when it comes to fiscal responsibility as well as to product quality. The market might not start buying inexpensive storage clones, but it will learn to make do with what it has until the requirements for storage are more clearly defined. It also means that associated products, such as storage-area networks, are coming under closer scrutiny.
Excitement over the potential for storage-area networks has allowed SAN-only companies such as Brocade and Emulex to build fast-growing, successful businesses that serve early adopters. However, SANs are confusing, and most vendors and customers have difficulty explaining their bottom-line value.
Until their value can be translated into hard dollars, SAN purchases are now less likely to get funded. The integration of SANs is still extremely difficult, and conservative customers who plan seriously to install one will likely be frightened off by the lack of qualified integration partners and the costs of the projects.
The SAN market won't heat up again until "killer apps" emerge that justify the expense to a wider range of customers.
Despite what some industry pundits are saying, storage "virtualization," or pooling, is not a killer application--yet. For now, it is completely proprietary, which makes it expensive and impossible to integrate in a mixed-vendor environment.
The cure shouldn't be worse than the disease, and it isn't clear that the market will continue to view SANs through rose-colored glasses. The resulting disappointments will be spread throughout the storage industry.
SAN-only companies such as Brocade and Emulex will be most deeply affected. Companies that derive a high percentage of their revenue from SANs, such as EMC and Qlogic, will also be affected. Veritas, as a company that provides storage-management software to all types of storage will likely have an easier time weathering the storm, but they will likely see a downturn, too, as overall storage spending takes a breather.
Storage is a necessity and has a strong future in the long term, but it is also an industry that has been overvalued. Fibre Channel has shrunk too far to be a major force again.
The next big wave of growth will accompany the next fundamental technology advancement. That will probably be deployment of IP-based storage, so long as the IP networking industry takes storage more seriously and eschews the easy, "we-are-the-world" approach. If the networking world misses the point, InfiniBand will be close at hand ready to make its run at storage success.