Sorry, you need to enable JavaScript to visit this website.
Home / OpenStack* on Intel® Architecture / Blogs / 2014 / OpenStack* on Intel® Architecture - DevOps: Enterprise Datacenters Can Adopt DevOps or Gradually Lose Their Workloads to Cloud Service Providers

DevOps: Enterprise Datacenters Can Adopt DevOps or Gradually Lose Their Workloads to Cloud Service Providers

01 Staff
DevOps:  Enterprise Datacenters Can Adopt DevOps or Gradually Lose Their Workloads to Cloud Service Providers

The IT Ops mind-set tends to revere “reliable/repeatable processes” while tenaciously holding on to the tried-and-true.  There’s a good reason for that:  the enterprise runs on its IT applications and infrastructure; if their operation gets disrupted, the business could grind to a halt, and everyone could lose their jobs. 

So new services, patches, and bug fixes don’t get deployed until a complex set of release criteria are signed off by the affected stakeholders.  But the back-and-forth associated with that step alone can take weeks. 

That’s not acceptable to Web companies such as Facebook, Google, Yahoo!, and Baidu who know their users can vote with their feet.  Indeed, in what many consider a defining moment in the history of the DevOps movement five years ago, Flickr famously claimed in the June 2009 O’Reilly Velocity Conference that they routinely did 10+ deploys a day

But Web companies are not alone is seeking the competitive advantage of responding more rapidly to customer needs.  And failure to deploy services rapidly enough can have devastating consequences to any business.  For example, a retailer launching a key “Back to School” promotion might need a just-in-time roll out of their revamped shopping Website along with an updated Sales & Distribution application.  If release to production is delayed by even a day, it would mean lost revenues and lost market share that might never be recovered.

What’s the answer?  According to Gene Kim, IT organizations should take a DevOps approach to dramatically reduce cycle time, just as leading manufacturers practice design for manufacturability to dramatically improve time-to-market.   Even IT stalwarts like IBM today tout the benefits of adopting the DevOps approach.

DevOps:  a mash-up of application development and IT operations.  Image courtesy of IBM


It sounds simple:  Development and Operations work together with their business sponsors to plan and manage a holisticdevelopment and operations lifecycle.  But challenges abound: 

  • Enterprise IT organizations often resist change strenuously due to their Ops-heavy, conservative points of view.  In addition, the Operations team will lose at least some of its dominance, so political maneuvering can rear its ugly head.
  • Development and Operations lifecycles are complex, and orchestrating them into one composable set of interrelated processes is not something either team is likely to embrace, right off the bat.

So, top-down executive support is an essential ingredient for success.  The CEO, CFO, and CIO should let it be known that greater IT agility is paramount to the business’s success.  They should create an empowered, decision-making DevOps team comprised of key stakeholders from Development, Operations, and sponsoring lines of business.  That team should define and manage the new IT lifecycle.

Another key to success is to orchestrate the applications and operations lifecycles into a composable set of automated processes.  Fortunately, leading vendors such as Puppet LabsCloudifySerena Software, and others have risen to the occasion.

The question is, will enterprise IT shops embrace the DevOps approach in time?  Savvy business decision makers see the competitive business advantage to deploying services in minutes rather than months.  But if their IT organizations don’t adopt DevOps soon enough, business unit sponsors can easily find cloud service providers that already have done so.  Is that where most enterprise workloads are headed?  What do you think?  


Further Reading

01 Staff

This coming week, our team will share how Intel is helping address the requirements demanded by data-centric, compute-intensive workloads quickly growing across data center and edge.

Manjeet Bhatia

Devops principles, like continuous integration (CI) and continuous delivery (CD), are attracting considerable attention given their propensity to increase software development efficiency and facili