Red Hat Summit + AnsibleFest 2025 | Chris Farrell, IBM

Clips
More from Red Hat Summit + AnsibleFest 2025

Chris Farrell

Group Product Manager Instana Oberservability

IBM

play_circle_outline Unlocking Efficiency: AI Automation in Modern Application Architecture and Observability for Management

play_circle_outline Generative AI for code generation and automation

play_circle_outline Enhancing Operations: Ansible's Power Expanded Through IBM and Red Hat Collaboration

play_circle_outline Revolutionizing Observability and AI in Application Development: The Future of Automation

Info
Transcript

Chris Farrell, IBM

Chris Farrell

Group Product Manager Instana Oberservability IBM

Chris Farrell, group product manager of Instana observability at IBM Corp., joins theCUBE’s Rebecca Knight and Rob Strechay during Red Hat Summit 2025 to explore how AI and automation are shaping the future of application management. The conversation examines IBM’s integration of observability with technologies such as watsonx to accelerate detection and resolution.

Farrell highlights the role of AI-driven automation in streamlining performance and reducing manual effort across complex environments. He also shares how IBM’s work with Red Hat and Ansibl... Read more

explore Keep Exploring

What is the importance of incorporating AI into observability and cost monitoring of applications? add

What role do you see generative AI, like Code Assist and watsonx Code Assist, playing in the automation and generation of code? add

What enhancements have been made to the integration of Instana with Ansible and watsonx, and what future plans involve incorporating Turbonomic into the mix for automation and resource planning? add

What are some of the new features and integrations that have been released in the observability tooling, and how do they aim to improve the efficiency and effectiveness of the application operations team? add

bolt Powered by CUBE AI

Chris Farrell, IBM

search

Rebecca Knight

>> Good afternoon, everyone, and welcome back to theCUBE's Live coverage of Red Hat Summit AnsibleFest 2025. We are in the home stretch, this is our final interview of the day before we get to our analyst angle. I'm your host, Rebecca Knight, alongside Rob Strechay, who's been my esteemed co-host and analyst for the show. It's been a joy working with you, and we're about to talk about your favorite topic.

Rob Strechay

>> One of my favorite topics.

Rebecca Knight

>> One of your favorites.

Rob Strechay

>> Definitely. I am partial to-

Rebecca Knight

>> Yeah. Beer, cars, observability.

Rob Strechay

>> Yes.

Rebecca Knight

>> Yeah.

Rob Strechay

>> Maybe not in that order. Sometimes, it depends on where we are in the day, but we're getting close to that being the right order.

Rebecca Knight

>> Exactly, okay. And with that, I would like to welcome our next guest, Chris Farrell, group product manager, Instana observability at IBM. Thank you so much for coming on theCUBE, Chris-

Chris Farrell

>> Rebecca, glad to be here....

Rebecca Knight

>> and for being our final interview of the day.

Chris Farrell

>> I didn't know I was going to be the last one in the day, yeah.

Rebecca Knight

>> We saved you for last.

Chris Farrell

>> So excited about that.

Rebecca Knight

>> Yeah, so-

Chris Farrell

>> And I'm also interested in the beer and the cars. Yes, like I said, observability first, then beer, then cars-

Rob Strechay

>> Yes, -

Chris Farrell

>> But not together.

Rob Strechay

>> Not together.

Chris Farrell

>> Yeah, exactly.

Rebecca Knight

>> So automation is changing everything. How, Chris, do you think, how does IBM think and how do you personally think about automation in the context of AI and modern application architecture?

Chris Farrell

>> . That's a great question. You really have to separate it into two different things. First is what does automation mean for applications? And then, what does automation or what does AI mean for applications, and what does AI mean for managing those applications, and how do you manage it? And we're actually focused on both those things at the same time, simultaneously. So one of the things that we're doing is putting AI into the observability aspect of managing the applications. We have recently released integration with watsonx, to create summarizations of problems in plain English, so that anyone can get a summarization and print it out. We've also put watsonx integration in to start making decisions on how to remediate incidents. But at the same time, it would be silly to think about observing applications without also thinking about observing the AI aspect of those applications. So we have built the ability to measure and monitor performance of AI requests. We've also put in there the ability to understand what it's costing. Now, we haven't made cost reports just yet, but we're actually tracking usage stats that, of course, at the end of the day, once we all really start using AI more and more, the AI providers are probably going to start charging us more and more. So it'll be an important aspect of an application, to understand what your operational cost of doing that is.

Rob Strechay

>> Well, the good thing is you have watsonx, which is your own, which is IBM. So you have your internal chargeback, but that's besides that the .

Chris Farrell

>> .

Rob Strechay

>> What role do you see generative AI, like Code Assist, watsonx Code Assist and things of that nature really playing from automation and automation of code generation?

Chris Farrell

>> I mentioned the fact that we're starting to see actions taken.

Rob Strechay

>> Yeah.

Chris Farrell

>> And watsonx and Code Assistant as part of it make a big part of what we're doing. So we started with just using watsonx to curate some things for us upfront. So last quarter, we released some curated actions that are built into the product. So they're built by AI, but they're available for people to look at and take advantage of. But last month, we released a new version of that, where you can actually have watsonx generate actions for you, for something that you might not have seen before and that we haven't necessarily curated for you. And that way, you can start taking advantage of what we've learned and know from previous incidents and actions that we've taken. And we just announced three weeks ago, we just previewed, it's not available yet, we previewed an agentic version of that, where watsonx will start to generate these actions, and then will start using agentic AI to start understanding which ones are best to run and whether or not they worked.

Rob Strechay

>> .

Rebecca Knight

>> So Ansible is a core tool in many enterprise automation stacks. How is IBM working with Red Hat to extend, enhance Ansible's power?

Chris Farrell

>> In multiple ways. One of the great things that we've done with Instana was we added an Ansible action catalog into our products, so that users of Instana observability could actually start using Ansible actions to remediate issues and get things solved. But what we've done with the integration with watsonx is start to make that more automatic, take advantage of Ansible's event-driven features, and now we're looking forward to adding Turbonomic to the mix, so that you can actually get a little bit... It's not necessarily predictive, but we're thinking of it as predictive. Because when you add Turbonomic's ability to understand what resources are going to be necessary to operate the level of performance you want, and you throw Ansible actions into that mix, you can now start to automate the actions you need, to always make sure everything is available that you need to run your applications.

Rob Strechay

>> So you just brought up Turbonomic, and it kind of made me think about the fact that one of the big things we've been talking about all this week has been OpenShift Virtualization. And with the changing of that, and people trying to figure out what are the right resources as they move, how does Instana and Turbonomic really help these organizations figure out what to do with that? Because if they're going from a previous hypervisor to now OpenShift Virt, they need to understand what it's going to be, the performance, and the right resources, and things of that nature.

Chris Farrell

>> It's actually all the things, a lot of the reasons that people would look to move to OpenShift Virt are similar to reasons why you want Turbonomic and Instana in the mix. As you're moving, you want to take advantage of the new technologies, you want to perform, you'd like to perform better than you did before. You'd like to perform at a lower cost than you did before. Well, how are you going to verify that? How are you going to make sure of that? One thing is to have tools that operate in all the environments the same way, which again, a nice benefit of OpenShift, that's a benefit of Instana is that as you migrate, whether from some legacy systems to a cloud, or one hypervisor to another, or to a multi-cloud environment, Instana is always monitoring, measuring the same things, the same way, so you get an apples to apples comparison. So you get to benchmark, and you can monitor as you move, and you can monitor after it's done, and you get to really verify that what you've done did not impact your performance, did not make the application worse, hopefully made it better. And Turbonomic has the same value, in that you can take a baseline understanding of performance, and cost, and resource usage. And now, as you go through this transition and migration, Turbo's there to make sure that there's nothing incorrectly provisioned that might prevent you from reaching the full benefit of what you're trying to do.

Rebecca Knight

>> So break it down, what success looks like, because as you said, there are so many different challenges, with costs, and inefficiencies, and complexity. What does success look like for these organizations that adopt AI-powered automation?

Chris Farrell

>> I think one of the things that happens is that we start to move the tasks that we do into more valuable situations. This is actually a benefit of automated observability to begin with, before you apply AI. But once you apply AI, it takes it even better. One of the problems with old-school monitoring tools is that you have to manually configure a lot of things to get the full value of what you hoped when you got the tool. Everything works a little bit out of the box, but to get the real value, old-school tools required constant manual configuration. And one of the things that Instana was focused on doing was to create automation across that entire monitoring lifecycle. So not only did you get this benefit upfront of being able to deploy quickly, but as changes occur, and let's face it, with new, modern applications, changes occur more often, and they're usually pretty volatile, or they can be, it's important to also then, also not require manual configuration. Well, AI only makes that one step better, because if you think about the trust, once people trust the actions that are there, AI can come in, start to make those decisions that says, "Based on this incident that's going on, this is the right action to take, and I have a score, and it's been checked," and they trust that it is. So you start to automate that. And now, I'm no longer worried about quote, "smaller", unquote actions of cycling servers or decommissioning something because I don't need it anymore. Rather, now I can focus on, "How do we engineer this backend system to operate a little bit better?"

Rob Strechay

>> When you look at the landscape of this approach, a lot of companies are embracing multi-cloud. We kind of talk about it. It's cloud is not a place, it's an operating model. So you have your on-premise cloud, you have your cloud in somebody else's data center, hyperscaler, you have it in one of the big co-los, and things of that nature. How do you look at this, and how is IBM helping facilitate the automation across this, with this tooling and the observability aspects of it? Because I think, again, sometimes, to your point on metrics, and KPIs, and parameters, and understanding, they're not the same everywhere. They're in different details, they're in different logs, they're in different trace files and things of that nature. How do you look at that?

Chris Farrell

>> One of the things you have to do is kind of ground everyone on a common language, right? And to your point, if I'm using different data sources from different clouds and different monitoring tools from different clouds, I'm creating that siloed environment that we're trying to eliminate in the first place. I'm just creating it at a different spot. So it's really important, as you think about, no matter what you do to operate, whether all on-prem, a hybrid environment, a multi-environment, or all on cloud somewhere, or on a private cloud hosted somewhere, it's important to be able to ground everyone in the same types of understanding of what you're measuring, how you're measuring it. And I'm going to go back to OpenShift, different OpenShift, regular OpenShift, Kubernetes distribution. One of the reasons I love working with Kubernetes OpenShift distribution is because I can run the same thing everywhere, but I don't have to translate from one Kubernetes distribution to another as I go multi-cloud. And the power of Instana, and Turbo, and really all IBM automation products is that we operate the same across all. So you have this grounded messaging, grounded monitoring, and grounded kind of a language that everyone can speak as you move across. Otherwise, everyone walks into the ops meeting with their own tool and says, "Look, all my lights are green, but yet our users are experiencing yellow or red conditions."

Rob Strechay

>> Right.

Rebecca Knight

>> So a lot of the organizations you work with operate in highly-regulated environments, and I'm thinking telecoms, I'm thinking financial services. How do you help those organizations? How does IBM think about helping them balance the need for automation, the desire to automate certain kinds of tasks, with governance, and compliance, and risk management?

Chris Farrell

>> That really comes in a couple of ways. First of all, a lot of the regulated clients and customers around the world, they have their own service levels that they have to meet regulations on, or it costs them additional money. But the reality is that the value of an application is well beyond anything just fines are now, right? And especially in the industries you just mentioned, those are classic APM buyers. But what they used to do would be buy an APM solution and put it on one application or two. The applications that were regulated, those were the ones they had to make sure. But most organizations have dozens of applications they're monitoring now, they have hundreds of applications that they should be monitoring, and they're having to draw the line somewhere. And that's where automation helps, because they're drawing the line for two reasons. One, probably license costs, but observability is helping with that. But what observability is not necessarily helping with is the amount of effort it takes to roll out monitoring across dozens or hundreds of applications. And that's where IBM Automation is focused on, is helping you get what you need on any application. Because at the end of the day, every application is important to somebody. They developed it, they put investment in to create it. So everyone, there's people out there that want and need observability there, and the automation tools allow that to happen.

Rob Strechay

>> So what are some of the patterns or best practices around that automation that IBM has seen and had success, that helps people get that long-term success?

Chris Farrell

>> The first thing that I see is that once you have that performance monitoring in place, adding the optimization with Turbonomic on top and integrating them together, so that you're assuring application performance while also assuring that you don't over-provision, I think that's a best practice that I'd like to see more often. But it's a really important one, especially as we start to move into AI workloads, because now we're going to not only be provisioning a lot of on-prem and cloud resources for our applications, but also for these calls that we're going to make, and you start to implement GPUs across the board. Understanding how that comes into play is a really big part of that. So you can't really monitor applications even in their own silo anymore. You have to have an ear, an eye on infrastructure and an eye on your GPUs. And I think as we move forward, the other best practice that I see is the embracing of the automated actions, trusting the tools as they make recommendations to go forward. Now, you want to test them yourself first, but once you have that trust, turning them loose so that you can focus on the harder things that are going to come down the pipe with integrated AI, hybrid cloud, multi-cloud solutions, I think those are the two places where I see companies succeeding the most.

Rebecca Knight

>> Excellent. Last question, what are you most excited about right now? What are we going to be talking about next year at Red Hat Summit 2026 or even IBM Think?

Chris Farrell

>> so for me personally, and within the observability tooling itself, what we're working on right now, what we've just released with our AI tooling and where we're going next, we've just put in a whole bunch of integrations with the other automation products. So we have vulnerabilities brought in by Concert, we have optimization ideas brought in by Turbo, we have costing information coming in from Qcost, and now we're starting to look at how we bring all that together with things like the AI token usage, and our probable root cause determination, which is powered by watsonx, and these automatic actions. And when you start to put all those things together, you start to create this hub where the application operations team now can truly relax, to a point, and not be totally stressed all the time. And I think where we're going, what excites me is seeing where AI will actually come in and enhance those parts of the systems, so that the application development and deployment teams can truly take what they're building with AI to another level.

Rebecca Knight

>> I like the vision. Chris, thank you so much. It's a pleasure having you on the show.

Chris Farrell

>> Thanks, guys. I appreciate it.

Rebecca Knight

>> That is a wrap for Red Hat Summit 2026 AnsibleFest. I'm Rebecca Knight for Rob Strechay. We still have one more segment, one analyst angle coming up, so keep it right here. You're watching theCUBE, the leader in enterprise tech news and analysis.