RHBA-2021:0235 - Bug Fix Advisory

Topic

Red Hat OpenShift Container Platform release 4.6.15 is now available with
updates to packages and images that fix several bugs.

Description

Red Hat OpenShift Container Platform is Red Hat's cloud computing
Kubernetes application platform solution designed for on-premise or private
cloud deployments.

This advisory contains the container images for Red Hat OpenShift Container
Platform 4.6.15. See the following advisory for the RPM packages for this
release:

https://access.redhat.com/errata/RHBA-2021:0237

Space precludes documenting all of the container images in this advisory.
See the following Release Notes documentation, which will be updated
shortly for this release, for details about these changes:

https://docs.openshift.com/container-platform/4.6/release_notes/ocp-4-6-release-notes.html

This update fixes the following bugs among others:

• When installing on AWS previously, the Cloud Credential Operator automatically created an OpenID discovery endpoint that could be used to externally authenticate service account tokens obtained with the TokenRequest API. This involved changing the `serviceAccountIssuer` on the kube-apiserver, which invalidated any previously minted tokens. This functionality has been removed. Creation and configuration of the OpenID discovery endpoint and `serviceAccountIssuer` now must be done by the cluster admin at cluster installation time. The `serviceAccountIssuer` can no longer be changed after cluster installation begins. (BZ#1907613)
• Previously, the log level for the kube-rbac-proxy was set to level 10. This caused excessive detail in the log messages for the kube-rbac-proxy. The log level for the kube-rbac-proxy has been decreased to 3 to match other components, resulting in more appropriate logging detail for most scenarios. (BZ#1908316)
• Previously, a missing guard in the Red Hat OpenStack Platform (RHOSP) validations could have led to fetching the list of subnets with an empty subnet ID, and caused some non-RHOSP clouds to return unexpected values. The unexpected error code would fail validation and prevent a cluster installation on these non-RHOSP clouds. This has been fixed by adding the missing guard against empty subnet ID. This validation now returns the expected result, allowing a cluster to be installed on more clouds. (BZ#1913813)

You may download the oc tool and use it to inspect release image metadata as follows:

(For x86_64 architecture)

$ oc adm release info quay.io/openshift-release-dev/ocp-release:4.6.15-x86_64

The image digest is sha256:b70f550e3fa94af2f7d60a3437ec0275194db36f2dc49991da2336fe21e2824c

(For s390x architecture)

$ oc adm release info quay.io/openshift-release-dev/ocp-release:4.6.15-s390x

The image digest is sha256:4f113aa813f96701ea62474bcaa5545451a9cbfbdf3437fa9c24f8a8ac288010

(For ppc64le architecture)

$ oc adm release info quay.io/openshift-release-dev/ocp-release:4.6.15-ppc64le

The image digest is sha256:ef0937ce67e3cc8eafbb613a90ea4434330ae2b2546d0291c164242ef87d67a8

All OpenShift Container Platform 4.6 users are advised to upgrade to these
updated packages and images when they are available in the appropriate
release channel. To check for available updates, use the OpenShift Console
or the CLI oc command. Instructions for upgrading a cluster are available
at
https://docs.openshift.com/container-platform/4.6/updating/updating-cluster-between-minor.html#understanding-upgrade-channels_updating-cluster-between-minor.

Solution

For OpenShift Container Platform 4.6 see the following documentation, which
will be updated shortly for this release, for important instructions on how
to upgrade your cluster and fully apply this asynchronous errata update:

https://docs.openshift.com/container-platform/4.6/release_notes/ocp-4-6-release-notes.html

Details on how to access this content are available at
https://docs.openshift.com/container-platform/4.6/updating/updating-cluster-cli.html.

Affected Products

• Red Hat OpenShift Container Platform 4.6 for RHEL 8 x86_64
• Red Hat OpenShift Container Platform for Power 4.6 for RHEL 8 ppc64le
• Red Hat OpenShift Container Platform for IBM Z and LinuxONE 4.6 for RHEL 8 s390x

Fixes

• BZ - 1886062 - Logging may be broken due to mix of k8s.io/klog v1 and v2
• BZ - 1898160 - [OVN] EgressIP cannot fail over to available nodes after one egressIP node shutdown
• BZ - 1907461 - Init containers restart when the exited container is removed from node.
• BZ - 1907613 - oidcdiscoveryendpoint controller invalidates all TokenRequest API tokens during install
• BZ - 1908316 - kube-rbac-proxy exposes tokens, has excessive verbosity
• BZ - 1908463 - OLM continuously printing debug logs
• BZ - 1913099 - Ingress VIP assigned to two infra nodes simultaneously - keepalived process running in pods seems to fail
• BZ - 1913813 - Validation is not robust enough and may prevent to generate install-confing.
• BZ - 1914910 - Some pods are stuck in ContainerCreating and some sdn pods are in CrashLoopBackOff
• BZ - 1915247 - AggregatedAPIDown permanently firing after removing APIService
• BZ - 1915583 - p&f: carry upstream pr 97860
• BZ - 1915905 - When the marketplace-operator pod get's restarted, the custom catalogsources are gone, as well as the pods
• BZ - 1916112 - [Assisted-4.6 ]Read only admin cannot fetch openshift_versions
• BZ - 1916335 - [sig-storage] Managed cluster should have no crashlooping recycler pods over four minutes
• BZ - 1917411 - cluster will not gracefully recover if openshift-etcd namespace is removed
• BZ - 1917937 - upgrade version of dnsmasq package
• BZ - 1918371 - Too many haproxy processes in default-router pod causing high load average

CVEs

• CVE-2020-25681
• CVE-2020-25682
• CVE-2020-25683
• CVE-2020-25684
• CVE-2020-25685
• CVE-2020-25686
• CVE-2020-25687
• CVE-2021-20182

References

(none)