BMC's exit from open systems storage management is a wake-up call to the remaining vendors in that market.
Here's the wake-up call: Selling unproven software with many zeros in the price that takes weeks to set up and can actually only manage some functions on some of your storage is a tough sell to storage managers who haven't been able to add any staff in months and are barely keeping their noses above water as is. We've heard that from many storage managers and now software vendors are feeling the pain.
Multibillion-dollar BMC had dozens of customers for Patrol Storage Manager, often at more than $100,000 a pop, before it decided there was a bigger gold mine somewhere else. Some high-profile startups number their customer lists in single digits.
What do vendors have to do to get you to buy something you readily admit you need badly? For one, they need to address today's market, not some imaginary, lights-out storage world. Let's keep it simple and affordable.
Structure products so that functions come in more discrete modules that can be deployed in a phased manner, giving users a quick return without having to figure everything out up front.
Some shops only want reporting. Some shops that used to only want reporting now want policies with some automated enforcement--quota management comes to mind, for example. Some shops that have gone through that evolution are now ready to move on to some minor level of automated provisioning. No one should have to buy a full-fledged suite because one day they want to get there, even though now they only want a decent utilization report.
Above all, make this stuff easy to deploy and configure. It shouldn't take too much time for already harried staff to get an SRM package working.
The storage management software market is staring down a crisis in this awful economy. And the fault doesn't lie just with bad product management decisions at software companies. Some of the proprietary baggage of the storage industry is having a serious detrimental effect on the whole storage community.
At the top of the list is the high start-up cost for every storage management software vendor to buy proprietary APIs or to reinvent the wheel to manage every proprietary piece of gear. An industry with robust standards would have a much more streamlined development cycle.
Subsystem vendors spend a lot of time and money ensuring their disk controllers, cache and disk talk to each other in unique ways, and a lot less time and money making sure that it can be managed by the user in a heterogeneous environment. In a sane world, EMC would pay BMC to write software to make EMC arrays more useful. BMC--or anyone else--would be able to charge less and more people would buy the software and get the benefit of it.
OK, it's not a sane world. But making storage management software the bottleneck for the growth of storage networking is just plain crazy, and I have to believe the vendors that figure that out and fix it will be the winners.