Releases · longhorn/longhorn

22 Jan 07:50

derekbit

v1.8.0

7c7c779

Longhorn v1.8.0 Latest

Latest

Longhorn v1.8.0 Release Notes

This latest version of Longhorn introduces several features, enhancements, and bug fixes that are intended to improve system quality and the overall user experience. Highlights include new V2 Data Engine features, multiple backupstores, automatic RWX volume expansion, installation and upgrades via Helm Controller, and V2 Data Engine Support for Talos Linux.

The Longhorn team appreciates your contributions and anticipates receiving feedback regarding this release.

For more information about release-related terminology, see Releases.

Important

The CSI external-snapshotter was upgraded to v8.2.0. Ensure that all clusters are running Kubernetes v1.25 or later before upgrading to Longhorn v1.8.0 or a later version.

Deprecation & Incompatibilities

The default block size for block-type disks in earlier Longhorn releases is 4096 bytes. However, 512 bytes is more commonly used and aligns with the V1 Data Engine's configuration. Additionally, the 4096-byte block size is incompatible with backing images generated by the V1 Data Engine. To address these concerns, the default block size was changed to 512 bytes.

If you have existing V2 volumes, perform the following steps:
1. Back up the V2 volumes.
2. Remove the V2 volumes.
3. Delete the block-type disk with a 4096-byte block size from node.spec.disks.
4. Erase the old data on the block-type disk using tools such as dd.
5. Add the disk again to node.spec.disks with the updated configuration.
6. Restore the V2 volumes.
For more information, see #10053.
A V2 volume data corruption issue that affects earlier Longhorn releases has been resolved in v1.8.0. The issue involves potential continual changes to the checksum of files in a V2 volume with multiple replicas. This occurs because SPDK allocates clusters without initialization, leading to data inconsistencies across replicas. The varying data read from the volume can result in data corruption and broken backups. For more information, see #10035.

Primary Highlights

New V2 Data Engine Features

Although the V2 Data Engine is still considered an experimental feature in this release, the core functions have been significantly enhanced.

Configurable CPU cores: Support the global and node-specific configuration options provide greater control and flexibility for optimizing performance and resource allocation.
Disaster recovery (DR) volumes: Designed to store data in a backup cluster in the event of a failure in the main cluster. DR volumes enhance the resiliency of Longhorn volumes by ensuring data can be quickly restored in case of cluster outages.
Auto-salvage volumes: Automatically repair volumes in the event of a failure.
Live migration: Allow for the migration of volumes from one node to another without interrupting the services using those volumes.
Volume encryption: Ensure data stored in volumes is protected through encryption in transit and at rest.
Delta replica rebuilding using snapshot checksum.
Backing image update and download.

Multiple Backupstores and Default Backup Target

Starting with v1.8.0, Longhorn allows you to use multiple backupstores for storing backups of Longhorn volumes. Longhorn v1.8.0 also creates a default backup target (default) during installation and upgrades. The default backup target is used for system backups and volumes that were created without an assigned backup target name.

Demo | Documentation | GitHub Issue

Automatic RWX Volume Expansion

Longhorn v1.8.0 supports fully automatic online expansion of RWX volumes without the need to scale down the workload or apply manual commands. To use this feature, ensure that the v1.8.0 versions of Longhorn Manager, Share Manager, and the CSI plugin are all running.

Documentation | GitHub Issue

Helm Controller

You can now install and upgrade Longhorn on clusters running RKE2 or K3s using the Helm Controller that is built into those distributions. The Helm Controller manages Helm charts using a HelmChart Custom Resource Definition (CRD), which contains most of the options that would normally be passed to the Helm command-line tool.

Documentation | GitHub Issue

V2 Data Engine Support for Talos Linux

Longhorn v1.8.0 supports usage of V2 volumes in Talos Linux clusters. To use this feature, ensure that all nodes meet the V2 Data Engine prerequisites.

Documentation | GitHub Issue

Installation

Important

Ensure that your cluster is running Kubernetes v1.25 or later before installing Longhorn v1.8.0.

You can install Longhorn using a variety of tools, including Rancher, Kubectl, and Helm. For more information about installation methods and requirements, see Quick Installation in the Longhorn documentation.

Upgrade

Important

Ensure that your cluster is running Kubernetes v1.25 or later before upgrading from Longhorn v1.7.x to v1.8.0.

Longhorn only allows upgrades from supported versions. For more information about upgrade paths and procedures, see Upgrade in the Longhorn documentation.

Post-Release Known Issues

For information about issues identified after this release, see Release-Known-Issues.

Highlight

[FEATURE] Support encrypted v2 volumes 7355 - @mantissahz @chriscchien
[FEATURE] Support v2 volume delta replica rebuilding based on snapshot checksum: data plane and control plane 9488 - @shuo-wu @roger-ryao
[FEATURE] Multiple backup stores support 5411 - @mantissahz @roger-ryao
[FEATURE] Support v2 volume delta replica rebuilding based on snapshot checksum: SPDK part 5573 - @DamiaSan @roger-ryao
[IMPROVEMENT] v2 volume supports data locality 9371 - @derekbit @roger-ryao
[FEATURE] SPDK volume supports live migration 6361 - @PhanLe1010 @chriscchien @roger-ryao
[FEATURE] Support backing image for v2 volume 6341 - @yangchiu @ChanYiLin @chriscchien
[FEATURE] Make CPU core configurable for v2 data engine 8835 - @yangchiu @derekbit
[FEATURE] Support/verify installing Longhorn Helm chart using helm-controller which is built in k3s and rke2 9506 - @yangchiu @james-munson
[FEATURE] v2 volume supports auto salvage 8430 - @c3y1huang @chriscchien
[FEATURE] v2 volumes supports DR volume 6613 - @c3y1huang @roger-ryao
[FEATURE] Talos support for v2 data engine 7791 - @yangchiu @c3y1huang

Feature

[FEATURE] Support Data Engine Option For Backing Image in CSI flow 10084 - @ChanYiLin @roger-ryao
[FEATURE] Missing option to modify backup-task retain value from UI while editing recurring job 8810 - @houhoucoop @roger-ryao
[UI][FEATURE] Multiple backup stores support 8647 - @a110605 @roger-ryao
[UI][FEATURE] Support V2 Backing Image in UI 9880 - @yangchiu @chriscchien @houhoucoop
[FEATURE] Add periodic HugePages (2Mi) configuration check to ensure v2 data engine compatibility 9983 - @yangchiu @jangseon-ryu
[FEATURE][UI] Fill in the secret and secret namespace when restoring the BackupBackingImage 9490 - @chriscchien @houhoucoop
[FEATURE] Add kernel module check for data engine v2 9915 - @jangseon-ryu @roger-ryao
[FEATURE]: Automatic online filesystem resize for RWX volumes. 9736 - @yangchiu @james-munson
[FEATURE] Add recurring job label to backup metrics or empty value 9429 - @ChanYiLin @roger-ryao
[FEATURE] Longhorn CLI supports darwin platform 9532 - @derekbit @chriscchien

Improvement

[IMPROVEMENT] ...

Contributors

COLDTURNIP, tserong, and 26 other contributors

Assets 7

20 Jan 08:41

derekbit

v1.8.0-rc5

d86e2e7

Longhorn v1.8.0-rc5 Pre-release

Pre-release

DON'T UPGRADE from/to any RC/Preview/Sprint releases because the operation is not supported.