Get the most from your IT Ops analytics tools

There are over 50 tools on the market for IT operations analytics, and sales are on the rise, according to a report from Research and Markets. However, companies don't always fully utilize the tools to most effectively help them collect and analyze the data they need for smooth IT operations.

If you're an IT operations manager or administrator who already has IT Ops analytics tools—or who is about to acquire them—here are some best practices you should follow to ensure that you're getting the full value from your investment.

The State of Analytics in IT Operations

Embed IT Ops analytics in your process

Ensuring proper adoption is key, said Torrey Jones, principal consultant of the GreenLight Group.

"All too often, we see these tools get installed, and the executive management advertises that they want people to use them, but they're never truly operationalized by the organization to put them into the day-to-day practice," he said.

IT Ops analytics tools should be embedded in the standard operating procedures of the company's Tier 1 NOC (network operations center). That's where the processes start, said Jones, and that's where organizations will see the most usage. Furthermore, adoption at that level will drive usage through the Tier 2 and Tier 3 deeper-level support teams.

The more data sources, the better

Data integration is also critical. "Obviously, make sure that there's data going in there," he said. "But in the world of big data analytics, you don't really know what the right data is until you need it. So integrating with as many data sources as possible is key." 

IDC analyst Stephen Elliot agreed that the most important best practice is recognizing the tremendous amount of data that exists from any number of data sources, including IT, human resources, external partners, blogs, applications, and hardware.

"You have to have a sense of what matters to your objective," he said. "So you have to recognize what data you need or don't need for what you're going to measure."

Look at the bigger IT Ops picture

David Levy, vice president of marketing at J9 Technologies, said that organizations typically implement numerous monitoring tools to keep track of what's going on in their environments. These can include tools that look at bandwidth utilization, that look at devices such as servers and storage, and that look at applications that are critical to the success of the enterprise.  

But organizations often don't get the full value from those tools because they manage each tool in a separate silo. As a result they miss out on a lot of potential correlations and insights into the way their environments behave, he said.

"That's why we always recommend looking at your environment as one integrated whole, and having a tool to do correlations between the different events and metrics that are coming off your different monitoring tools," Levy explained. "And that will give you the insight into where the root cause of a problem is."

Give users what they need

It's important to have dashboards that are relevant to the use cases, according to GreenLight's Jones. For instance, system administrators are going to care about system and host metrics, but application administrators will be concerned with application performance metrics and related information.

"So be sure that you have the right dashboards for the personas" who are going to be using the tools, he said.

There also has to be someone on staff responsible for maintaining the analytics tools, according to Jones. "Sometimes people go out and buy a Ferrari engine but they don’t hire a Ferrari mechanic," he said.

There has to be someone to ensure that the correct data is flowing, that users are getting their questions answered, and that the system is online and available. Otherwise, IT Ops analytics is "not going to get adopted," Jones said. This role is most likely an application admin type of person instead of a data admin role, he said. 

IT Ops analytics should change everything

Roy Illsley, a principal analyst at Ovum, said that an IT Ops analytics tool is only as good as the processes a company applies to it and the people who use it. "It changes the role of the job, and it changes what you're doing and how you're doing it," he said.

Consequently, IT operations managers must change the processes that they use to interact with the business as well as with DevOps—and everybody else—to get the maximum value out of their IT operational tools.

"So, whatever tool you choose, talk to your development team and make sure that they're comfortable with either getting the output from it, so they can do their stuff, or they can have access to it and see the bits they need," especially related to testing, release management, and faults, Illsley explained. "That way you're getting the collaboration, you're getting the DevOps message out there, and that's where the big value will come from."

It's not enough for an analytics tool to sound an alert when something is about to fail in six hours. Rather, IT Ops managers must have a process in place that will enable them to identify the same error in every other instance where it may be happening and fix it, Illsley said.

However, as IT operations managers and administrators get comfortable with this manual intervention, they may want to evolve those processes to do things automatically, saidIllsley.

"The best practice is to look for tools that can actually be extended to do the automation," he said. Many IT Ops analytics tools are "very good at telling you something, but they don't necessarily integrate with automation tools, such as Chef or Puppet, that can do the actual work."

Additionally, many IT operations analytics tools are geared toward specific verticals. Companies in the healthcare industry, for instance, should look for tools that understand the regulations and nuances that pertain to healthcare, Illsley said.

Get bang for your buck

To get the most out of IT operations analytics tools, you need to have a good sense of what you want to achieve with the tools, ensure that the tools are actually adopted in a meaningful way, integrate the correct data from various data sources into your core processes, and make certain that you have the right staff in place to manage everything.

[ Upcoming Webinar (Oct. 23): Simplify Discovery and Change Management for Cloud and Container Environments ]