Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-42523

NUMA resources operator can allow more than 1 machineconfigpool per nodegroup

XMLWordPrintable

    • None
    • CNF Compute Sprint 261, CNF Compute Sprint 262, CNF Compute Sprint 263, CNF Compute Sprint 264, CNF Compute Sprint 265, CNF Compute Sprint 266
    • 6
    • False
    • Hide

      None

      Show
      None
    • Hide
      Previously, you could configure the NUMA resources operator to map a `nodeGroup` to more than one `MachineConfigPool`. This implementation is contrary to the intended design of the Operator, which assumed a one-to-one mapping between a `nodeGroup` and a `MachineConfigPool`. With this release, if a `nodeGroup` maps to more than one `MachineConfigPool`, the Operator accepts the configuration, but the Operator state moves to `Degraded`. To retain the previous behaviour, you can apply the `config.node.openshift-kni.io/multiple-pools-per-tree: enabled` annotation to the NUMA resources operator. However, the ability to assign a `nodeGroup` to more than one `MachineConfigPool` will be removed in a future release. (link:https://issues.redhat.com/browse/OCPBUGS-42523[*OCPBUGS-42523*])
      Show
      Previously, you could configure the NUMA resources operator to map a `nodeGroup` to more than one `MachineConfigPool`. This implementation is contrary to the intended design of the Operator, which assumed a one-to-one mapping between a `nodeGroup` and a `MachineConfigPool`. With this release, if a `nodeGroup` maps to more than one `MachineConfigPool`, the Operator accepts the configuration, but the Operator state moves to `Degraded`. To retain the previous behaviour, you can apply the `config.node.openshift-kni.io/multiple-pools-per-tree: enabled` annotation to the NUMA resources operator. However, the ability to assign a `nodeGroup` to more than one `MachineConfigPool` will be removed in a future release. (link: https://issues.redhat.com/browse/OCPBUGS-42523 [* OCPBUGS-42523 *])
    • Bug Fix
    • Done

      Description of problem:

      Due to insufficient internal validation, the NUMA resources operator can allow defining NodeGroups that map to more than a single MachineConfigPool. The operator was however designed in such a way that a NodeGroup must not map to more than a single pool of nodes. Users that wishes to manage more than a single pool should instead use more node groups.

      Version-Release number of selected component (if applicable):

      4.14

      How reproducible:

      not enough data. Issue discovered internally through code review.

      Steps to Reproduce:

          1. create 2 mcps that share 1 label under the metadata section
    2. install nrop and create CR with node group with the shared mcp label from 1.
    3.

      Actual results:

      The issue manifests itself with multiple MachineConfigPools reported in status when a single NodeGroup is configured in spec

      Expected results:

          Each NodeGroup must match one pool, so the number of MachineConfigPools in status must match the number of NodeGroups in spec.

      Additional info:

              rhn-support-shajmakh Shereen Haj
              fromani@redhat.com Francesco Romani
              Roy Shemtov Roy Shemtov
              Votes:
              0 Vote for this issue
              Watchers:
              9 Start watching this issue

                Created:
                Updated: