This month, we look at how the introduction of low-price disk drives will have a far larger effect than just lowering that portion of the cost of disk arrays. Along with existing RAID and storage area network (SAN) technologies, serial ATA (SATA) could become what it is trendy to call a disruptive technology--one which dramatically changes the way things are done. I might add iSCSI to that list, as well, and I fully expect to see the combination of all four technologies make storage networking as everyday as local networks in the next five to ten years.
But here's a less welcome kind of disruption to the progress of storage: the lack of tools. They're better than they were two years ago, but still not good enough.
What's wrong with the current crop? They either don't do enough or try to do too much; they don't work on enough platforms; they cost too much; they take too long to install and they just plain don't inspire confidence. There are exceptions, and I don't want to tar every vendor with the same brush, but that's what I hear from most of you.
What's gone wrong? There sure was a lot of venture funding thrown at tools startups in recent years. That's part of the problem. Startups aimed too high in an effort to convince investors that they were sitting on a diamond mine, not just a gold mine. Products seek to automate too much based on policies that don't yet exist in most shops.
OK, what about the big guys? They've got stars in their eyes, too. They think you'll plunk down $500,000 and up for a suite just because it has a recognizable trademark on the box. If that's not a good enough reason, how about because it integrates with their multimillion-dollar, manage-everything-in-the-enterprise suite? What, you never really fully deployed that either? What, there's actually no individual on staff who has to know the zoning on your switch, which jobs ran on your mainframe and your router traffic?
Engineers always like to analyze systems in terms of bottlenecks. Right now, the bottleneck is management software and without serious improvements, you won't be able to operate large-scale production storage networks with any assurance that they're actually doing what you want and not without an army of admins.
Vendors, here's some places to start in breaking this bottleneck:
- Forget strategic lock-in. Do something useful now, and build on that.
- Let's lose the get-rich-quick mentality. Sacrifice some of those high margins for more customers and repeat sales.
- Tell your development teams that if a product can't be installed in a week, go back to the drawing board.
- Come up with licensing schemes that are flexible and don't penalize growth.
- No standards, no future.