IT staffs spend an inordinate amount of time and effort configuring the stacks of software to run applications. But if anything was made clear at the OpenStack Summit 2015 conference last week, it’s that much of the work associated with deploying tools like Hadoop is about to become automated.
Case in point is Piston Computing, which used the conference to showcase a turnkey platform for deploying a private cloud. Via the click of a single button, IT administrators can use Piston CloudOS 4.0 to not only launch OpenStack in minutes, but also to deploy Hadoop and Spark on bare metal servers. In the near future, Piston Computing plans to add support for container orchestration tools such as Kubernetes, Mesos and Docker Swarm.
OpenStack has a reputation for being difficult to deploy. But Steve Bagby, chief operating officer for Piston Computing, says one of the things that will differentiate IT vendors in a world where most of them are embracing OpenStack is the degree to which they can automate the deployment processes associated with creating a private cloud. Today, that’s mainly an exercise in manual labor for most IT organizations. But as private cloud computing platforms mature, it’s apparent that the expertise required to deploy a private cloud is going to be considerably less hands-on.
The timing of these advances couldn’t be better. Most IT organizations today are under severe pressure to be more agile, which is something that is very difficult to do if you’re spending the majority of your time manually configuring complex pieces of software that can quite easily be misconfigured.
Fortunately, advances in IT automation are about to make the whole software deployment process a lot less painful. Also, the private cloud computing environment is about to become a major battleground between IT vendors that have historically counted on the patience of IT administrators willing to persevere no matter how complicated things get.