Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-42323

IRQBALANCE_BANNED_CPUS is not updated with isolated cpus when irq load balancing is disabled

XMLWordPrintable

    • Yes
    • Proposed
    • False
    • Hide

      None

      Show
      None
    • Hide
      When globallyDisableIrqLoadBalancing is switched from true to false in Performance profile can cause the number of CPUs available for rebalancing to reduce. This should not affect workloads.

      Note for doc team: This is a regression introduced in 4.17.0 and fixed already in nightlies that will be included in 4.17.0
      Show
      When globallyDisableIrqLoadBalancing is switched from true to false in Performance profile can cause the number of CPUs available for rebalancing to reduce. This should not affect workloads. Note for doc team: This is a regression introduced in 4.17.0 and fixed already in nightlies that will be included in 4.17.0
    • Known Issue
    • In Progress

      Description of problem:

          when globallyDisableIrqLoadBalancing is set to true in Performance profile IRQBALANCE_BANNED_CPUS doesn't get updated with isolated cpus set in the profile instead IRQBALANCE_BANNED_CPULIST gets updated 
      
      

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

      4.17.0-rc.5
          

      How reproducible:
      Everytime

          

      Steps to Reproduce:

          1. Setup performance profile as show below:
      
      spec:
        cpu:
          isolated: 1,3-38,41,43-78
          reserved: 0,2,39,40,42,79
        globallyDisableIrqLoadBalancing: true
        hugepages:
          defaultHugepagesSize: 1G
          pages:
          - count: 1
            node: 0
            size: 1G
          - count: 128
            node: 1
            size: 2M
        machineConfigPoolSelector:
          machineconfiguration.openshift.io/role: worker-cnf
        net:
          userLevelNetworking: true
        nodeSelector:
          node-role.kubernetes.io/worker-cnf: ""
        numa:
          topologyPolicy: single-numa-node
        realTimeKernel:
          enabled: false
        workloadHints:
          highPowerConsumption: true
          perPodPowerManagement: false
          realTime: true
      
          2.  check /etc/sysconfig/irqbalance in the worker-cnf node:
      IRQBALANCE_BANNED_CPUS=0
      IRQBALANCE_BANNED_CPULIST=1,3-38,41,43-78
      
          3.
          

      Actual results:

      IRQBALANCE_BANNED_CPUS=0
      IRQBALANCE_BANNED_CPULIST=1,3-38,41,43-78
      
          

      Expected results:

      IRQBALANCE_BANNED_CPUS=1,3-38,41,43-78
          

      Additional info:

          

            yquinn@redhat.com Yanir Quinn
            mniranja Mallapadi Niranjan
            Liquan Cui Liquan Cui
            Votes:
            0 Vote for this issue
            Watchers:
            9 Start watching this issue

              Created:
              Updated: