Mar
23

Data Center Automation

Data Center Automation

March 23
By

Bill Gates recently caused a stir when he said that robots that replace humans in the workforce should be taxed. Gates explained that because we tax employee’s earning, we should also tax robots that reduce costs. Gates believes that such a tax may be temporary or just long enough for people to gain new skills. EU lawmakers proposed a similar tax on robot owners to pay for the training of employees who lost their jobs.

Whether or not you agree with Gates on this issue, he shines a light on an increasingly important issue: How do we adapt to automation? This is a tough question to answer because it affects people in many different ways.

human playing chess with a machine - data center automation

Gates reminded me of a friend I met in Germany many years ago. He welded sections of the chassis as they moved down the assembly line at the Audi auto plant. Over time, robots took over his job, and his boss him assigned to the warehouse. As part of his new job, he organized, prepared and inventoried thousands of different parts. The work wasn’t as interesting as what he used to do on the assembly line. Over time, automation found its way into the warehouse. My friend began working alongside robots that could organize, sort and retrieve parts at a much faster pace than he could. Robots always show up ready to work. They are never late or call in sick. A robot is the perfect employee in many ways.

Human Involvement

Two weeks ago, Amazon S3 service suffered a major outage at its North Virginia data center. The outage took down products like Slack and Trello which affected millions of users. Once Amazon restored their service, they admitted that an engineer caused the outage. An Amazon employee executed a simple command in order to take a handful of servers offline for maintenance. But he inadvertently removed a much larger set of servers when he committed a typo in the command.

The small typo had major implications. This is one reason automation is so attractive to business owners. Amazon brass has promised they have patched their system so a similar issue could not happen. I suspect that some of those patches rely on automating the process and taking it out of the hands of the engineer. The risks are too great and the losses too substantial for businesses to ignore automation.

Unless you’re running a lights out data center, humans will make mistakes. But as Amazon learned, it helps to have a safely net. Automating repetitive tasks makes a lot of sense. That frees your employees to decisions that require nuance. As AI improves and computers are able to learn from their mistakes, they will take on larger roles. Who knows where this will end, but two years ago a V.C. firm named a robot to their board of directors.

Along this same vein, I’d like to take a look at a topic that’s closer to home for us in the computer industry: data center automation. As more companies move their products and services to the cloud, the more cloud service providers will focus on reliability and business continuity. Even a small outage can cost businesses hundreds of thousands of dollars if not more. With that in mind, let’s take a look at some of the most popular automation products found in today’s data center.

Open Stack

Last month I wrote about the rise of OpenStack and the role it plays in cloud computing. But it will also bring automation opportunities to the data center. OpenStack software controls large pools of computer, storage and networking resources through a dashboard. It helps companies manage local resources as though they were in the cloud. That opens opportunity for companies to automate the building, taking down and management of virtual resources such as virtual machines and storage.

OpenStack still has a ways to go in terms of implementation. But it has support from the largest cloud service providers, and they will be looking for efficiencies to set their services apart. The goal is to make moving to the cloud as easy as possible, and automation along with OpenStack will play a big part.

Microsoft System Center

Microsoft System Center (MSC) is management and automation system that integrates Microsoft Windows and Linux servers, cloud and on-premise infrastructure as well as variety of compute, security and networking components. In short, it provides the administrator a number of tools that allow him to automate many operations on a server platform. MSC includes the following components:

  • VM Manager – Provisions, deploys and manages VMs.
  • Configuration Manager – Configures, patches and updates clients and servers.
  • Ops Manager – Monitors virtual and cloud infrastructure and workloads.
  • Service Manager – IT process management and self-service tools.
  • Data Protection Management – Backup private clouds, physical clients and servers.

With MSC 2016, Microsoft is positioning itself as a major player in data center management and automation arena. In addition to deployment tools that manage Windows Servers, it also provides mature tools for monitoring and managing heterogeneous and open systems that include Hyper-V and VMWare. Microsoft promises that by implementing MSC companies will free up IT managers to tackle more complex problems. It’s a lofty promise, and one that many might question a decade ago when Microsoft made the bulk of its profits selling Windows and Office products. But with a new CEO in Satya Nadella acting as chief cloud advocate at the world’s largest software company, one must take Microsoft seriously. Nobody knows this better than Amazon.

Eucalyptus

With the inclusion of MSC, it seems only fair to include a data center automation tool that bridges Amazon Web Service (AWS) across any on-premise infrastructure. Eucalyptus was first born as an open grid project at Rice University. Today it’s an open source tool that treats every resource it manages as a cloud instance whether it resides in the cloud, on-premise or in AWS.

While it’s clearly marketed at AWS customers, Eucalyptus was purchased by HP in late 2014 and was renamed to HPE Helion Eucalyptus. It has six key component that include:

  • Cloud Controller – Basically the dashboard for all high level resource management and scheduling.
  • Walrus – The Eucalyptus equivalent AWS Simple Storage Service. It offers persistent storage to all the VMs.
  • Cluster Controller – Front end interface for cluster sets within Eucalyptus.
  • Storage Controller – Communicates with the cluster controller to manage Eucalyptus block volumes and snapshots.

HP has improved the level of support and documentation for Eucalyptus which has given it increased credibility. With the torrid growth of AWS, you can expect it to flourish for many years to come.

Conclusion

We are witnessing the convergence of a number of technologies that bring automation into our lives. These and other automation tools make it easier for companies to manage their resource and infrastructure in the cloud. Do these tools allow the data center of the future to operate with much human intervention? And what does that mean for those employees managing data centers today? We are approaching the point in time where computers are able to not only organize information, but make complex decisions based on the data we give them.

This is great news for those with advanced skills to program the computers and robots. But if you’re one of those people like my friend who worked for Audi, your job may be replaced by advanced automation. We need to discuss these and similar questions because the answers have important ramifications in terms of employment and education. Advances in technology are leaving many people behind. We’ve seen cities like San Francisco and Seattle produce booming economies while soaring housing prices in those areas leave all by the most privileged behind.

We still have a lot of work ahead of us. Automation is here to stay, but we need to consider its effects on not just business but society.