What's new in IBM Cloud Pak for AIOps 4.6.0

Get a quick overview of what is new with IBM Cloud Pak® for AIOps.

IBM Cloud Pak for AIOps leverages AI to streamline IT operations, increase uptime, improve efficiency, and reduce costs. The solution supports Operations teams to deliver optimum performance and assurance by:

  • Breaking down data silos
  • Providing real-time insights into IT systems
  • Augmenting incident management with AI
  • Automating incident resolution

With Version 4.6.0, IBM Cloud Pak for AIOps delivers new and improved tools for installing, integrations, anomaly detection, alert management, process automation, resource management, and infrastructure automation.

Overall, the following new features and changes are introduced with this IBM Cloud Pak version:

Install and upgrade

Incremental adoption You can now choose to deploy a base deployment of IBM Cloud Pak for AIOps that has a smaller footprint and does not have log anomaly or ticket analysis capabilities. The option to deploy an extended deployment of IBM Cloud Pak for AIOps with these capabilities remains. Learn more

Installing on Linux® Technology preview feature: you can now install a nonproduction IBM Cloud Pak for AIOps deployment directly on Linux, without the need to install Red Hat® OpenShift® Container Platform. A new tool, aiopsctl, is introduced to install IBM Cloud Pak for AIOps on a Linux cluster. Learn more

AWS native storage Support for Amazon Web Services (AWS) native storage is introduced, using Amazon Elastic Block Store (EBS) and Amazon Elastic File System (EFS). Learn more

Additional install and upgrade new and changed features:

  • IBM Storage Fusion 2.8 support IBM Storage Fusion 2.8 support is introduced for the backup and restore of IBM Cloud Pak for AIOps (multi-region cold standby disaster recovery). Previously only IBM Storage Fusion 2.7.2 was supported.

  • Usage data collection Additional usage data about the deployment type, deployment platform, and secure tunnel enablement is collected in this release. For more information, see Usage data collection.

Defining integrations (connections)

Dynatrace Observer proxy support and updates: The Dynatrace Observer is updated to add proxy support, new parameters, and new and enhanced entitySelectors for the Dynatrace v2 API. See Configuring Dynatrace Observer jobs

The following integrations are also updated:

AI model management

Metric blocklist for data processing Metric anomaly detection is updated to support purging and blocking metric data. From the Metric Search UI, you can define a block list to prevent specific metric data from being processed for the metric anomaly detection. You can also purge the metric data for all resources that are inlcuded in the block list.

Additional AI model management updates:

  • Log anomaly detection performance The log anomaly detection UI performance during template editing is enhanced to address slow response times that were caused by Flink refreshes.

Resolving incidents and alerts

Alert count breakdown: You can now see a more comprehensive breakdown of the alert counts from within the Alert Viewer. There is also a countdown timer for table refresh in the bottom right corner of the page. For more information, see the Alert Viewer.

Managing applications (services) and viewing topologies

  • New 'Structure view' tab: When viewing resource details, the new structure view tab displays all resources that are part of a host or parent resource (linked together in a relationship via the 'partOf' edge) regardless of their position; that is, it displays linked resources even if they are several hops removed from one another or from the host (parent) resource.

  • New 'compare resources' functionality: New functionality lets you compare resources side-by side in a properties table.

  • New manual layout option (for 'exact' templates): You can now manually adjust your topology layout while creating an exact template in the template builder, and then replicate that layout when resource groups are generated from the template. You can also adjust your layout in the topology viewer and then launch the template builder from the navigation bar with your template definition and exact layout already selected.

  • New advanced option when assigning icons to resource types: You can now use scripting to refine the way icons are assigned to resource types.

  • Administration - Topology data: Updated the script that clears out the topology and inventory databases. See Clearing out data without reinstalling.

Automating processes (Policies, Runbooks, Actions)

  • Runbook Automation: You can use "SSH key" as an authentication type in the SSH jump server configuration if you are using the configuration API. This feature is not yet supported in the SSH configuration UI. For more information, see the Automation (runbook) API Swagger document.

  • Policies: New countdown timer for policy table refresh is in the bottom right corner of the page. See About policies.

  • Policies: When the alert.details property is selected in a policy condition set, the Details name field allows you to minimize the scope of the policy to a singular key within the alert's details. You can now apply Details name to other areas of policies. Note, for Scope ID and runbook parameters you can send only specific alert details by name and not the entire alert.details object. You must enter a string value in the Details name field for Scope ID and runbook parameters.

    • Scope ID: you can now add details.name references for the value of properties that define the scope of how alerts are grouped. See Group alerts based on scope.
    • Alert update in Policy triggers now supports details.name (not applicable to suppression policies). See Policy editor.
    • Runbook parameters support details.name for string variable type parameters only. See Assign a runbook to alerts.

Infrastructure Automation

Non-production (tech preview)

  • Installing on Linux® is now available for non-production use. See Install and upgrade.

  • The following non-production features are no longer available with Version 4.6.0:

    • Selecting alerts as valid or invalid (as feedback to improve training).
    • Log anomaly automated training data selection.
    • Feedback learning for statistical baseline log anomaly detection.
    • Log anomaly model drift detection.

Security and compliance

  • Fixed security vulnerabilities: Fixes and updates are included to address reported security-related vulnerabilities with IBM Cloud Pak for AIOps. For more information about the included fixes, see Security bulletins and fixes.

Documentation changes

In addition to the documentation for the preceding new and changed features, the following documentation is updated:

  • References to Podman replaces Docker for container management commands, because Docker is not shipped or supported for Red Hat Enterprise Linux (RHEL) 8 or 9. For more information, see the Red Hat customer portal Opens in a new tab.

  • The instructions for Restarting the cluster that IBM Cloud Pak for AIOps is deployed on are updated.

  • Documentation for alert enrichment is now available. See About golden signal alert enrichment.

  • Documentation references to the following internal components are removed. These components were used in previous versions of IBM Cloud Pak for AIOps, but are now removed:

    • API gateway Kong
    • Luigi daemon

  • Documentation related to the following disabled non-production (tech preview) features are removed:

    • Log anomaly automated training data selection
    • Feedback learning for statistical baseline log anomaly detection
    • Log anomaly model drift detection

  • The documentation now lists Support reported issues that are now fixed with Version 4.6.0. See Fixed known issues.

  • Addition of storage classes and storage provider renaming.

  • Removal of disruption cost chart in the AIOps Insights dashboard page.

  • Addition of supported languages page. For more information, see Supported languages.

Previously introduced features and changes

For more information about new features and changes that were introduced in previous versions of IBM Cloud Pak for AIOps, see: