The consolidation within the storage software industry continued last week with EMC's purchase of Astrum Software, a Boston storage resource management vendor. Astrum is focused on policy management, storage clean up, file migration, and archiving services for mid-to-large sized enterprises that have a heterogeneous mix of storage products.
Astrum had made only a little headway as it tried to carve out market share: the software was slow in coming and the first release of the product, as is often the case, was heavier on reporting capability than it was on management. By the time EMC stepped in Astrum had downsized to the bare minimum and had churned its management.
Three other things had also happened, however, by the time EMC bought the company: much of the existing product had been rewritten, management functionality was expanded and Astrum had picked up a first class reseller in Overland Storage.
Why did EMC want Astrum? Well, it was probably cheap. But why spend the money? After all, EMC had already purchased a number of software companies over the last few years as it tried to shake loose from its image of just selling iron. Most likely, the Astrum product will be wrapped together with the software it inherited through last year's Prisa acquisition.
The Prisa product, VisualSAN, had been aimed at managing performance, networks and configuration of midrange storage systems, and was "InfiniBand ready" (for those of you interested in the concept of InfiniBand ready, see Samuel Beckett's "Waiting for Godot"). Although several of us were impressed with what Prisa had created by the time of the software's second revision, its investors sold the company to EMC last September when it failed to pump enough sales through its OEM pipeline. (At the time, Prisa was working with storage providers Compaq, EMC and Hitachi).
Although the Astrum product was still limited in terms of functionality, what it did, it did well. Creating a single entity out of the Prisa and Astrum code may indeed contribute to a more robust offering, filling in the gaps in VisualSAN. With that, EMC can support its own midrange Clariion line and go after midrange competitors Sun and HP. It may also serve to preempt further inroads by Veritas into what EMC would like to see as its new marketplace.
One last point: Prisa and Astrum only achieved viable product with the second release of their respective products, and by then it was too late. The first release, in effect, was little more than a placeholder. Because of that, neither company could make market headway on its own.
Smart vendors will see a basic imperative here that they should all take to heart. And what is that insight? Why, it is Karp's Third Rule for Storage: when it comes to developing software, a guiding principle must always be BETTER NEVER THAN LATE.