-
Bug
-
Resolution: Not a Bug
-
Major
-
None
-
4.14, 4.15, 4.16
-
Important
-
No
-
False
-
-
-
Description of problem:
Cloud-event processes are running on isolated cpus instead of reserved cpus.
isolated: "2-71,74-143" reserved: "0-1,72-73"
[core@e32-h22-r750 ~]$ ps e | grep "cloud"
863598 ? 00:00:00 cloud-event-con
863706 ? 00:00:00 cloud-event-pro
[core@e32-h22-r750 ~]$ taskset -pc 863598
pid 863598's current affinity list: 2-71,74-143
[core@e32-h22-r750 ~]$ taskset -pc 863706
pid 863706's current affinity list: 2-71,74-143
[core@e32-h22-r750 ~]$ taskset -pc 863706
This issue is also observed in ocp 4.15.2 and 4.14.15.
Version-Release number of selected component (if applicable):{code:none}
OCP version: 4.16.0-ec.5
containerRuntimeVersion: cri-o://1.29.2-9.rhaos4.16.git5052e1a.el9
kernelVersion: 5.14.0-432.el9.x86_64+rt
kubeProxyVersion: v1.29.2+258f1d5
kubeletVersion: v1.29.2+258f1d5
OperatingSystem: linux
osImage: CentOS Stream CoreOS 416.94.202403300731-0
# oc get csv -A
NAMESPACE NAME DISPLAY VERSION REPLACES PHASE
openshift-local-storage local-storage-operator.v4.15.0-202403220332 Local Storage 4.15.0-202403220332 Succeeded
openshift-operator-lifecycle-manager packageserver Package Server 0.0.1-snapshot Succeeded
openshift-ptp ptp-operator.v4.15.0-202403221440 PTP Operator 4.15.0-202403221440 Succeeded
How reproducible:
1. Deploy SNO with DU configuration
2. Create cloud-events namespace (add the following annotation):
workload.openshift.io/allowed: management
3. Deploy PTP Operator config and ptp events CRs
4. Check cpu-affinity of cloud-event processes on the cluster
Steps to Reproduce:
1.
2.
3.
Actual results:
Cloud-event processes are running on isolated cpus
Expected results:
Cloud-event processes should run on reserved cpus
Additional info:
Attaching all related config yaml files.
- relates to
-
-
- Verified
-
-
-
- Verified
-