“It’s important to note that it’s not automation just for the sake of automation.”
IBM’s recent acquisition of Instana brings automated application performance monitoring (APM) and observability capabilities to IBM Cloud Pak® for Watson AIOps. Instana goes beyond traditional APM, bringing enterprise observability that delivers automated visibility, actionable information and understanding to DevOps — helping users ensure the performance of cloud native applications.
Let’s explore how Instana fits into the IBM Automation portfolio and what the value is to organizations looking to drive intelligent IT Operations through automation and artificial intelligence (AI). Mike Mallo, Offering Management Lead, Hybrid Cloud DevOps at IBM and Raphael Weiner, VP of Product at Instana share their thoughts below.
1. With the increased focus on Automation from IBM, how do you see Instana playing into the Automation story?
Raphael: Instana was built from day one with STAN in mind — a robot that helps DevOps teams deliver their tasks faster and with lower effort. From the beginning, Instana started to automate most of the manual work that has normally been necessary to set up and configure monitoring: installing and configuring the different infrastructure and application monitoring sensors, instrumenting and configuring traces, alert set up and even root cause analysis.
It’s important to note, though, that it’s not automation just for the sake of automation. When dealing with modern, cloud native applications, change is constant — and even the infrastructure is ephemeral. Automation is the only way to ensure that you can maintain the appropriate levels of observability needed to ensure proper application operation and performance.
With this in mind, we will play into the IBM Automation story and be able to take this to the next level together with IBM Cloud Pak for Watson AIOps (more on that later).
Mike: If you think about the Automation of IT, you can relate it to a self-driving car. For the self-driving car, there is a lot of engineering put into how the car senses its external environment. It has to ‘see’ traffic, including other vehicles, bicycles, people crossing the street and road markings like road lines, intersections and train crossings. From this it can guide the vehicle’s speed, steering, safe following distance and more.
For the Automation of IT, Instana provides the enterprise observability, or the ability to sense the external environment. This environment includes bare metal, VMs, Kubernetes, Tanzu PCF environments and the applications running on 250+ different kinds of middleware — including the context of how they all map together, automatically, so the rest of the automation portfolio can then do its work.
2. Do you have examples from Instana clients using IBM AI?
Raphael: Nothing, yet, from the perspective of Instana’s clients, although AI has played an important part in application performance monitoring, especially when trying to both automate and deliver assistance in troubleshooting scenarios. It’s only a matter of time before technology from both sides of the IBM and Instana equation work together to build on that value.
Mike: Both ICBC and Caixa Bank are leveraging AI to drive to new outcomes. Caixa Bank has been able to leverage IBM Cloud Pak for Watson AIOps to more effectively work with and extract insight from unstructured data, helping them deliver greater reliability and performance with their IT infrastructure.
Without AI and automation, they would not be able to get the necessary insight to manage risk and impact. Similarly, ICBC has leveraged AI to help recognize and manage “new normals” — helping them triage resources and human insight on anomalies far more effectively. Taken together, these clients show how AI can help automate efforts that can’t be managed at scale manually.
3. Where do you see the future of enterprise observability heading?
Raphael: Ultimately, I see enterprise observability going from reactive to predictive. Where today, the focus tends to be lowering mean-time-to-resolution (MTTR), the future will be more focused on auto-remediation based on predicting potential outages or other problems.
As development organizations become more agile and DevOps teams accelerate their pipelines, they’ll want their observability, monitoring and AIOps platforms to also maintain the velocity achieved through optimized pipeline.
Achieving this level of predictive action will likely incorporate lots of different technologies:
Machine learning, applied on top of events and data (metrics, logs, traces)
- Service level indicator and objective monitoring (SLI/SLO) aligned with business metrics
- Programmatic understanding of architectural and informational context and dependencies
Mike: I see enterprise observability continuing to remove as many blind spots as possible for our enterprise clients. Tactically, while Instana’s coverage is excellent in the cloud native space — and that is a continuous investment item as cloud native progresses — we must also invest to cover the rest of the enterprise, such as the mainframe and legacy middleware stacks that our clients depend on for business-critical applications.
Strategically there will be three areas that the industry will push on for more enterprise observability:
- First is to consume business data and provide context to that business data to the application space so automation can steer the most important parts of the business as priority.
- Second is to consume log data in context with the application space so when corrections need to be made, automation has the deeper logging insights to make those decisions.
- Third is to move left into the developer code bases to put code changes in context with the application space so problems can be caught before they manifest into production problems
Learn more about IBM and Instana
The combination of the Instana platform and IBM Cloud Pak for Watson AIOps extends and enhances IBM’s AI-powered automation portfolio. Learn more about Instana and how it compliments IBM Cloud Pak for Watson AIOps, and see Instana in action by checking out this “play with” demo.