Reality IT: Successful SAN Implementation

Don't Speak Too Soon

We made the mistake of talking with vendors too early. They started pushing to close the sale as soon as we asked for information, but we didn't know what we wanted or needed at that point. And we hadn't done enough research to pin down the specific requirements for the bids.

Apparently, the vendors thought we were playing hardball, because the prices kept going down. But Eugene couldn't decide which bells and whistles he wanted, and anyway it was hard to figure out which bells worked with which whistles. Some vendors more directly experienced with the chosen products finally helped us articulate what we really wanted: a scalable, true Fibre Channel SAN that would be totally installed by the vendor. We ruled out iSCSI or serial ATA options.

The lords of procurement required us to issue a public RFP for such a large purchase, and we dutifully complied--after all, we're experts in such administrative CRUD (critical research for ultimate decisions).

In addition, we had to post a notice about the RFP on our Web site. You can imagine the inquiries we got in return. Needless to say, we passed on the vendors whose primary e-mail account was with Hotmail. We received quite a few e-mail messages from vendors that essentially build SANs from spare parts. I'm sure this approach works just fine for some customers, but it seemed too risky for ACME. We put some stiff requirements in our RFP, which promptly scared off most of the lower-end bidders.In the end, we received bids from just a handful of vendors--mostly trusted names, including a few of the large PC vendors that are now reselling SAN technology. We looked closely at which one was actually going to provide support; we didn't want one vendor installing the SAN, only to toss support to a different vendor.

Packing Slips

After a great deal of consideration, we finally selected a vendor. I can't tell you who it is, because I don't want it or other vendors guessing my true identity. The key factors in our decision were price, the proposed system, support, the vendor's understanding of our requirements and the vendor's level of experience. Our experience turned out to be an essential factor: The vendor we chose was one we had used before and greatly trusted.

We're now in the midst of installing the SAN, and we believe we'll face fewer problems--and a lot less risk--because of the extra time we put into mapping the requirements for the RFP, planning the technology and reviewing the bids. It's better to pay a price early in the process than to pay it later, when you're living with the consequences of making the wrong choice.

Hunter Metatek is an enterprise IT director with 15 years' experience in network engineering and management. The events chronicled in this column are based in fact--only the names are fiction. Write to the author at [email protected].0