RHSA-2026:65343 - Security Advisory

Topic

Red Hat OpenShift distributed tracing platform (Tempo) 3.11.0 has been released

Description

This release of the Red Hat OpenShift distributed tracing platform (Tempo) provides new features, security improvements, and bug fixes.

Breaking changes:

• High availability defaults for TempoStack deployment sizes: Before this update, only the ingester component was scaled for high availability while all other components defaulted to a single replica, leaving the overall ingest and query paths vulnerable to single points of failure. With this update, when you set a non-demo deployment size (1x.pico, 1x.extra-small, 1x.small, or 1x.medium) in the spec.size field of the TempoStack custom resource, the Operator defaults each component to at least 2 replicas, and scales throughput-bound components at larger sizes. Watch your resource consumption, as running more replicas per component increases the total CPU and memory consumed by the deployment. Explicit per-component replicas values always take precedence. For more information, see https://redhat.atlassian.net/browse/TRACING-6169.

Deprecations:

• Jaeger Query is deprecated: The Jaeger Query component in the Tempo Operator is deprecated and will be removed in a future release. The Tempo Operator emits a warning when the Jaeger Query feature is enabled. To visualize traces, use the distributed tracing UI plugin as the recommended replacement for the Jaeger UI. For more information, see https://redhat.atlassian.net/browse/TRACING-6509.

Technology Preview features:

• None

Enhancements:

• High availability for TempoStack deployments: You can use the spec.replicationZones field in the TempoStack custom resource to define zone-aware replication. By using this feature, you can deploy TempoStack instances with higher resilience by spreading replicas across topology zones and ensuring minimum pod availability during disruptions. For more information, see https://redhat.atlassian.net/browse/TRACING-6168.
• Service name and namespace auto-complete with query RBAC enabled: When you enable query Role-Based Access Control (RBAC) in a multi-tenant TempoStack deployment, the Tempo gateway supports auto-complete for non-sensitive attributes such as service name and namespace. Before this update, all tag value API endpoints were blocked when query RBAC was enabled, which prevented the Service Name drop-down filter in the distributed tracing console plugin from displaying results. With this update, the gateway allows the search tags API for non-sensitive attributes while still protecting sensitive span data. As a result, the Service Name filter and other non-sensitive attribute drop-down list items work correctly in multi-tenant RBAC mode. For more information, see https://redhat.atlassian.net/browse/TRACING-6485.

Bug fixes:

• TempoStack and TempoMonolithic status conditions accurately reflect pod readiness: Before this update, the status field of the TempoStack or TempoMonolithic custom resource might show Ready=false even when all pods were ready, or vice versa. With this update, the Operator watches Deployment and StatefulSet status changes, and the status conditions accurately reflect actual pod readiness. Deleted pods are removed from the pod status map. As a result, the status field reliably indicates the current state of the deployment. For more information, see https://redhat.atlassian.net/browse/TRACING-6453.
• Gateway redirect URL respects the custom ingress hostname: Before this update, setting a custom hostname with the spec.template.gateway.ingress.host field in the TempoStack custom resource did not update the OpenShift OAuth redirect URL, so authentication redirects pointed to the wrong URL. With this update, the Operator uses the custom hostname when generating the redirect URL, and gateway authentication works correctly with custom ingress hostnames. For more information, see https://redhat.atlassian.net/browse/TRACING-6435.
• Gateway starts on clusters that use external OIDC authentication: Before this update, on clusters using external OpenID Connect (OIDC) authentication instead of the built-in OpenShift OAuth server, the Tempo gateway container failed to start because it could not auto-discover the OpenShift OAuth endpoints. With this update, the gateway handles external authentication and starts correctly on such clusters. For more information, see https://redhat.atlassian.net/browse/TRACING-6646.

Known issues:

• None

Solution

For details on how to apply this update, refer to:

https://docs.redhat.com/en/documentation/openshift_container_platform/latest/html/operators/administrator-tasks#olm-upgrading-operators

Affected Products

• Red Hat OpenShift distributed tracing

Fixes

• TRACING-6509 - Add a deprecated warning to the Tempo operator webhook when jaeger query is enabled
• TRACING-6167 - Support PodDisruptionBudget in TempoStack
• TRACING-6168 - Support topologySpreadConstraints in TempoStack
• TRACING-6169 - Support HA deployments for TempoStack
• TRACING-6485 - Support service name and namespace auto-complete with query RBAC enabled
• TRACING-6453 - status field can become stale
• TRACING-6631 - Improve Tempo status reporting
• TRACING-6636 - Improve pod rollout by avoiding spec changes after certificate refresh
• TRACING-6520 - The Tempo gateway's query RBAC middleware (`WithTraceQLNamespaceSelectAndForbidOtherAPIs`) blocks the `/api/v2/search/tag/{tag}/values` and `/api/v2/search/tags` endpoints with HTTP 403
• TRACING-6435 - Tempo Operator is not generating correct redirectURL
• TRACING-6646 - Tempo Gateway container fails when OCP BYO is enabled

CVEs

• CVE-2026-32283
• CVE-2026-33818
• CVE-2026-41178
• CVE-2026-42502
• CVE-2026-48586
• CVE-2026-55969
• CVE-2026-56852
• CVE-2026-56853
• CVE-2026-56858
• CVE-2026-56859
• CVE-2026-56860
• CVE-2026-56862

References

• https://access.redhat.com/security/updates/classification/
• https://docs.redhat.com/en/documentation/openshift_container_platform/latest/html/distributed_tracing/distributed-tracing-platform-tempo