Well…well…well. This morning when I jumped on Twitter I saw a few tweets that weren’t totally unexpected. They were concerning Cisco acquiring a storage company. How long have we heard rumors about them being interested in NetApp? Then over the last year or so the rumors about Nimble have been loud and often so when I saw a tweet that said “I guess Cisco is now in the storage business!” I assumed it was Nimble. Wrong.
We’ll come back to that in a minute. But first, Why? That’s the question everyone is asking today and it’s already one I’m being asked like Chambers called me up to brief me. But there are many different things going on at Cisco and I’ve had a few “off the record” conversations today. The bottom line is that Cisco is extending SDDC further than many have expected. It’s not just about SDN or SDS (Software Defined Servers – Cisco UCS and Service Profiles). They want to software define everything and they don’t feel they can do that without owning at least some part of the storage piece.
CIsco hasn’t said a lot about Insieme, the spin-out SDN/ACI (Application Centric Infrastructure) focused company. But I think we’re going to start hearing more and more about that and this acquisition will make more sense over time as that entire strategy starts to gel. Until then this just makes for interesting conversation. But…the push to get the Insieme developed products has quickened so it’s going to be sooner rather than later.
Is this a shot across the bow to EMC/VMware? I don’t think so. In my opinion if it was they’d have chosen differently and that leads us to the discussion around “Why WHIPTAIL?”. If Cisco really wanted to ruffle feathers they’d have gone with a platform with a wider use case like the previously considered Nimble. Instead you need to think about this outside the constructs of just general storage…even high-speed general storage. It’s about low-latency application performance which gives you some insight as to why WHIPTAIL.
WHIPTAIL provides a platform that already speaks many protocols. FC, iSCSI, CIFS, NFS, and SCSI RDMA along with Infiniband. This adds up to a pretty interesting low-latency play for Cisco and one that isn’t just confined to the virtual space (such as Tintri). Depending on how deep they go here it may be a niche move for some very latency sensitive applications (think trading applications) or just a great platform to accelerate all applications by putting an all flash array out there…possibly as a cache tier (ala Project Thunder from EMC) or just flash storage. Then you layer on Cisco’s SDDC vision to make that totally software defined and you get a pretty interesting overall strategy.
WHIPTAIL will be in the UCS business unit and as Stu mentions here it is going to be a UCS-first push and then position WHIPTAIL. Word is that we’ll see Cisco branded/sold WHIPTAIL in its current form but that will change over time and become more integrated with the rest of Cisco’s SDDC strategy. This initial foray in to storage may cause friction with the usual players that Cisco works with but I thnk it will be hard to convince many existing storage partners to go that direction. A pure Cisco partner may have different ideas, but do they have the relationships within their customer bases to do anything?
Is Cisco making the right move? If Chad Sakac isn’t sure (or at least won’t say) here how can we mere mortals be expected to know? We don’t…but their overall plans are starting to take shape. My personal opinion is that Cisco will have a hard time convincing people to let them own all these pieces of the datacenter and much of it will come down to how well they play with others. Either way, Cisco is putting together a pretty good hand of cards across SDN, SDDC, networking hardware, server hardware, automation (CIAC and UCS Director/Cloupia), and more.