Merge pull request #77581 from mburke5678/cpuset.cpus.exclusive

cgroupsv2: failed to write on cpuset.cpus.exclusive

Author: mburke5678

release_notes/ocp-4-16-release-notes.adoc

@@ -3125,6 +3125,8 @@ In the following tables, features are marked with the following statuses:
 
 * When a {product-title} {product-version} cluster is installed or configured as a postinstallation activity on a single VIOS host with virtual SCSI storage on {ibm-power-name} with multipath configured, the CoreOS nodes with multipath enabled fail to boot. This behavior is expected as only one path is available to the node. (link:https://issues.redhat.com/browse/OCPBUGS-32290[*OCPBUGS-32290*])
 
+* When using CPU load balancing on cgroupv2, a pod can fail to start if another pod that has access to exclusive CPUs already exists. This can happen when a pod is deleted and another one is quickly created to replace it. As a workaround, ensure that the old pod is fully terminated before attempting to create the new one. (link:https://issues.redhat.com/browse/OCPBUGS-34812[*OCPBUGS-34812*]) 
+
 [id="ocp-telco-ran-4-16-known-issues_{context}"]
 * The current PTP grandmaster clock (T-GM) implementation has a single National Marine Electronics Association (NMEA) sentence generator sourced from the GNSS without a backup NMEA sentence generator.
 If NMEA sentences are lost before reaching the e810 NIC, the T-GM cannot synchronize the devices in the network synchronization chain and the PTP Operator reports an error.