NetApp/trident v26.06.0

on GitHub

Changes since v26.02.0

Trident

Fixes:

• Kubernetes: Fixed race conditions and state handling for concurrent publish/unpublish, clone, cache, and backend update operations.
• Kubernetes: Fixed export policy race conditions and concurrent publish/unpublish handling for subordinate volumes and read-only clones in ONTAP-NAS and ONTAP-NAS-Economy drivers.
• Kubernetes: Fixed ONTAP-NAS-Economy FlexVol discovery after MetroCluster failover when snapshot policies differ (Issue #1082).
• Kubernetes: Fixed ONTAP-NAS-Economy volume delete retries that interrupted long-running qtree deletes (Issue #1121).
• Kubernetes: Fixed ONTAP-SAN and ONTAP-SAN-Economy import and resize behavior, including fsType validation, volume metadata handling, and autogrow mode behavior.
• Kubernetes: Fixed NVMe/TCP namespace race condition during concurrent volume creation (Issue #1089).
• Kubernetes: Improved stale node and multipath cleanup by removing ghost multipath artifacts and stale TridentNode resources after soft-deleted node cleanup.
• Kubernetes: Fixed iSCSI resize failures when available paths and portals did not match, and improved repeated resize behavior.
• Kubernetes: Fixed ONTAP-SAN volume delete failures caused by volume busy conditions.
• Kubernetes: Fixed ONTAP REST error handling and improved SVM state timeout behavior.
• Kubernetes: Fixed GCNV SAN zonal pool and host-group handling, normalized volume identifier conversion, and normalized create requests below minimum size.
• Kubernetes: Fixed native GCNV capacity-pool selection to exclude ONTAP-mode pools.
• Kubernetes: Fixed KubeVirt failover for VMs with Trident Automated Workload Failover to speed up VM failover.
• Kubernetes: Fixed Operator-managed installation RBAC issues, including missing node ClusterRoles and incorrect ClusterRole/ClusterRoleBinding reconciliation during upgrades (applies to Operator and Helm-based installs via the Trident Operator chart).
• Kubernetes: Fixed persistent GitOps diffs on TridentConfigurator and related CRDs by omitting default priority values and correcting YAML template indentation (Issues #915, #1120).
• Kubernetes: Fixed patching of TridentVolumePublication CRDs during Helm upgrade (Issue #1151).
• Kubernetes: Fixed CSI startup ordering by starting CSI gRPC before node registration retries.
• Kubernetes: Fixed secure SMB clone configuration overlays during concurrent operations.
• Kubernetes: Fixed the issue of missing annotations in Trident controller pod after upgrade (Issue #1004).
• Kubernetes: Improved slow Qtree provisioning on large SVMs (Issue #1142).
• Kubernetes: Fixed ControllerPublish to use volume config file system type to avoid using incorrect default file system type.
• Kubernetes: Fixed an issue where orphan dm devices might cause repeated CSI NodeUnstageVolume failures.

Enhancements:

• Kubernetes: Added support for Kubernetes 1.36.
• Kubernetes: Concurrency support is now generally available (GA) for ONTAP-NAS-Economy, ONTAP-SAN-Economy, both graduated from Tech Preview, and for ANF and SolidFire drivers.
• Added ability to use Unified ONTAP balanced placement for optimal aggregate selection when provisioning new volumes (Issue #64).
• Kubernetes: Added support for ONTAP-SAN iSCSI intra-cluster volume move.
• Kubernetes: Added ONTAP AFX support for ONTAP-NAS-Flexgroup driver.
• Kubernetes: Added integration with NetApp Shift Toolkit to simplify virtual machine migrations when using Red Hat OpenShift Migration Toolkit for Virtualization.
• Kubernetes: Autogrow support for GCNV drivers (NAS and SAN) is now generally available (GA).
• Kubernetes: Trident containers' root filesystems are now mounted as read-only.
• Kubernetes: Added Cloud Identity support for the GCNV SAN driver.
• Kubernetes: Added subordinate volume support for Trident Volume Publication (TVP) propagation and autogrow workflows.
• Kubernetes: Added dynamic GCNV gRPC rate limiting.
• Kubernetes: Added controller readiness exec probe and tuned default Kubernetes API QPS/Burst settings for controller and sidecars.
• Kubernetes: Improved node operation sequencing by decoupling attach, encryption setup, and filesystem formatting steps.
• Kubernetes: Added support for volume updates and volume group snapshots when running Trident with concurrency enabled.
• Enhanced AWS ARN handling to support special regions.
• Added a best-effort snapshot removal for clone volume snapshots in ONTAP backends when the source snapshot of clone volumes is deleted.
• Kubernetes: Added backup clone PVC recognition to keep temporary PVCs for backups created by Veeam Kasten, Trilio, and Cohesity from getting stuck in the ONTAP Volume Recovery Queue.
• Kubernetes: Prevent shrinking of Flexvols during resize operation for the ONTAP-SAN-ECONOMY driver.

Experimental Enhancements:

NOTE: Not for use in production environments.

• [Tech Preview] Added support for ONTAP-NAS-Economy qtree import.
• Kubernetes: [Tech Preview] Added GCNV ONTAP-mode HTTP transport for ONTAP SAN and NAS drivers.
• Kubernetes: [Tech Preview] Added support for concurrency for ASA-R2 (SAN and NVMe) drivers.

Trident Protect

Fixes:

• Fixed cross-cluster VM restore failing with a webhook error (vrestore-creation denied: VM namespace not found) when the source namespace does not exist on the target cluster.
• Fixed selective VM in-place restore failing at CreateApplication due to a spec mismatch on an existing Application; the Application is now reused.
• Fixed the IPR job hanging in Running when both the VM and its PVC are deleted; a deleted PVC during IPR is now treated as a fatal error.
• Fixed an immediate backup/snapshot being triggered when editing an existing schedule by correcting the scheduled timestamp for overdue runs.
• Fixed the Snapshot finalizer attempting to fetch the Application CR even after it was deleted; it now skips when the Application is not found.
• Fixed protected capacity increasing drastically after IPR by using the trident-protect-utils pvc-copy image with sparse copy (cp --sparse).
• Fixed Trident Protect deleting the Kubernetes snapshot object without confirming ONTAP deletion; the finalizer now waits for VolumeSnapshot deletion.
• Fixed snapshots not continuing when a VM freeze fails; freeze/unfreeze now continues on failure.
• Fixed KubeVirt VM apps recreating PVCs during AMR reverse / reverse-resync / resync operations.

Enhancements:

• Made the temp-volume splitOnClone annotation configurable for Kopia backups.
• Added StorageClass mapping support to the Neptune CLI.
• Added includedVirtualMachines support for Restore CRs in protectctl.
• Added a SkipApplicationCreation field on Restore and implemented conditional Application CR creation.
• Added StorageClass mapping to AppMirrorRelationship (AMR).
• Propagated all Job pod statuses and events to their parent CRs.
• Added resource transformations support to all Restore CRs.
• Surfaced resource-restore failure messages from the pod into the Restore CR status.
• Volume import now uses the source request size (to be reverted once fixed upstream).