Important Notes

With efficient cloning enabled, a newly cloned and detached volume is degraded and has only one replica, with its clone status set to copy-completed-awaiting-healthy. To bring the volume to a healthy state, transition the clone status to completed and rebuild the remaining replica by either enabling offline replica rebuilding or attaching the volume to trigger replica rebuilding. See Issue #12341 and Issue #12328.

Due to the upgrade of the CSI external snapshotter to v8.2.0, you must be running Kubernetes v1.25 or later to upgrade to SUSE Storage v1.8.0 or a newer version.

When you upgrade with Helm or the Rancher App Marketplace, SUSE Storage performs pre-upgrade checks. If a check fails, the upgrade stops and the reason for the failure is recorded in an event.

For more details, see Upgrading Longhorn Manager.

Automated pre-upgrade checks does not cover all scenarios. A manual check is recommended using kubectl or the SUSE Storage UI.

SUSE Storage v1.11.0 introduces the manager-url setting that allows explicit configuration of the external URL for accessing the Longhorn Manager API.

Background: When Longhorn Manager is accessed through Ingress or Gateway API HTTPRoute, API responses may contain internal cluster IPs (for example, 10.42.x.x:9500) in the actions and links fields. This occurs when the ingress controller does not properly set X-Forwarded-* headers, causing the API to fall back to the internal pod IP.

Solution: Configure the manager-url setting with your external URL (for example, https://longhorn.example.com). The Manager injects proper forwarded headers to ensure API responses contain correct external URLs.

Configuration:

For more details, see Manager URL.

SUSE Storage v1.11.0 introduces built-in support for Gateway API HTTPRoute as a modern alternative to Ingress for exposing the SUSE Storage UI.

For detailed setup instructions, prerequisites and advanced configuration, see Create an HTTPRoute with Gateway API.

SUSE Storage v1.11.0 introduces the Snapshot Heavy Task Concurrent Limit to prevent disk exhaustion and resource contention. This setting limits concurrent heavy operations—such as snapshot purge and clone—per node by queuing additional tasks until ongoing ones complete. By controlling these processes, the system reduces the risk of storage spikes typically triggered by snapshot merges.

For further details, refer to Snapshot Heavy Task Concurrent Limit and Issue #11635.

To improve data distribution and resource utilization, SUSE Storage introduces a balance algorithm that schedules replicas evenly across nodes and disks based on calculated balance scores.

For more information, see Scheduling.

SUSE Storage CSI now supports StorageClass allowedTopologies, enabling Kubernetes to automatically restrict pod and volume scheduling to nodes where SUSE Storage is available.

For more information, see Issue #12261 and Storage Class Parameters.

Starting with SUSE Storage v1.11.0, disk health monitoring is available for both the V1 and V2 data engines. SUSE Storage collects disk health data and exposes it through Prometheus metrics and the Node custom resources.

For more information, see Disk health monitoring.

SUSE Storage v1.11.0 introduces support for the ReadWriteOncePod (RWOP) access mode, addressing the need for stricter single-pod volume access guarantees in stateful workloads. Unlike ReadWriteOnce (RWO), which permits multiple pods on the same node to mount a volume, RWOP ensures that only one pod across the entire cluster can access the volume at any given time. This capability is particularly valuable for stateful applications requiring exclusive write access, such as databases or other workloads where concurrent access could lead to data corruption or consistency issues.

For more information, see Access Modes and Issue #9727.

Starting with SUSE Storage v1.11.0, a new scale replica rebuilding feature allows a rebuilding replica to fetch snapshot data from multiple healthy replicas concurrently, potentially improving rebuild performance.

For more information, see Scale Replica Rebuilding.

Starting with SUSE Storage v1.11.0, the Offline Replica Rebuilding setting is updated from a data engine-specific setting to a global setting. Previously, users could configure offline replica rebuilding separately for v1 and v2 data engines. During the upgrade to v1.11.0, SUSE Storage automatically checks the existing configuration. If offline replica rebuilding is enabled for either the v1 or v2 data engine, the new global setting defaults to true. Otherwise, it remains disabled (false).

For more information, see Offline Replica Rebuilding Setting.

SUSE Storage v1.11.0 enhances the Longhorn CLI preflight install and check behavior. When /etc/os-release does not match a known distribution, the CLI attempts to detect a supported package manager and continues in a compatibility mode.

For more information, see Issue #12153.

Live upgrades of V2 volumes are not supported. Before you upgrade, make sure all V2 volumes are detached.

The V2 Data Engine is a Technical Preview feature in SUSE Storage v1.11.0.

It is nearly complete, with no significant functional changes expected, and has been validated in controlled environments. Users should evaluate the feature thoroughly before enabling it in production.

Starting with SUSE Storage v1.11.0, the SPDK UBLK front-end exposes performance-tuning parameters that can be configured globally or per-volume:

These parameters can be configured:

For more information, see Issue #11039.