Problem Statement

When evaluating OpenShift Container Platform cluster configuration the oc adm must-gather tool is indispensable for collecting diagnostic information. A critical feature, must-gather --all-images, is designed to list all container images required by the cluster's deployed components, which is vital for setting up disconnected environments, pre-pulling images, and validating image availability.

However, many "Day-2" Operators, do not include the necessary Kubernetes annotations on their operand deployments or custom resources defined in the ClusterServiceVersion. This omission results in the must-gather --all-images command providing an incomplete list of required images, forcing manual and potentially error-prone discovery processes for critical components.

This gap is particularly problematic in:

• Complex deployments with numerous Day-2 Operators.
• Scenarios where all logs, configuration, etc. required for troubleshooting needs to be gathered at the same time to enable rapid Root Cause Analysis (RCA) and to avoid support escalation.

Proposed Solution (Enhancement)

In order to be able to use the must-gather --all-images feature the Red Hat-supported Day-2 Operators reference in this document must be updated to consistently include the relevant and accurate Kubernetes annotations on their deployed operands and associated Custom Resources (CRs). This is to ensure that must-gather --all-images can correctly identify and list all operators for the purpose of checking the deployed state of the Operator's configuration.

Extra Information
This link explains how must-gather works with the –all-images parameter.  https://github.com/openshift/enhancements/blob/master/enhancements/oc/must-gather.md

Operators include a special annotation in their CSV (ClusterServiceVersion), typically like:

annotations:
operators.openshift.io/collection-image: "quay.io/my-org/my-operator-must-gather:latest"

Specific Day-2 Operators for annotation

The Day-2 Operators are as follows.   

ACM - Advanced Cluster Management

CLO - Cluster Logging Operator

Compliance Operator

KMM - Kernel Module Management Operator

LCA Operator

LSO - Local Storage Operator

LVMS Operator

MCE - MultiCluster Engine

OADP - OpenShift API for Data Protection

ODF - OpenShift Data Foundation

OpenShift GitOps Operator

SRIOV Network Operator

TALM - Topology Aware Lifecycle Manager

Why does the customer need this? 

Ensuring Day-2 Operators properly annotate their images for must-gather --all-images provides significant benefits, especially for our partners and customers:

• Validation of Reference Architectures:
• The ability to collect a complete accurate snapshot of the cluster's current configuration state through must-gather is a critical aspect of the blueprint design validation and deviation detection process.

• Reduced Partner/Customer Burden:
• Currently, Red Hat support teams, professional services and partners must manually inspect Operator manifests, pod definitions, and ClusterServiceVersion (CSV) resources to piece together a complete view of the Day-2 Operators deployed. This has the potential to be time-consuming and error-prone. Correct annotations automate this burden.

• Enhanced Operational Consistency and Automation:
• Enables automated workflows for cluster configuration validation and security scanning.

• Provides a standardized way for partners and customers to gather must-gathers for the complete deployment without requiring knowledge of the individual must-gather images that should be used for each operator

• Simplified and Automated Troubleshooting:
• Comprehensive Data Collection: The all-images option relies on these annotations to build a complete list of images used by the cluster. Without them, must-gather misses critical images, making initial diagnostics incomplete and delaying troubleshooting.

• Reduced Mean Time To Resolve (MTTR): Field engineers and Red Hat Support can obtain a full and accurate list of images in one command, drastically reducing the time spent manually identifying missing images or verifying image versions. This accelerates the problem identification and resolution process.

Improved Support Experience: Customers receive faster, more accurate support when comprehensive data can be gathered efficiently, preventing multiple back-and-forths for image information.