IBM Watson for AIOps – Bringing AI to IT Operations Management

Written by Farooq Zubairi

Practice Lead

AIOps uses artificial intelligence capabilities such as big data, machine learning, and other analytics from different data sources to simplify complex IT operations and automate data processing by providing proactive insights. We can understand Watson as a service that automates the real-time detection, diagnosis, and remediation of network and operations anomalies.

The difference between IBM Watson and other monitoring tools like App Dynamics, SolarWinds, etc. is that it automatically learns data behaviors, detects anomalies, and then suggests the appropriate solutions. Read this blog to act on your intelligence with AI & Chatbot with IBM Watson.

Why do we need AIOps?

Today, most organizations are moving from traditional infrastructure to a more dynamic & hybrid infrastructure, where some apps are hosted on the cloud, and some are on the ground. Applications and systems across these multiple environments generate tons of data, and it keeps on growing with time.

Traditional domain-based IT management solution

  • Legacy analytics solutions and BI tools cannot cope with the volume of data the applications are generating these days.
  • They cannot sort the meaningful data out of raw data generated by different applications.
  • They can't correlate data across different environments and applications.
  • They can't provide real-time insight and predictive analysis to IT operations teams.

AIOps benefits

AIOps enables IT operations to identify, address, and resolve slow-downs and outages faster than they can by examining manually through alerts from multiple IT operations tools. This results in several specific benefits:

How does AIOps work

  • Data: Collect and aggregate large volumes of operations data generated by multiple IT infrastructure components, applications, & performance-monitoring tools.
  • Machine Learning: Intelligently filter accurate alerts out of the noise to identify significant events & patterns related to system performance and availability issues.
  • Automation: Rapidly diagnose Root Causes and, in some cases, automatically resolve these issues without human intervention.

AIOps use cases

AIOps can help drive other important business and IT initiatives:

  • Digital transformation: Every organization needs digital transformation in either process, business model, domain, or cultural/organizational transformation. AIOPS can help give a 360-degree view of existing and new systems to better implement change in a more controlled manner.
  • Cloud adoption/migration: Cloud migration is a portfolio management effort. You'll decide to move some assets, invest in others, & obsolete or unused assets during the migration process. Therefore, migration always needs a framework where organizations can see potential issues upfront & can be fixed at the initial stage.
  • DevOps adoption: DevOps is also one of the use cases where AIOPs are very helpful for the operation team to automate repetitive tasks and have a holistic view of deployments.

Case Study – Enterprise System

In typical enterprises, we have the following:

  • Enterprise System/Applications

  • Hybrid Environments

  • Multiple Application and log files

  • System and Application logs

  • Millions of events

  • Streaming and messaging

It's not easy for traditional monitoring applications to handle all of these in a single place; AIOPs can help define the data correlation and semantic understanding of data between discrete systems. AIOps is a better fit for medium to large enterprises.

Watson AIOps on Cloud Pak for Data v3.0.1 - v2.1

IBM Watson OpenScale can now be run in any environment – on-premises or on any private, public, or hybrid multi-cloud – enabling businesses to predict based on the data available, no matter where it is hosted. In addition, companies will be able to build models into their apps, regardless of where they reside. This flexibility can remove one of the major obstacles to scaling AI since businesses can now leave data in secure or preferred environments and take Watson to that data.

IBM Watson® AIOps 2.0 is composed of four components:

  • IBM Watson AIOps AI Manager
  • IBM Watson AIOps Metric Manager (IBM® Operations Analytics Predictive Insights 1.3.6)
  • IBM Watson AIOps Topology (IBM Netcool Agile Service Manager 1.1.9)
  • IBM Watson AIOps Event Manager (IBM Netcool® Operations Insight® 1.6.2)

Watson AIOps on Cloud Pak for Data v3.0.1 - v2.1

IBM Watson OpenScale can now be run in any environment – on-premises or on any private, public, or hybrid multi-cloud – enabling businesses to predict based on the data available, no matter where it is hosted. In addition, companies will be able to build models into their apps, regardless of where they reside. This flexibility can remove one of the major obstacles to scaling AI since businesses can now leave data in secure or preferred environments and take Watson to that data.

IBM Watson® AIOps 2.0 is composed of four components:

  • IBM Watson AIOps AI Manager
  • IBM Watson AIOps Metric Manager (IBM® Operations Analytics Predictive Insights 1.3.6)
  • IBM Watson AIOps Topology (IBM Netcool Agile Service Manager 1.1.9)
  • IBM Watson AIOps Event Manager (IBM Netcool® Operations Insight® 1.6.2)

IBM Watson AIOps Installation and License

  • IBM Watson® AIOps 2.1 brings together a set of complementary components to create a unified, robust AIOps solution for your organization.
  • You can install Watson AIOps AI Manager and Event Manager together, with Metric Manager separately. Purchasing a license for Watson AIOps entitles you to all three components, but installation is performed on a per-component basis.

IBM Watson AIOps Implementation

Watson can be implemented in any kind of environment, whether it's local, on-premise, cloud, hybrid, or multi-cloud.

Input

Watson collects structured or unstructured data from different applications, systems, and environments in the form of logs, tickets, events, metrics, and topologies.

Processing

  • Consolidate Data: Consolidate, correlate and remove any unwanted data.
  • Detect Anomalies: Anomalies are also known as noise, deviations, and exceptions in input data. Watson algorithms remove all possible anomalies and use unique, filtered, and noise-free data.
  • Event Grouping: An event indicates that something noticeable happened in a production environment. For example, a response time of an application is slow, or a hard disk capacity reached its limit, etc. The purpose of event grouping and classification is to quickly understand the issues in IT operations and focus on a few critical events that need immediate attention. Fault radius: Draw fault radius to check the scope of its impact on the ecosystem.
  • Incident resolution: Watson AIOps consumes incident ticket data by connecting to tools such as ServiceNow to provide timely next-best action.

Output

  • Chatops/APIs/Dashboard: In Watson AIOps, all insights are delivered via ChatOps and dashboards.
  • Runbooks: Watson AIOps provides a customizable means of suggesting solutions from automated run books for the subsequent best action recommendations. IBM Runbook Automation can automate procedures that do not require human interaction, increasing IT operations processes' efficiency.

Summary

In the next generations of Watson AIOps solutions, Royal Cyber envisions fully instrumented, visible, self-aware, automated, and autonomic IT operations environments. Watson AIOps solutions will help resolve issues in a responsive mode and avoid issues from taking place in the first place by creating the Development-Security-Operations (DevSecOps) lifecycle activities for effective operations.

Drive Down Service Outages with AIOps, watch this on-demand webinar to know how we helped reduce noise by 99% using ServiceNow Event Management. Our team is striving to shape the future and take your business to the next level. For more information, you can email us at info@royalcyber.com or visit www.royalcyber.com.

Leave a Reply