A quote yesterday from Adrian Cockcroft's blog (if you don't know, he's Netflix's performance guru):

There has been a sometimes heated discussion on twitter about the term NoOps recently, and I've been quoted extensively as saying that NoOps is the way developers work at Netflix. However, there are teams at Netflix that do traditional Operations

-- Adrian Cockroft

So for anyone who was under the impression that large companies can function without operations in this "NoOps" cloudy fantasy land, you now know that you are wrong, and for John Vincent It Sucks to be Right.

Here was the list of things that Netflix does in the form of Operations (as compiled by John from Cockroft's blog):

• Metrics collection
• PaaS/IaaS evaluation/investigation
• Automation (auto-build, auto-recovery)
• Fault tolerance
• Availability
• Monitoring
• Performance
• Capex and Opex forecasting
• Outage response

John Vincent then goes on to nit-pick a couple of other things that Adrian says in his post.  But I think the key takeaway from John and Adrian was this:

The developers used to spend hours a week in meetings with Ops discussing what they needed, figuring out capacity forecasts and writing tickets to request changes for the datacenter.

There is no ops organization involved in running our cloud, no need for the developers to interact with ops people to get things done, and less time spent actually doing ops tasks than developers would spend explaining what needed to be done to someone else.

-- Adrian Cockroft

Netflix is still doing operations. What should be telling and frightening to operations teams everywhere is this:
The Netflix response to poorly run operations that can’t service the business is going to become the norm and not the exception. Evolve or die.

-- John Vincent

John Vincent also wanted to note that he doesn't want to seem like he's blaming the development or operations side of the story at Netflix and is grateful to Adrian for writing this insightful post.