Skip to content

Commit a3a220c

Browse files
committed
OSDOCS-14369: 4.18.10 z-stream RN
1 parent c5276de commit a3a220c

File tree

1 file changed

+44
-0
lines changed

1 file changed

+44
-0
lines changed

release_notes/ocp-4-18-release-notes.adoc

+44
Original file line numberDiff line numberDiff line change
@@ -3016,6 +3016,50 @@ This section will continue to be updated over time to provide notes on enhanceme
30163016
For any {product-title} release, always review the instructions on xref:../updating/updating_a_cluster/updating-cluster-web-console.adoc#updating-cluster-web-console[updating your cluster] properly.
30173017
====
30183018

3019+
// 4.18.10
3020+
[id="ocp-4-18-10_{context}"]
3021+
=== RHSA-2025:4019 - {product-title} {product-version}.10 bug fix update and security update
3022+
3023+
Issued: 22 April 2025
3024+
3025+
{product-title} release {product-version}.10 is now available. The list of bug fixes that are included in the update is documented in the link:https://access.redhat.com/errata/RHSA-2025:4019[RHSA-2025:4019] advisory. The RPM packages that are included in the update are provided by the link:https://access.redhat.com/errata/RHBA-2025:4021[RHBA-2025:4021] advisory.
3026+
3027+
Space precluded documenting all of the container images for this release in the advisory.
3028+
3029+
You can view the container images in this release by running the following command:
3030+
3031+
[source,terminal]
3032+
----
3033+
$ oc adm release info 4.18.10 --pullspecs
3034+
----
3035+
3036+
[id="ocp-4-18-10-enhancements_{context}"]
3037+
==== Enhancements
3038+
3039+
* In the bootstrap phase of the installation process, the Transport Layer Security (TLS) between the `metal3` `httpd` server and the node's Baseboard Management Controller (BMC) is enabled by default in {product-title} 4.18 and later. The `httpd` server is on port 6183 instead of port 6180 when TLS is enabled. Disable the TLS setting by adding 'disableVirtualMediaTLS: true' to the provisioning custom resource (CR) file that is created on the disk. (link:https://issues.redhat.com/browse/OCPBUGS-39404[OCPBUGS-39404])
3040+
3041+
3042+
[id="ocp-4-18-10-bug-fixes_{context}"]
3043+
==== Bug fixes
3044+
3045+
* Previously, the Prometheus remote-write proxy configuration was not correctly applied to the Prometheus user workload custom resource (CR), which caused communication and data collection problems in the cluster. With this release, the user workload monitoring (UWM) Prometheus configurations, including user workload Prometheus, correctly inherit the proxy settings from the cluster proxy resource. (link:https://issues.redhat.com/browse/OCPBUGS-38655[OCPBUGS-38655])
3046+
3047+
* Previously, when running {op-system-first} in an active environment, the `rpm-ostree-fix-shadow-mode.service` systemd service that used to run caused that service to fail. With this release, the `rpm-ostree-fix-shadow-mode.service` systemd service does not activate when {op-system} does not run from an installed environment. (link:https://issues.redhat.com/browse/OCPBUGS-41625[OCPBUGS-41625])
3048+
3049+
* Previously, an incorrect component import in the `SimpleSelect.tsx` file caused an undefined function `r` function in the `react-dom.production.min.js` file. This component caused error messages on the *Dashboards* and *Metrics* pages related to dropdown lists. With this release, the dropdown lists on the affected pages function correctly, eliminating the error message. (link:https://issues.redhat.com/browse/OCPBUGS-42845[OCPBUGS-42845])
3050+
3051+
* Previously, an error in the rotation logic of the image pull secret controller's secret token caused a temporary, invalid token for authentication. As a consequence, the image pull process was disrupted. With this release, the updated image pull secret controller eliminates the period when the token is not valid while the token rotates. As a result, the image pull process is smooth and continuous. (link:https://issues.redhat.com/browse/OCPBUGS-54304[OCPBUGS-54304])
3052+
3053+
* Previously, an error occurred in {hcp}-managed clusters because of the omission of the `shutdown-watch-termination-grace-period` setting in the `kube-apiserver` configuration. This error led to the unstable shutdown of applications in {hcp}-managed clusters. With this release, an update improves the shutdown process of applications in {hcp}-managed clusters, providing a grace period for the `kube-apiserver` configuration. During a shutdown, the application stability is improved and potential errors are decreased. (link:https://issues.redhat.com/browse/OCPBUGS-53404[OCPBUGS-53404])
3054+
3055+
* Previously, an issue with the version of the `github.com/sherine-k/catalog-filter` element stopped, causing instability in the mirroring process. With this release, the `github.com/sherine-k/catalog-filter` element in the `go.mod` file is updated, which solves the problem and ensures a stable and reliable mirroring process. (link:https://issues.redhat.com/browse/OCPBUGS-54727[OCPBUGS-54727])
3056+
3057+
* Previously, an iteration counter increment omission in the `scrapeCache` setting led to an incorrect series count for subsequent scrapes. As a result, monitoring was interrupted and data could potentially be lost during the Prometheus scrape process. With this release, an update ensures uninterrupted monitoring, because Prometheus continues scraping and processing data while parsing errors. (link:https://issues.redhat.com/browse/OCPBUGS-54940[OCPBUGS-54940])
3058+
3059+
[id="ocp-4-18-10-updating_{context}"]
3060+
==== Updating
3061+
To update an {product-title} 4.18 cluster to this latest release, see xref:../updating/updating_a_cluster/updating-cluster-cli.adoc#updating-cluster-cli[Updating a cluster using the CLI].
3062+
30193063
// 4.18.9
30203064
[id="ocp-4-18-9_{context}"]
30213065
=== RHSA-2025:3775 - {product-title} {product-version}.9 bug fix update and security update

0 commit comments

Comments
 (0)