Those who follow me on Twitter were treated to the perspective of my new travelling companion, Blue Cow, during my trip to VeeamON. This week, I continue the tradition. Alas, Blue Cow was not allowed to join me, so Clarence the horse has come with me instead.
Clarence is a little cross-eyed, so his perspective is a little more sceptical than the general enthusiasm of Blue Cow. He thinks he’s a unicorn (his horn is invisible, apparently. What, like unicorns have to follow your rules?) so I guess he’ll fit right in with the Amy Lewis crowd. Maybe he’ll make a guest appearance on #EngineersUnplugged?
But I Digress
This week Clarence and I are at NetApp’s Insight conference. I’ve written a little about it before, but recent news events have me in the mood to prognosticate a little. That and not a lot of sleep on my flight MEL->LAX.
EMC are buying Cisco out of VCE. This has been coming for some time now, despite protestations from EMC that the partnership is strong. This little piece on the background to EMC/VMware’s acquisition of Nicira puts an interesting perspective on things.
Commentary from several people at EMC whom I respect, such as Chad Sakac and Mark Twomey, led me to second guess my own instincts and analysis of the issue and strategy. Even after getting some credible information from contacts before the VCE story broke at Bloomberg. The way this has come about has taught me two things: firstly, I should trust my own analysis more, and secondly, that I didn’t have enough information to trust my analysis because I wasn’t watching enough of the right things.
Now part of this is because I have an actual day job that isn’t as an analyst, and my clients don’t pay me to watch the infrastructure industry the way the buy-side guys do. It’s unlikely that I’ll ever have access to that level of information. But what my experience with other areas of information analysis has taught me is that less volume, higher quality information is a lot better than acres of mindless guff. There’s a lesson there for the BigData boosters, too, btw.
So with that in mind, I’m gonna make predictions about this week based on what I know so far that’s on the public record. I’ve not signed any NDAs (because I don’t do that) and I haven’t had any embargoed briefings. Yet.
Justin Gets It Wrong
Cisco is pissy at EMC right now, so they’ll go large on their NetApp partnership, and its vBlock competitor FlexPod. FlexPod is doing decent numbers all by itself, close to $2bn annualised demand, apparently. VCE is doing about $1.8bn (found via EtherealMind). And with reports that Cisco didn’t make any money on their investment in VCE, FlexPod makes a lot of sense for them.
It also makes a lot of sense for NetApp, because, well, what other options do they have? They don’t have the full breadth of software and hardware that EMC has, and they don’t have a VMware, and their latest offerings in the FlashRay feels kinda lacklustre. NetApp, in my opinion, has been sort of on-the-wane since their Spinnaker acquisition. It just took up too much attention, took too long, and when it landed, was too little too late. The world has moved on. Continuing on their previous path would have been a slow road to irrelevance.
But. And this is a big But. There are a lot of smart minds at NetApp, and it would be a mistake to think they’ll die quietly. There will be some significant announcements at Insight this year, and they are announcements they want the world to hear. I know this because this is the first year NetApp have invited media to what is traditionally a customer and partner event, full of secret squirrel knowledge. And not just media, also analysts and bloggers. This is now their major tech conference of the year, though happily they didn’t rebrand it as NetAppWorld. (As an aside, where’s the OracleUniverse or something? Missed opportunity there, Larry. Branson has Virgin Galactic.)
It was at the last Insight I attended that Cisco gave a copy of the project California book on it to every single attendee as part of their NetApp partnership announcement (Rats! I should have brought it with me to maybe get it signed. Doh!). Cisco announcing some sort of significant partnership makes sense. But what sorts of partnerships make sense in the current climate?
Cisco are big on ACI being better than NSX (and while I think they’re complementary technologies, I agree that ACI appeals for a wider range of use-cases, such as VoiP phones, security webcams, etc.) so there’s potential there. Cisco also lack a real “hyperconverged” platform in a world that is hyperconverged mad of late. They have a partnership with SimpliVity, which I reckon is a precursor to an acquisition. With VMware announcing EVO:RAIL (and EVO:RACK to come, which is the bigger problem for Cisco), Cisco are not on the list of vendors bringing out a RAIL box. Dell is, despite OEMing Nutanix. HP are onboard. EMC will be doing something. But no UCS EVO:RAIL, because it doesn’t make sense.
Maybe NetApp get to use the SimpliVity software on hardware they build? NetApp has the supply chain and support organisation built. If I were Cisco, I’d be more likely to just do all that myself, though, because Cisco also has supply chain and support. Perhaps not a great idea.
But then, UCS doesn’t really play at the entry level where EVO:RAIL plays. NetApp has some stuff that’s closer to the commodity/volume hardware business that EVO:RAIL matches, so maybe, but really it’s not a great idea for either of them unless it’s used as a hook to get people into the ecosystem. Which is what VMware is doing with it, and why being a software vendor is super useful.
What’s the competing software that could run on Cisco’s, NetApp, and SimpliVity hardware? Bare-metal and containers? OpenStack? Hyper-V? OpenStack is the best match with the enterprise workloads side of things, and automation of things with an ACI APIC might make sense.
This has turned into a bit of a rambling, flow-of-consciousness thing, which implies that only the FlexPod idea is a strong one. Expansion of relationship, maybe a line extension or two.
I still think there might be something in the ACI/UCS side of things from Cisco, but I’m not sure what it would look like. Ah well, I’ll just have to wait for the announcements.
And Now We Wait
Prediction is very difficult, particularly about the future.
There’s some sleep-deprived reasoning for you. Now let’s see how wrong I am!