Description of problem: When applying the default DU profile as provided in ztp today, there is a critical mismatch between what PAO does and what our custom 'tuned-performance-patch' change does. Version-Release number of selected component (if applicable): N/A How reproducible: 100% of the time Steps to Reproduce: 1. Build the default DU profile, editing the performance profile to match your hardware 2. Apply the resulting DU profile 3. Check if the command line in /proc/cmdline contains the expected CPU pinning arguments (or anything else tuned normally is supposed to add) Actual results: Tuned did not make any changes Expected results: Tuned should have applied our performance profile including kernel commandline changes. Additional info: When PAO generates a tuned profile, it takes the associated PAO policy object name and prepends openshift-node-performance-. This means that because our examples have a PAO policy named openshift-node-performance-profile the tuned performance patch needs to refer to a tuned policy named openshift-node-performance-openshift-node-performance-profile
I've got a fix for this under way.
Verified on 4.9 rc.0. kernel arg is updated properly. [yliu1@yliu1 ~]$ oc get tuneds.tuned.openshift.io -n openshift-cluster-node-tuning-operator performance-patch -o yaml apiVersion: tuned.openshift.io/v1 kind: Tuned metadata: creationTimestamp: "2021-09-10T01:55:43Z" generation: 1 name: performance-patch namespace: openshift-cluster-node-tuning-operator resourceVersion: "21298" uid: aeeda45a-07e0-452e-84a1-11fb286c01f4 spec: profile: - data: | [main] summary=Configuration changes profile inherited from performance created tuned include=openshift-node-performance-openshift-node-performance-profile [bootloader] cmdline_crash=nohz_full=2-39,42-79 [yliu1@yliu1 ~]$ oc get performanceprofiles.performance.openshift.io -o yaml apiVersion: v1 items: - apiVersion: performance.openshift.io/v2 kind: PerformanceProfile metadata: creationTimestamp: "2021-09-10T02:01:48Z" finalizers: - foreground-deletion generation: 1 name: openshift-node-performance-profile resourceVersion: "26602" ... runtimeClass: performance-openshift-node-performance-profile tuned: openshift-cluster-node-tuning-operator/openshift-node-performance-openshift-node-performance-profile [yliu1@yliu1 ~]$ oc exec -n openshift-machine-config-operator machine-config-daemon-4ztgz -- cat /proc/cmdline Defaulted container "machine-config-daemon" out of: machine-config-daemon, oauth-proxy BOOT_IMAGE=(hd4,gpt3)/ostree/rhcos-af9203de93d5eeb77503e85b9dd8bb3677661f3c43934153c6827501b2659823/vmlinuz-4.18.0-305.17.1.rt7.89.el8_4.x86_64 random.trust_cpu=on console=tty0 console=ttyS0,115200n8 ignition.platform.id=metal ostree=/ostree/boot.1/rhcos/af9203de93d5eeb77503e85b9dd8bb3677661f3c43934153c6827501b2659823/0 ip=eno1:dhcp root=UUID=607f528d-93c8-414a-a342-b2faa5b1c95d rw rootflags=prjquota intel_iommu=on iommu=pt skew_tick=1 nohz=on rcu_nocbs=2-39,42-79 tuned.non_isolcpus=00000300,00000003 intel_pstate=disable nosoftlockup tsc=nowatchdog intel_iommu=on iommu=pt isolcpus=managed_irq,2-39,42-79 systemd.cpu_affinity=0,1,40,41 default_hugepagesz=1G hugepagesz=1G hugepages=32 idle=poll nohz_full=2-39,42-79
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: OpenShift Container Platform 4.9.0 bug fix and security update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2021:3759