Netflix: Not interested in cloud 'heavy lifting'
Amazon gets them up in the sky, we bring the rain
If you're thinking that you might be able to sell Netflix an alternative cloud to Amazon Web Services, or better still, convince the online TV show and movie distributor to go back to building its own infrastructure using an open source cloud stack such as OpenStack in the hopes of saving money, just forget it.
Adrian Cockcroft, the cloud architect at Netflix, pours a whole lot of cold water on the OpenStack effort despite the fact that he very much wants Netflix to have an alternative to AWS.
Virtualization juggernaut VMware would like to convince IT shops that virtualization is the new hardware, and therefore worthy of the kinds of profit margins hardware used to command but no longer does (unless you are Apple). But if the way the top techies at Netflix are thinking is any indication, then the control point is a little higher up the stack somewhere between the infrastructure and platform cloud layers. And perhaps more significantly for new companies and the young entrepreneurs that will by and large be establishing them, they don't want to know about the underlying hardware that runs applications and will never consider directly investing in it.
Having been there already, explains Cockcroft on his personal blog, Netflix does not want to try to compete with Amazon in the cloud business – even if Amazon Prime is taking on Netflix in the online entertainment distribution business. (Netflix is one of the dominant players, but has not yet achieved verb status, as Google has in search.)
"I come to use clouds," explains Cockcroft, "because I work for a developer oriented company that has decided that building and running infrastructure on a global scale is undifferentiated heavy lifting, and we can leverage outside investment from AWS and others to do a better job than we could do ourselves, while we focus on the real job of developing global streaming to TVs."
And iPads and PCs. Don't forget those. In my house, we watch more Netflix on the iPad than we do on the widescreen.
Cockcroft joined Sun Microsystems back in 1988 and rose up through the ranks, helping the upstart Unix server vendor crack the corporate market and tune their Java applications until he moved to online auctioneer eBay in 2004 as an operations architect. Cockcroft was among those who set up eBay Labs and its public Web API stack. Three years later, Cockcroft jumped to Netflix, and now that company has become one of the biggest users – and biggest fans – of Amazon's cloudy infrastructure.
Neither Cockcroft nor Netflix has anything against OpenStack and the companies that are trying to create a more scalable alternative to the various cloud fabrics that are out there either trying to emulate AWS or are built using VMware's ESXi hypervisor and vCloud extensions or Microsoft's Hyper-V and soon Azure platform services. Cockcroft doesn't talk about VMware or Microsoft clouds in his blog, but it is pretty clear from his statements that he doesn't think OpenStack has much of a chance over the next couple of years of competing head-to-head with AWS.
"Some of the proponents of OpenStack argue that because it's an open source community project it will win in the end," writes Cockcroft. "I disagree. The most successful open source projects I can think of have a strong individual leader who spends a lot of time saying no to keep the project on track. Some of the least successful are large multi-vendor industry consortiums."
Cockcroft said that he attended an OpenStack development meeting a few weeks ago and was dismayed that OpenStack would not have a "production-level" release until the "Essex" version in March 2012. More importantly the OpenStackers were "the consortium of people who can't figure out how to compete with AWS on their own," and they were still down in the guts of the cloud fabric defining features by committee while AWS was adding features driven by real customers.
"While it seems obvious that adding more members to OpenStack is a good thing, in practice it will slow the project down," says Cockcroft. "I was once told that the way to kill a standards body or consortium is to keep inviting new people to join and adding to its scope. With the huge diversity of datacenter hardware and many vendors with a lot to lose if they get sidelined I expect OpenStack to fracture into multiple vendor specific 'stacks' with narrow test matrixes and extended features that lock customers in and don't interoperate well."
But don't get the wrong idea. Netflix is still rooting for OpenStack, and would probably like to see the full AWS API set implemented in OpenStack. This will almost certainly not happen unless the industry demands it over the heads of NASA and Rackspace Hosting, which founded the OpenStack cloud fabric project back in July 2010 after NASA became frustrated with the EC2-alike cloud created by Eucalyptus Systems.
"I haven't yet seen a viable alternative to AWS, but that doesn't mean I don't want to see one," explains Cockcroft. "My guess is that in about two to three years from now there may be a credible alternative. Netflix has already spent a lot of time helping AWS scale as we figured out our architecture, we don't want to do that again, so I'm also waiting for someone else (another large end-user) to kick the tires and prove that an alternative works."
Unlike the content delivery networks, clouds are not yet interoperable and not quite commodities. But in the end, the clouds that run Netflix's applications might be as anonymous and uninteresting as the CDNs that Netflix actually uses to stream those movies that its applications control. ®
Sponsored: Benefits from the lessons learned in HPC