DevOps is no excuse for cowboy devs. Right. Let's talk Composable Infrastructure

Bollox. HP(E) lost it when they were taken over by EDS. Before that they engineered solutions - reasonably honestly, even if not always successfully (see rest of industry). The Compaq reverse takeover started the rot, then EDS, and now it really is all about how you can screw the customer - step one, big wool blanket, eyes, apply.

"While the comments section of this article will no doubt fill up with innumerable snippets about how the old-timers are perfectly happy with their pet servers running on an abacus they build out of rocks and crushed flower stems,"

It's like you have 1980s_coder in the room with you while you're writing....

The thing to remember about pets and cattle is that the latter get slaughtered. Whilst you're busy herding your cattle your salary is probably being paid by business running on some pet server. Or, looking at it another way, when some service goes TITSUP maybe it was because one of the cattle got slaughtered.

Minor point

"DevOps or Continuous Integration (CI) types"

You actually mean Continuous Deployment here. Continuous integration is a different thing not related to pushing servers and code out, although often done within the same workflows and toolsets. In this instance though it looks like you definitely are talking about CD rather than CI.

I worry that..

software defined infrastructure takes the same route as agile development which around here mostly means waterfall without any testing.

Its all fun and games until some moron installs oracle on a silicon graphics UV3000.

"a significant chunk of enterprises and government agencies are gearing up"

Exactly. It's no suprise it's HPE talking about that. You need a significant amount of resources to make it work properly, which may be still outside the capabilities of many mid-sized and small companies. As soon as you start to spend more and more resources on automating most tasks, you still need enough resources to deliver new features that can actually be sold (and bring in enough revenues to pay it all), because no customers of yours will buy your DevOps efforts. It is true more quality pays, but still, you need more capable, appealing products - especially if you're an ISV, and not part of an internal development team without market competition.

I won't blame too much mid- and small-sized business if they can't really afford the costs in hardware, software and people to tackle such an effort.

Two lines of code...

"Far more time is spent automating testing, quality assurance, regression testing and error handling than is spent pushing out new code or automating deployments.

The result is a remarkably agile company, one with processes in place to handle a variety of crises, from security problems to physical disasters to massive spikes in demand for resources."

Indeed and those two lines of code per quarterly release will be fabulously good, along with all the changes to support regression, QA, automation, etc. Pretty soon we won't need coders at all and we won't bother altering the product. We'll just keep improving the testing until everything is a known issue...

Humbug (and a good way to be overtaken by someone who actually ships product...).

CI/CD and devops are just the start

You need a lot more than that to be able to run Chaos Monkey non-disruptively. A highly redundant, loosely coupled application architecture for starters.

5 years of HP[E} Declining overall revenue

HPE can borrow other company's innovation and label it their own. Intel RSA and Brocade NFV are examples. None of the Composable Synerygy technology claims are possible without the actual innovation from processor, memory and HDD/SDD hardware companies. HPE's contribution appears to be marketing. Notice HP revenue in 2010 was $127.25B and it is now $103.36B. in 2011 Meg W claimed she would make China 25% of HP revenue. At the time China revenue was less than $700M and China had a massive sales and management staff turnover issue. Today China is probably less than 1% like was before Meg. Memristor was supposed to be available in 2013. The Machine was supposed to be here last year. Has anyone seen either? HP and HPE have twice the number of employees to revenue as Cisco, Dell, Lenovo or IBM. When I see HPE I see compost.