Can AIOps get us to better ITSM? The impact of visibility for your mission critical IT systems

Profile picture for user Crystal Miceli By Crystal Miceli February 3, 2021 Audio mode
Summary:
How do you mitigate IT systems risk, when budgets are stretched thin? And is AIOps a true asset to better ITSM and systems visibility - or are we getting ahead of ourselves? Crystal Miceli of Cherwell shares her lessons learned.

Cloud question mark © Thuat Images - Fotolia.com

About 20 years ago, I implemented an IT asset discovery tool for my organization. The software found something that took my breath away. On the same network responsible for supporting surgery, medical records and lab servers for a busy hospital, was a beast of a gaming computer consuming copious network resources.

We found the machine in a closet - and the errant employee at his desk - and quickly eliminated the issue. 

Today, every business-critical application is supported by a complex fabric of servers, network devices and services - both physical and virtual - local and in the cloud. The impact of an outage or vulnerability for any one of those components can be extreme. It's of the utmost importance that IT is aware of and monitors not just every device, but also the dependencies between them. 

Fast forward 15 years and at a different company, I found out through a casual conversation that there was a single server responsible for processing all of the audit logs for hundreds of customers - sitting under the vacant desk of a prior employee. This situation put customer data and the company's reputation at risk. Mind blown.

These are true, but extreme examples of what can happen when you don't have visibility over the IT infrastructure that supports your mission critical systems. As the architectures behind applications become even more complicated with the cloud, virtualization and shared services, manual documentation of dependencies is no longer a feasible option.

Understanding all of the pieces that make an application available is about more than knowing the up or down status. You need to know that old server that you thought wasn't in use is actually a key contributor to the availability of your ecommerce site - and that a simple reboot could actually impact sales and your bottom line. With security vulnerabilities putting customer data at risk, having awareness of all of the priority systems that need to be secured is also vital. 

So how do you mitigate these risks proactively at a time when IT resources are already stretched thin? I'd like to be able to say there's a silver bullet, a magic application that solves all of these problems. Unfortunately, that's not reality. As with most IT efforts, a combination of people, process and technology will be required.

The good news is that the solutions available to automatically discover your systems and map dependencies are much more sophisticated today than they were when I found that gaming machine in a closet. These lightweight DDM tools can handle the complexity of distributed hybrid environments - giving you the opportunity to visualize and manage the components of your e-commerce, supply chain, ERP and other critical applications. 

DDM and AIOps tools - an asset to system visibility, or overhyped?

You likely already have some degree of automated discovery implemented, but it's worth looking at the innovative DDM and AIOps solutions on the market that extend the capabilities to fully map your business-critical applications to the underlying hybrid architectures that make them accessible. The best solutions on the market allow you to discover and map services including components such as web servers, application services, APIs and physical/virtual/cloud infrastructure.

They depict the interdependencies in a graphical format, so IT teams can readily see the connections between assets, and which services they support - critical to troubleshooting incidents or preventing impact of change. Some leading solutions even offer real-world, practical use of artificial intelligence with predictive analytics, anomaly detection and self-healing capabilities. 

After years of hearing AI touted as the solution to all IT problems, you'd be wise to be skeptical of vendor hype. Success with AITSM, DDM and AIOps solutions requires a practical implementation plan that includes process change, especially with regards to management of your CMDB and change and release management processes. Adoption of these process changes is necessary to reap the benefits of the technology, so you have to plan for governance over process execution by the teams involved.  

How to get to better ITSM

What does "nirvana" look like? Who knows? But maybe your IT reality could look a little more like this: 

  • A well-meaning IT team member initiates a change request to decommission an outdated virtual machine that supported a deprecated application. Impact analysis indicates that the server actually still supports a live mission-critical service. Instead of being shut down, a plan is put in place to bring it up to date on the lasted security patches so no vulnerabilities are introduced that could result in business impact.
  • Your team receives an alert about performance issues on your ecommerce site during a busy shopping period. Before the team can react, you get another alert that capacity has automatically been added to one of the web servers supporting that site by the self-healing feature of your AIOps solution. A ticket was opened and closed without intervention. 

Thankfully, both the situations I described earlier in this article were managed before the worst could happen. We know from our own experiences and those unfortunate incidents that make the news that not every IT organization is so fortunate. Learn more about how you can protect your mission critical applications with DDM and AIOps technologies at our webinar on March 4, 2021.