v2.10.2 - 30/Jun/2026

Release Notes: SUSE® Observability Helm Chart v2.10.2

Component Versions

Component Version

SUSE Observability

2.10.2

SUSE Observability Agent

1.5.10

New Features & Enhancements

  • Metrics Explorer Screen Expansion (/#/metrics): The metrics explorer screen now natively supports the full suite of metrics visualization charts:

    • Line Charts, Bar Charts, Gauges, and Stat panels.

    • Each visualization supports an extensive range of visual layouts and backend data configuration settings.

    • Users can now directly convert any configured metrics explorer chart into a permanent dashboard widget using the new Add to dashboard button.

  • Metrics Preview Drawer Enhancements: The preview drawer now comprehensively supports Line, Bar, Gauge, and Stat charts, fully respecting all custom chart settings. A new Explore link safely redirects users to the full metrics explorer page with all settings and related PromQL queries loaded for further editing.

  • Stat and Gauge Pagination: Added pagination support to Stat and Gauge charts to provide a smoother, more responsive user experience when rendering large time-series collections.

  • Agent Default Namespace Change: The default Kubernetes namespace for fresh SUSE Observability Agent installations has been updated from suse-observability to suse-observability-agent.

  • Standardized OCI Labels: All SUSE Observability container images now bundle a standardized, common set of OCI labels for enhanced compliance and metadata tracking.

  • Overview Table Clarity: Missing instant metrics results within the overview table now cleanly display an explicit n/a (not available) indicator instead of blank spaces.

  • Cluster naming It is now possible to choose a different a cluster name than the one used in rancher, including the local cluster.

Bug Fixes

  • Platform Stability & Data Integrity:

    • StackGraph Resiliency: Fixed a critical issue where StackGraph data could become corrupted during a Transaction Manager shutdown or a sudden loss of network connectivity.

    • Tephra Network Partitioning: Fixed a bug where the Tephra state could become corrupted following a prolonged network partition.

    • Restore Job Retries: Resolved an issue where a failed restore job would unexpectedly retry indefinitely.

  • Sizing Profile Fixes: Fixed a configuration bug where the resources for larger deployment profiles (specifically 4000-ha) were not being allocated or scaled correctly.

  • Rancher extension

    • PV/PVCs Persistent Volumes and Persistent Volume Claims now get a health state.

    • Onboarding clusters Clusters which do not have the agent installed now show UNOBSERVED rather than CONNECTION ERROR.

  • Metrics & Tracing Corrections:

    • OpenTelemetry Service Metrics: Corrected the metrics displayed in the trace performance tabs for OpenTelemetry services and service instances. This fixes an issue where span durations were completely missing or where client and server metrics were accidentally swapped.

    • ClickHouse Backup Monitoring: Fixed node-agent scraping errors targeting the ClickHouse backup metrics endpoints.

  • UI & Monitor Refinements:

    • Node Readiness Alerts: Fixed an issue where the node readiness monitor failed to fire under alert conditions.

    • Duplicate Monitors: Resolved a bug causing the monitors list on component highlight pages to generate duplicate rows over time.

    • Conditional Dashboard Menu: The "Dashboards" menu item is now hidden for users who do not possess the explicit RBAC permissions required to view or use dashboards.

Agent Bug Fixes

  • OS & Kernel Compatibility:

    • RHEL 8.10 Container Metrics: Fixed an issue where container metrics failed to register or display on Red Hat Enterprise Linux (RHEL) 8.10 environments.

    • SELinux Log Shipping: Resolved an issue where the agent was blocked from shipping system logs on clusters with SELinux enabled in enforcing mode.

  • Disk Metrics Restoration: Fixed a bug that entirely blocked disk usage metrics from being transmitted from the agent to the platform.