Get the most out of data center automation tools

Siloed and inconsistent tool sets can complicate infrastructure and operation automation. Heed this advice to get the most out of automation tools.

This Content Component encountered an error

ORLANDO, Fla. -- While organizations rely on automation tools to manage workloads, the truth is these tools can also pose serious challenges.  

Automation tools can manage workloads between increasingly diverse environments like local data centers, remote or outsourced locations and cloud providers. But there are often too many tools to handle, which can overcrowd an enterprise and lead to errors.

Adopting some best-practices can smooth the use of infrastructure and operations (I&O) tools and position your organization to integrate future automation platforms.

I&O automation tool challenges

Perhaps the most obvious problem with I&O automation tools is that there are simply too many of them at work in the enterprise. It's easy to understand why; most organizations trace them back to deeply-rooted silos that are still in place today. Each team brings in tools to meet the needs of their unique silo, such as server automation, scripting and job scheduling.

"That's how the tools have emerged and evolved," said Ronni Colville, an analyst with Gartner Inc., during a session here at this week's Gartner IT Infrastructure and Operations Management Summit 2014.

To exacerbate the problem, there is little inter-silo communication or collaboration. This prevents organizations from planning long-term or overarching automation strategies. The vendors haven't helped, happily providing purpose-built tools that solve specific problems -- but are rarely extensible or interoperable outside of their own product family. 

For example, server automation typically involves initial provisioning, discovery and inventory, software deployment and patching, configuration and auditing, and even broader lifecycle management. Each of these steps may rely on a different tool or module -- sometimes from different vendors.

Just consider a Windows-based organization that adds System Center 2012 R2 which includes individual components like App Controller, Configuration Manager, Data Protection Manager, Orchestrator, Operations Manager, Service Manager, and Virtual Machine Manager within the same platform.

Future of Automation tools

By 2017, 75% of large enterprises will have more than four diverse automation technologies within their IT management portfolio; up from 20% in 2014.
-- Gartner

By 2017, 75% of large enterprises will have more than four diverse automation technologies within their IT management portfolio; up from 20% in 2014, Colville said.

Another challenge is that the fundamental design of I&O automation tools is based on workflows, which expect certain results produced from known inputs or steps; extensions of scripting. When one input or step is missing or unexpected, the tool does not produce the desired result, so errors occur and manual interaction is required.

"We're not using a lot of automation besides scripts," said Eric Devaul, manager of Windows Server and security teams for a global glass manufacturer. "We're looking to see what [I&O automation] will do."

To move past tightly-defined script-type behavior, more sophisticated tools will be needed in the future.

Make the most of automation tools

For the near term, organizations must streamline I&O automation strategies. It's a major priority; polling during the session revealed that 55% of respondents plan to add I&O automation tools within the next 12 to 18 months.

"We're looking for opportunities to automate what we do manually now," said Kevin Weaver, director of infrastructure at David's Bridal, based in Conshohocken, Pennsylvania. "Everybody starts small and gradually builds repeatable processes."

Colville recommends a set of best-practices that start with establishing and filling new automation roles within the IT staff.

These roles include automation manager, automation architect, system administrator, or automation specialist. The goal is to get an IT professional thinking about what needs to be automated -- and how. This helps ensure someone can take overarching responsibility for the tool selection and integration.

Next, conduct an inventory and discover what needs to be automated. In most cases, this involves looking for scripts, and a typical organization may have scripts under Windows, Unix, PHP, Ruby on Rails, file transfer scripts, command like scripts and automation scripts under countless other platforms. Once the scripts are identified, it's easier to implement uniform version control and strategize tool changes that might allow some tools to be extended into other areas while less-flexible tools might be retired.

Look for mature tools that offer value. Immature tools that cater to a relatively new automation goal will undergo more change and cost considerably more. For example, Colville said scripts tend to be immature with little value, while business process management tools are very mature and can offer significant value.

However, tools can complement or compete with one another. Some features may overlap, while some technologies may disrupt each other unintentionally. The goal, Colville said, is to carefully evaluate each tool's features and determine what it can tell you right out of the box, and then note its integration to see how it bridges or connects to other tools.

Weigh the service that you'll get from each tool, and then study its workflow. This can help IT leaders choose the best tools for the current environment and business needs.

Future I&O tool trends

There are two developments in the I&O automation tool market worth watching, Colville said. First, it will be practically impossible to replace disparate tools with a single ubiquitous automation tool -- at least in any foreseeable timeframe. Instead, look for the emergence of higher-level products that can tie together and interoperate with existing diverse products; a "manager of managers."

But maybe the more interesting prediction is the eventual move to heuristic automation which relies on knowledge and analytics to drive the automation process instead of traditional deterministic tools. Heuristic tools can potentially act with more autonomy in response to available or dynamic resources and prevailing conditions within the environment without a blind reliance on pre-defined steps.

This was first published in June 2014

Dig deeper on IT process automation, data center automation

Pro+

Features

Enjoy the benefits of Pro+ membership, learn more and join.

1 comment

Oldest 

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to:

-ADS BY GOOGLE

SearchWindowsServer

SearchEnterpriseLinux

SearchServerVirtualization

SearchCloudComputing

Close