IBM Cloud Pak for AIOps workflows

Installing IBM Cloud Pak for AIOps

Deployments on Red Hat® OpenShift® Container Platform

Platform engineers prepare all required hardware and software, and then install the Red Hat® OpenShift® Container Platform. They prepare OpenShift clusters and install IBM Cloud Pak for AIOps into those clusters.

For more information about installing IBM Cloud Pak for AIOps, see the following sections:

• Planning installation, configuration, and overall cluster requirements
• Preparing for and installing IBM Cloud Pak for AIOps

Deployments on Linux®

Platform engineers prepare all required hardware and software. They prepare the Linux cluster and install IBM Cloud Pak for AIOps onto the cluster.

For more information about installing IBM Cloud Pak for AIOps, see the following sections:

• Planning installation, configuration, and overall cluster requirements
• Preparing for and installing IBM Cloud Pak for AIOps

Guided tours

To help first-time users get started with some workflows, guided tours are included within the UI console across the different dashboards. These tours introduce users to the various UI options and components, and demonstrate how to complete key tasks across the platform. Users can access the tours at any time by clicking the Tours icon in the console toolbar, and selecting their preferred tour.

For more information about the available guided tours, see Guided tours.

Setting up integrations

You can create integrations for logs, to establish a baseline of normal behavior and identify anomalies. Other integrations provide event, applications, and infrastructure data to IBM Cloud Pak for AIOps.

Systems engineers create integrations to ticketing and to ChatOps tools. The integrations to ticketing tools have a dual purpose: they enable ticket data to be collected for change request and similar ticket training. The same integration also enables synchronization between IBM Cloud Pak for AIOps incidents and corresponding ServiceNow tickets. The integrations to Microsoft Teams and Slack ChatOps tools enable incident notifications to be pushed to these ChatOps tools. These notifications are for the attention of Site Reliability Engineers who are tasked with resolving those incidents.

IBM Cloud Pak for AIOps provides many connectors, probes, and observers by default for setting up integrations with common IBM and third-party systems and services. In addition, if a default integration is not available for connecting to a system or service that you want to use with IBM Cloud Pak for AIOps, you can set up a custom integration to connect to that system or service.

For more information about setting up integrations, see the following sections:

• Defining outgoing integrations
• Defining incoming integrations

Loading data and training AI

As a prerequisite to training your AI to provide insights on your monitored environment, systems engineers load event, log, metric, ticket, and other data from various data sources. They then run different types of AI training to generate AI models such as log anomaly models, which can easily identify and report on anomalies in your log data. Insights that are generated by these AI models are used to enrich incidents with related alert, similar ticket, and event grouping information.

For more information about loading data and training AI, see the following sections:

• Planning data loading and training
• Defining incoming integrations
• Configuring AI training

Building applications

Systems engineers are responsible for building the applications that are monitored by Site Reliability Engineers. Systems engineers build applications from the ground up, by first defining observer integrations to import resource topology data, and then defining templates to automatically generate resource groups from the incoming observer streams. When they have an incoming dynamic stream of resource groups, they build business applications that are made up of multiple resource groups.

For more information about building applications, see the following sections:

• Defining observer jobs
• Defining templates with template builder
• Creating and editing applications
• Viewing applications

Creating automations

System administrators set up policies, runbooks, and actions to detect and remediate incoming alerts. Policies are rules that contain a condition and a set of outcomes that can be manual or automated. Runbooks help to solve common operational problems by automating procedures that do not require human interaction. Procedure-based documentation can be transferred into manual runbooks. Actions in runbooks are the collection of several manual steps into a single automated entity. Actions can be added to create semi-automated or fully automated runbooks.

For more information about creating automations, see the following sections:

• Creating policies
• Creating a runbook
• Creating an action

Resolving incidents

Site reliability engineers are responsible for investigating and resolving incidents and incidents that are associated with applications and infrastructure. They use ChatOps tools such as Microsoft Teams and Slack to monitor incident channels, and investigate incidents by using application, topology, ticketing, alert, and similar ticket information provided in the incident notification. They act to resolve the incidents, where necessary involving Applications developers.

For more information about resolving incidents, see the following sections:

• ChatOps insight management
• Application incidents

Administering the cluster

Platform engineers administer the cluster on which IBM Cloud Pak for AIOps is installed on an ongoing basis. They ensure that all nodes, pods, and other elements within the cluster are performing as expected. They also support the systems engineer with complex cluster administration tasks such as multitenancy management, project (namespace) management, and authorization and authentication of users within the cluster.

Administering IBM Cloud Pak for AIOps

Systems engineers, or in some organizations, Operations engineers, are responsible for authenticating and managing users. They can also customize the UI experience by setting console branding and customizing what content users see on the home page. For more complex cluster-based tasks such as multitenancy management, project (namespace) management, and authorization and authentication of users within the cluster, they collaborate with the Platform engineer.

For more information about administering IBM Cloud Pak for AIOps, see the following sections:

• Managing user access control
• Customizing the platform UI