Yesterday I read this article by Chris Mellor on the register. I always find Chris’s comments reveal good insight need, and he raised an interesting point about the future of the all-CISCO integrated stack (UCS+NEXUS+
Whiptail Invicta) but I too would like to put in my two cents.
From my point of view, CISCO has two options here and neither of them are ideal.
CISCO has, if not the best at least one of the best, server infrastructure products out there and the integration between it and its networking stuff makes the solution very appealing to a certain kind of customer.
Integrated stacks of various types have been hitting the market for a while now and their success is due to a couple of reasons:
1) Enterprises like end-to-end solutions and they always favor infrastructures which are easy to deploy and manage.
2) Hardware infrastructure is a small part of the IT budget, especially when you have to manage complex projects with tons of software and services. As such (the cost of an integrated system is often higher than the singles pieces combined), it works out to be affordable when compared to traditional approaches.
CISCO has been doing well on partnerships with NetApp and EMC in building integrated stacks (FlexPods and VCE vBlocks) and sales figures on the high end are showing interesting numbers. But, is this enough?
CISCO has a great but incomplete solution today and needs storage partners to make it integrated and “converged”. Some of its direct competitors (HP and Dell) can offer end-to-end solutions while hyper-converged guys Simplivity and Nutanix, for example) are becoming more and more relevant…
In any case, storage is a fundamental piece of any modern infrastructure and has shown the biggest growth. This obviously means higher future revenue.
From this point of view, the right strategy might be to further (and aggressively) invest in storage to build a full product line-up. In this case, a VSA (like Maxta) or a server-side caching software (like Pernix) could easily move more revenue from traditional storage to the software-defined (or, better, server-based) storage!
Leaving only the crumbs to the “partners”.
The drawback here is obvious. Losing partners like NetApp and EMC will hurt sales, especially in the short/mid term, and it won’t be easy to recover the situation… even though there are lot of reseller recruitment opportunities out there, especially now after the IBM/Lenovo deal.
…or give it up
The other solution would be to make Whiptail the next Flipcam.
Tactically speaking this is the best option, messing up with partners when you’re not ready to deal with the consequences is always a bad idea.
But as I said, that’s not a strategy, it’s a mere short term sales tactic: In a few months time both VCE and NetApp will be adding all flash arrays to vBlocks/FlexPods and then we will have three fully comparable stacks!
Will Cisco ditch Whiptail at that time? I don’t think so… but it’s clear that the more time passes, the harder it will be to do just that.
At the same time, not pushing Invicta will automatically give it (and its development team) an uncertain future and lead it towards a slow but sure death.
NetApp and EMC also have agreements with other server vendors and they can play on multiple tables at the same time. And I won’t go to mention the internal political problems that VCE will probably face now that VMware can also play an important role in the Storage/Networking space.
Why it matters
It looks like CISCO is in midstream now and needs to make a choice:
1) Invicta is a niche product at the moment and it covers only a small part of end user needs. CISCO has to invest (and push) more to become an end-to-end primary player.
2) CISCO can stick with its partners while playing a secondary role in the converged and hyper converged space.
I can’t see CISCO playing secondary roles in the DataCentre, can you?
In the next future, all traditional hardware vendors will have to seriously face new challenges like “software-defined”, hardware commoditization and hyper-convergence: many values in the field will be changing radically… CISCO already has some components that can make the difference and, simply by adding a few more (storage) pieces, they could build a solution with greater appeal.
Is it me or CISCO isn’t showing up the Invicta array in its website? I didn’t find any product web page and all the links are broken! (BTW, Whiptail website is still alive)