Some months ago I speculated on the future of IBM’s storage roadmap; a post which I believed caused some consternation in IBM as it foreshadowed some of their recent announcement. I expect that over time even more of that entry’s predictions will become fact. For what is basically a packaging exercise, the v7000 is actually an interesting announcement in that it shows that IBM do want a piece of the storage pie and they are going to use their own products to do so.
I believe that the v7000 will cause IBM some real problems around the XIV message; with the new GUI, the v7000 is pretty much as easy to manage as an XIV and realistically, that was XIV’s biggest selling point. The v7000 is more flexible, potentially more scalable and more performant that the XIV and I wonder how much of an influence that it’s development had on Moshe’s departure. I like to tease IBM that Barry Whyte is the new Moshe; he was brought up on the mean streets of Glasgow and no one crosses a WeeGie.
There are some features missing at launch which are interesting; the scalability for example seems to be being constrained for some reason. Certainly SVC can support an 8 node cluster at the moment and Barry has stated that there are no real issues going beyond this. So why doesn’t the v7000 not have clustering from day one? Perhaps to allow XIV some breathing space to get the long roadmapped clustering out of the door. A cluster-enabled 7000 would be a very interesting prospect and I wonder whether we’ll see a v9000 which offers clustering?
v7000’s integration with VMware does at first glance appear to be lacking; at the announcement, I don’t think VMware even got mentioned but v7000 shares all the current SVC/VMware integration points. However I do believe that IBM need to up their game on the VMware integration.
No compression or dedupe; this isn’t suprising really as IBM haven’t have StorWize for that long although it’s long enough to for StorWize to become the storage brand; certainly in the mid-range space. But I’d expect to see announcements around compression and dedupe in the next twelve months, I think IBM will motor on that one with their big focus on Storage Efficiency, it’s really obvious that they intend to do so. I’m pretty sure I saw Steve Kenniston sitting in the audience at the v7000 launch, I’d expect him to be on stage next time.
There’s a few really nice little touches which show that IBM are listening to customers and understand some of the day-to-day pain of the Storage Admin; for example, the v7000 comes with a USB key which can be plugged into a PC and all the initial configuration can be built there without the need for a serial cable. Once you have finished the configuration; plug the key into the v7000 and it’ll configure itself. So many laptops these days don’t come with serial ports, this shows some real thought on IBM’s part.
SONAS integration; will we see a unified interface from IBM to enable SONAS, SVC and the v7000 to all be managed from a common simple GUI? I know NetApp will jump up and down if anyone was to call that Unified Storage but I can see IBM doing something like that. There’s probably little reason why SONAS and SVC cannot co-exist on the same physical hardware. SONAS will certainly get Storwize integrated; I’d expect a rapid rebrand of the SONAS product to bring it in line with the Storwize branding.
OEM relationships with LSI, NetApp and DDN are probably going to become less crucial to IBM; IBM will have control of their own storage destiny again and I fully expect the number of OEMed products to reduce with the LSI relationship probably the first under serious pressure.
And IBM do amuse me; if EMC or NetApp had launched the v7000; there would have been fireworks, orchestras and ticker-tape parades; we would have had a teaser campaign and the whole works. This was all rather understated, there were bold statements made but there wasn’t the chest-thumping that we have come to expect from storage announcements.