Uploaded image for project: 'OpenShift Request For Enhancement'
  1. OpenShift Request For Enhancement
  2. RFE-8681

Provide native PXE server management in OpenShift for Hosted Control Planes (IPMI fallback support)

XMLWordPrintable

    • None
    • Product / Portfolio Work
    • None
    • False
    • Hide

      None

      Show
      None
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      1. Proposed title of this feature request

      Provide native PXE server management in OpenShift for Hosted Control Planes (IPMI fallback support)

      2. What is the nature and description of the request?

      Problem Statement

      Not all bare-metal hardware platforms implement the Redfish standard in a consistent or reliable manner. In several environments, Redfish functionality is incomplete, unstable, or non-functional (for example, issues observed on certain B200 platforms).

      Because of this, customers must continue to rely on IPMI-based provisioning with PXE/iPXE as a fallback mechanism. This fallback is critical for out-of-band management and recovery in scenarios such as:

      • Node power cycling or reboots
      • Virtual machine or hypervisor failures
      • Loss of in-band agent connectivity
      • Network disruptions during provisioning or upgrades

      In these cases, recovery depends on reliable out-of-band access. Without PXE/IPMI support, recovery becomes tightly coupled to agent health and in-band networking, which may not be available during failure scenarios.

      Current Limitations

      Today, PXE-based provisioning requires externally managed infrastructure, typically:

      • External DHCP servers
      • Manually maintained DHCP rules forwarding to specific PXE/iPXE paths

      This introduces several operational challenges:

      • Each new inventory, node type, or provisioning workflow often requires creating or updating unique PXE paths
      • DHCP configurations must be manually updated and kept in sync with inventory changes
      • At scale, especially in environments with heterogeneous hardware, this process is:
        • Error-prone
        • Difficult to automate
        • Hard to audit
        • Risky during failure recovery

      This external dependency undermines the declarative and automated operational model that OpenShift aims to provide.

      Requested Enhancement

      Provide the ability to manage a PXE server natively within the hosting OpenShift cluster (MCE / ACM) for Hosted Control Plane environments.

      The enhancement should enable:

      • PXE/iPXE server lifecycle management as OpenShift-managed components
      • Declarative PXE configuration driven by cluster, inventory, or bare-metal CRDs
      • Automatic generation and lifecycle management of PXE paths per node, role, or hardware profile
      • Seamless integration with IPMI-based bare-metal provisioning workflows
      • Reduced or eliminated need for manual external DHCP/PXE configuration

      This would allow PXE infrastructure to be treated as first-class, declaratively managed OpenShift resources.

       


      Use Cases

      • Bare-metal deployments where Redfish support is incomplete or unreliable
      • Large-scale IPMI-managed environments with heterogeneous hardware
      • Hosted Control Plane clusters requiring robust out-of-band recovery mechanisms
      • Edge and disconnected environments where manual PXE/DHCP management is costly and error-prone

      Acceptance Criteria (Suggested)

      • PXE/iPXE server can be deployed and managed as part of the hosting OpenShift cluster
      • PXE configuration is fully declarative and reconciled via OpenShift APIs
      • PXE paths are automatically created and updated based on inventory or cluster lifecycle events
      • IPMI-based provisioning and recovery workflows function without external PXE management
      • Failure recovery does not depend on in-band agent availability

      3. Why does the customer need this? (List the business requirements here)

      Impact / Business Value

      This enhancement would:

      • Reduce operational complexity by removing manual DHCP/PXE coordination
      • Improve recovery reliability during failure scenarios where in-band access is unavailable
      • Lower operational risk due to misconfiguration or stale PXE paths
      • Enable scalable and repeatable provisioning for large, heterogeneous bare-metal fleets
      • Preserve critical IPMI-based workflows where Redfish is unreliable or unavailable
      • Improve overall platform resilience and maintainability

      This is especially valuable for customers operating at scale, at the edge, or in environments with diverse hardware platforms.

      4. List any affected packages or components.

      ACM, MCE

              bweidenb@redhat.com Bradd Weidenbenner
              rhn-support-chdeshpa Chinmay Deshpande
              None
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                None
                None