diff --git a/release_notes/ocp-4-13-release-notes.adoc b/release_notes/ocp-4-13-release-notes.adoc index b03e37b73a73..9952bf99a5ef 100644 --- a/release_notes/ocp-4-13-release-notes.adoc +++ b/release_notes/ocp-4-13-release-notes.adoc @@ -1066,7 +1066,7 @@ This update adds the following features: For more information, see xref:../scalability_and_performance/cnf-numa-aware-scheduling.adoc#cnf-about-numa-aware-scheduling_numa-aware[Scheduling NUMA-aware workloads]. [id="ocp-4-13-ran-workload-partitioning-three-node-standard-node"] -==== Support for workload partitioning for three-node clusters and standard clusters +==== Support for workload partitioning for three-node clusters and standard clusters (Technology Preview) Before this update, workload partitioning was supported for {sno} clusters only. Now, you can also configure workload partitioning for three-node compact clusters and standard clusters. @@ -1075,20 +1075,17 @@ Use workload partitioning to isolate {product-title} services, cluster managemen For more information, see xref:../scalability_and_performance/enabling-workload-partitioning.adoc#enabling-workload-partitioning[Workload partitioning]. [id="ocp-4-13-configuring-power-states-using-ztp"] -==== Configuring power states using zero touch provisioning (ZTP) +==== Configuring power states using {ztp} -{product-title} 4.12 introduced the ability to set power states for critical and non-critical workloads. This release now enables the user to configure power states by using ZTP. +{product-title} 4.12 introduced the ability to set power states for critical and non-critical workloads. +In {product-title} 4.13, you can now configure power states with {ztp}. For more information about the feature, see xref:../scalability_and_performance/ztp_far_edge/ztp-advanced-policy-config.adoc#ztp-using-pgt-to-configure-power-saving-states_ztp-advanced-policy-config[Configuring power states using PolicyGenTemplates CRs]. -[id="ocp-4-13-etcd-overview"] -==== Documentation enhancement: Overview of etcd is now available - -An overview of etcd, including the benefits it provides and how it works, is now available in the {product-title} documentation. As the primary data store for Kubernetes, etcd provides a reliable approach to cluster configuration and management on {product-title} through the etcd Operator. For more information, see xref:../architecture/control-plane.adoc#etcd-overview_control-plane[Overview of etcd]. [id="ocp-4-13-scalability-and-performance-talm-updates"] -==== Pre-caching container images for managed cluster updates with {cgu-operator} and GitOps ZTP +==== Pre-caching container images for managed cluster updates with {cgu-operator} and {ztp} -This release adds two new {cgu-operator-first} features for use with GitOps ZTP: +This release adds two new {cgu-operator-first} features for use with {ztp}: * A new check ensures that there is sufficient available disk space on the managed cluster host before cluster updates. Now, during container image pre-caching, {cgu-operator} compares the available host disk space with the estimated {product-title} image size to ensure that there is enough disk space on the host. @@ -1097,6 +1094,36 @@ Now, during container image pre-caching, {cgu-operator} compares the available h For more information see xref:../scalability_and_performance/cnf-talm-for-cluster-upgrades.adoc#talo-precache-feature-image-filter_cnf-topology-aware-lifecycle-manager[Using the container image pre-cache filter]. +[id="ocp-4-13-ran-http-transport"] +==== HTTP transport replaces AMQP for PTP and bare-metal events (Technology Preview) + +HTTP is now the default transport in the PTP and bare-metal events infrastructure. +AMQ Interconnect is end of life (EOL) from 30 June 2024. +When you use HTTP transport for PTP and bare-metal events, you must persist the events subscription in the cluster using a `PersistentVolume` resource. + +For more information, see xref:../networking/using-ptp.adoc#cnf-about-ptp-fast-event-notifications-framework_using-ptp[About the PTP fast event notifications framework]. + +[id="ocp-4-13-ran-westport-channel-grandmaster"] +==== Support for Intel E810 Westport Channel NIC as PTP grandmaster clock (Technology Preview) + +You can now configure the Intel E810 Westport Channel NIC as a PTP grandmaster clock by using the PTP Operator. +PTP grandmaster clocks use `ts2phc` (time stamp 2 physical clock) for system clock and network time synchronization. + +For more information, see xref:../networking/using-ptp.adoc#configuring-linuxptp-services-as-grandmaster-clock_using-ptp[Configuring linuxptp services as a grandmaster clock]. + +[id="ocp-413-ran-ztp-crun-container-runtime"] +==== Configuring crun as the default container runtime for managed clusters in {ztp} +A `ContainerRuntimeConfig` CR that configures crun as the default container runtime has been added to the GitOps ZTP `ztp-site-generate` container. + +For optimal performance in clusters that you install with {ztp}, enable crun for control plane and worker nodes in {sno}, {3no}, and standard clusters alongside additional Day 0 installation manifest CRs. + +For more information, see xref:../scalability_and_performance/ztp_far_edge/ztp-reference-cluster-configuration-for-vdu.adoc#ztp-sno-du-configuring-crun-container-runtime_sno-configure-for-vdu[Configuring crun as the default container runtime]. + +[id="ocp-4-13-etcd-overview"] +==== Documentation enhancement: Overview of etcd is now available + +An overview of etcd, including the benefits it provides and how it works, is now available in the {product-title} documentation. As the primary data store for Kubernetes, etcd provides a reliable approach to cluster configuration and management on {product-title} through the etcd Operator. For more information, see xref:../architecture/control-plane.adoc#etcd-overview_control-plane[Overview of etcd]. + [id="ocp-4-13-insights-operator"] === Insights Operator // https://issues.redhat.com/browse/OCPBUGS-6832 @@ -1888,7 +1915,7 @@ In the following tables, features are marked with the following statuses: |PTP dual NIC hardware configured as boundary clock |Not Available |Technology Preview -|Technology Preview +|General Availability |PTP events with boundary clock |Technology Preview @@ -2210,6 +2237,7 @@ In the following tables, features are marked with the following statuses: |==== [discrete] +[id="ocp-413-scalability-tech-preview"] === Scalability and performance Technology Preview features .Scalability and performance Technology Preview tracker @@ -2232,10 +2260,10 @@ In the following tables, features are marked with the following statuses: |Not Available |Technology Preview -|Single-node OpenShift cluster expansion with worker nodes -|Not Available +|{sno-caps} cluster expansion with worker nodes |Not Available |Technology Preview +|General Availability |{cgu-operator-first} |Technology Preview @@ -2252,6 +2280,21 @@ In the following tables, features are marked with the following statuses: |Technology Preview |General Availability +|HTTP transport replaces AMQP for PTP and bare-metal events +|Not Available +|Not Available +|Technology Preview + +|Intel E810 Westport Channel NIC as PTP grandmaster clock +|Not Available +|Not Available +|Technology Preview + +|Workload partitioning for three-node clusters and standard clusters +|Not Available +|Not Available +|Technology Preview + |==== [discrete] @@ -2527,6 +2570,48 @@ To work around this issue, before you update to {product-title} {product-version * There is a disk discovery delay when attaching storage to workloads. (link:https://issues.redhat.com/browse/OCPBUGS-11149[*OCPBUGS-11149*]) +[id="ocp-4-13-ran-known-issues"] +* If you specify an invalid subscription channel in the subscription policy that you use to perform a cluster upgrade, the {cgu-operator-first} indicates that the upgrade is successful immediately after {cgu-operator} enforces the policy because the `Subscription` resource remains in the `AtLatestKnown` state. +(link:https://issues.redhat.com/browse/OCPBUGS-9239[*OCPBUGS-9239*]) + +* After a system crash, `kdump` fails to generate the `vmcore` crash dump file on HPE Edgeline e920t and HPE ProLiant DL110 Gen10 servers with Intel E810 NIC and ice driver installed. +(link:https://issues.redhat.com/browse/RHELPLAN-138236[*RHELPLAN-138236*]) + +* In {ztp}, when you provision a managed cluster that contains more than a single node using a `SiteConfig` CR, disk partition fails when one or more nodes has a `diskPartition` resource configured in the `SiteConfig` CR. +(link:https://issues.redhat.com/browse/OCPBUGS-9272[*OCPBUGS-9272*]) + +* In clusters configured with PTP boundary clocks (T-BC) and deployed DU applications, messages are intermittently not sent from the follower interface of the T-BC on the vDU host for periods of up to 40 seconds. +The rate of errors in the logs can vary. +An example error log is below: ++ +.Example output +[source,terminal] +---- +2023-01-15T19:26:33.017221334+00:00 stdout F phc2sys[359186.957]: [ptp4l.0.config] nothing to synchronize +---- +(link:https://issues.redhat.com/browse/RHELPLAN-145492[*RHELPLAN-145492*]) + +* When you install a {sno} cluster using {ztp} and configure PTP and bare-metal events with HTTP transport, the `linuxptp-daemon` daemon pod intermittently fails to deploy. +The required `PersistentVolumeClaim` (`PVC`) resource is created but is not mounted in the pod. +The following volume mount error is reported: ++ +.Example output +[source,terminal] +---- +mount: /var/lib/kubelet/plugins/kubernetes.io/local-volume/mounts/local-pv-bc42d358: mount(2) system call failed: Structure needs cleaning. +---- +To workaround the issue, delete the `cloud-event-proxy-store-storage-class-http-events` `PVC` CR and re-deploy the PTP Operator. +(link:https://issues.redhat.com/browse/OCPBUGS-12358[*OCPBUGS-12358*]) + +* RFC2544 performance tests show that the `Max delay` value for a packet to traverse the network is over the minimum threshold. This regression is found in {product-title} 4.13 clusters running the Telco RAN DU profile. +(link:https://issues.redhat.com/browse/OCPBUGS-13224[*OCPBUGS-13224*]) + +* Performance tests run on a {sno} cluster with {product-title} 4.13 installed show an `oslat` maximum latency result greater than 20 microseconds. +(link:https://issues.redhat.com/browse/RHELPLAN-155443[*RHELPLAN-155443*]) + +* Performance tests run on a {sno} cluster with {product-title} 4.13 installed show a `cyclictest` maximum latency result greater than 20 microseconds. +(link:https://issues.redhat.com/browse/RHELPLAN-155460[*RHELPLAN-155460*]) + [id="ocp-4-13-asynchronous-errata-updates"] == Asynchronous errata updates