* StorageTek attracts crowds at its StorageWorld Conference booth
Last week’s StorageWorld Conference in Long Beach, Calif., was interesting on a number of levels. Unlike many such gatherings, this was no vendor love fest with vendors outnumbering corporate IT types. A lot of this is probably due to efforts of the Association of Storage Networking Professionals, which is heavily involved in many aspects of the conference. Whatever the reason, there were many real-world aspects to this conference that often seem missing at other such get-togethers.
Two booths were clear winners, competing successfully with both the other vendors’ offerings and with the guy running the espresso machine.
The StorageTek booth had good crowds throughout. In addition to demos, the company made several product and strategy announcements. The most significant of these concerns its new IntelliStore data archiving solution, a disk and tape product offering tiered storage and compliance solutions for companies lacking the budget to invest in enterprise-sized devices.
StorageTek also delivered one of the keynotes. Unfortunately, much of this was lost in the brouhaha surrounding the company’s impending acquisition by Sun (see last week’s column). I estimate about half the chatter in the booth was on that subject, which must have been distracting. Still, StorageTek seemed to have more booth traffic than anyone else, so on balance it looks like it came out way ahead. See what publicity having someone buy you for $4.1 billion can bring! Try it.
British tech firm Njini was right next to the StorageTek booth, but obviously didn’t need its neighbor to generate interest. Njini has come up with what appears to be a revolutionary method of attaching metadata tags to unstructured data (“unstructured data” in essence is any data – text, bitmaps, rich media, it makes no difference – that does not reside in a database or some other structured format). The metadata associated with each file is then held for use in subsequent searches. What is particularly interesting is that the metadata is not just stored, but is interpreted so that intelligent judgments can be made about it. It is thus far more than a simple string search as it should allow decision-making to be based not merely on the content of a text string but on the meaning of the string as well.
As unstructured data represents about half the data currently being stored, any success in making that corpus more manageable can have profound effects on a system’s ability to understand, retrieve and use stored information. The Njini approach looks at the raw data to understand file type, key worlds and end-user associations; groups like objects; and then provides automated policies to manage the movement of data objects across the various storage tiers. It also enforces policies that eliminate data duplication and that ensure compliance, however that is defined within the corporation.
All of this sounds like a front end to an information lifecycle management implementation, but because it enables intelligent searching of files based on their contents in fact it extends well beyond today’s standard ILM definitions. It is the not the ability to extract metadata content that does this; rather, it is Njini’s apparent ability to interpret the data more intelligently and then to create more intelligent metadata and make the metadata available for more sophisticated operations when data must be accessed.
And finally, one large vendor did something that clearly showed it hadn’t given much thought to training the people they sent out to this show. The Fujitsu booth had technology demos of Serial-Attached SCSI (SAS), a topic of interest to me and one that you’ve seen covered in this column on many occasions. The slide presentation accompanying the demo contrasted “SAS and SCSI,” when clearly the distinction it intended to make was between parallel and serial SCSI. I pointed this out to the person manning the booth, suggesting that as SAS is still brand new and just being rolled out into the marketplace this might cause some confusion among IT buyers. Fujitsu guy stared at me for a moment and then shared the following philosophical insight with me: “If they don’t understand the distinction, they shouldn’t be in the industry.”
I am guessing there may be some among you who, while you may not yet be up to speed on this new technology, nonetheless still plan on staying in the industry. If that is the case, and if you are looking for information on SAS before investing in your next round of arrays, I am sure that other vendors such as Maxtor and Seagate will be delighted to provide you with information on the topic. Alternatively, the SCSI Trade Association’s Web site (http://www.scsita.org) is an excellent repository for such knowledge, and includes a dandy paper on the topic that you can download for free (https://www.scsita.org/aboutscsi/sas/STS_SAS_WP.pdf). Modesty forbids that I mention the author’s name.
