Cloud computing is starting to come of age. It has fundamentally altered the IT landscape, dramatically boosting IT agility while lowering costs. What started out as a side project for companies like VMware has led to the proliferation of cloud providers and stacks from IaaS providers based on OpenStack, PaaS providers like Cloud Foundry, and SaaS providers like Dropbox and Salesforce.
This has led to the interest of the current giants of technology. EMC, for example, has purchased Virtustream, bringing it into its federation, and on Wednesday (June 3, 2015) Cisco announced that it has purchased Piston Cloud to fill out its Intercloud vision, although to date there has been no announcement on what it will cost.
With this purchase, Cisco is, according to its blog post, moving one step closer to realizing its ambitions. Now, with all that said, what I find more interesting are the side plays here. Firstly, this could be a tit-for-tat response. Former fremeny EMC bought a cloud company, so Cisco buys a cloud company. Very unlikely, especially as the acquisition of a company is not like buying milk from the grocery store. So, what exactly is their master plan?
Well, not being a fly on the wall, that is a difficult question to answer. However, to get a hint of the direction in which this is going, you need to look at a couple of Cisco’s other recent purchases, mainly Metacloud and Tail-f. The purchase of Piston Cloud will enhance those products’ automation, availability, and ability to scale.
It is only by looking closer at Piston Cloud and the abilities that that acquisition will bring into play that can you start to get a hint as to Cisco’s possible end game. John Chambers has made no bones about his desire for Cisco to own the next-generation data center. Cisco is still the champion of many a data center’s network stack, and it is gaining ground in the compute layer with its UCS portfolio. Its only real weakness is in the storage arena; the acquisition of WHIPTAIL appears to have amounted to nothing.
Piston Cloud believes that compute, network, and storage should be in the same device. This focus on being hyperconverged brings major benefits as your compute, network access, and storage becomes fungible, or interchangeable. This, coupled with UCS’s ability to provide inert compute devices (i.e., compute with no personality), merges neatly with Piston Cloud’s other main benefit: that of being able to deploy from ground zero a fully working OpenStack cloud to any device in fifteen minutes. This blank canvas can also deliver Hadoop, Spark, and container-based tools like Mesos, Kubernetes, and Docker. This is powerful stuff, and it takes the headaches out of delivering OSS-based services. No need for a PhD in software deployment, or a shrine to the gods of OpenStack pre-reqs.
So, back to my original statement about this being a poke in the eye for former frenemy EMC and the federation. Yes, it is, but it is not a tit-for-tat playground poke in the eye. This is a grown-up response to the threat that EMC has in its portfolio, with VMware products like EVO:RAIL, EVO:RACK, NSX, the vRealize Suite, and EMC’s storage products, Pivitol and now Virtustream.
Cisco now has the ability to deliver a fully orchestrated cloud that, after integration with Metacloud and Tail-f, could manage storage, compute, and more importantly, networking, both physical and SDN, from a single management point. This will give it a powerful position in the SDDC. All vendors worth their salt have an addressable API. With Piston Cloud, Cisco could quite easily become the one ring to manage them all.