发布日期: 2025-02-17
版本号: v1.7.3-rc2

此版本为Longhorn v1.7.3，重点提示不要升级至或从任何RC/Preview/Sprint版本，因操作不受支持。主要更新包括：新增HugePages配置检查以确保v2数据引擎兼容性；改进日志格式配置、设置变更验证逻辑、UI界面显示优化、存储调度策略增强；修复了多个关键问题，如CSI插件异常导致数据丢失、节点排空阻塞、备份超时清理失效、卷扩容卡顿、RWX卷创建限制等。同时修复了安全漏洞并更新组件依赖。此版本涉及功能增强、稳定性提升及用户体验优化，由多位开发者协作完成。

更新内容 （中文）

请勿从/向任何 RC/Preview/Sprint 版本升级，因为该操作不受支持。

本版本已解决的问题

新功能

  • [BACKPORT][v1.7.3][功能] 添加周期性 HugePages (2Mi) 配置检查以确保 v2 数据引擎兼容性 10029 - @yangchiu

改进

  • [BACKPORT][v1.7.3][改进] 设置变更验证应恢复使用卷状态判断"所有卷是否已分离" 10375 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][改进] 在 Longhorn 组件(UI、驱动)中支持 JSON 日志格式配置 10082 - @chriscchien
  • [BACKPORT][v1.7.3][改进] 记录实例管理器 Pod 即将被删除的原因 9887 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][改进] 为何无法在 v2 longhorn 卷中修改副本数？9806 - @chriscchien
  • [BACKPORT][v1.7.3][改进] 改为检查 /etc/nfsmount.conf 中的 NFS 版本 9831 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][改进] 在 CLI 中添加 dmsetup 和 dmcrypt 工具检查 9935 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][UI][改进] 优化 UI 上的卷大小信息显示 9964 - @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][改进] 防止卷扩容卡死 9914 - @c3y1huang @roger-ryao
  • [BACKPORT][v1.7.3][改进][UI] 在备份页面添加带标签页的 backupBackingImage 表格 9970 - @houhoucoop
  • [BACKPORT][v1.7.3][改进] 拒绝 strict-local + RWX 卷的创建 9930 - @COLDTURNIP @yangchiu
  • [BACKPORT][v1.7.3][改进] 通过 longhorn manager 设置配置其他系统和用户管理组件的日志级别 9617 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][改进] 将容易引起误解的错误信息降级为警告 9917 - @yangchiu @derekbit
  • [BACKPORT][v1.7.3][改进] 优化 longhorn-manager 构建时间 9693 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][改进] 在 longhorn manager 中增加 Talos 系统的环境检查支持 9723 - @yangchiu @c3y1huang

Bug 修复

  • [BACKPORT][v1.7.3][BUG] 在罕见竞态条件下 Longhorn CSI 插件执行错误文件系统格式化操作导致数据丢失 10417 - @yangchiu @PhanLe1010 @chriscchien
  • [BACKPORT][v1.7.3][BUG] kubectl drain node 因残留引擎进程阻塞 10427 - @yangchiu @PhanLe1010
  • [BACKPORT][v1.7.3][BUG] 测试用例 test_csi_mount_volume_online_expansion 因无法扩展 PVC 失败 10413 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] 备份操作耗时过长时工作负载 Pod 无法迁移至新节点 10173 - @yangchiu
  • [BUG][v1.7.x] RWX 卷/ganesha.nfsd 导致内存消耗过高 8523 - @james-munson @chriscchien
  • [BACKPORT][v1.7.3][BUG] WebUI 卷列表异常消失与重现 10331 - @PhanLe1010 @chriscchien @houhoucoop
  • [BACKPORT][v1.7.3][BUG] 每次指标收集时出现来自 metrics_collector.(*BackupCollector).Collect 的 “Error get size” 10362 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][BUG] 引擎卡在 “stopped” 状态导致卷无法挂载 9954 - @ChanYiLin @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Helm chart 中备份执行超时设置问题 10326 - @james-munson @chriscchien
  • [BACKPORT][v1.7.3][BUG] 断电后系统不稳定 10185 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][BUG] 通过 CSI 快照创建备份时，备份卷出现前 CSI 插件 Pod 持续崩溃 10024 - @mantissahz @chriscchien
  • [BACKPORT][v1.7.3][BUG] 节点临时关闭后出现存储不足；新增副本预检失败 10223 - @PhanLe1010 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 长期调用 rpm -q 导致主机 rpm 数据库崩溃 10022 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 备份进度不应将上传失败的块计入成功计数 9793 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][BUG][v1.8.x] 无法创建备份，备份立即进入错误状态 10180 - @PhanLe1010 @chriscchien
  • [BACKPORT][v1.7.3][BUG] v1.7.2 中存储无法重新调度 10109 - @PhanLe1010
  • [BACKPORT][v1.7.3][BUG] 超时后旧备份未被清理 9731 - @mantissahz @roger-ryao
  • [BACKPORT][v1.7.3][BUG] RHEL 系统显示未知操作系统信息 9833 - @yangchiu @mantissahz @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 卷挂载失败 - 输入/输出错误 10005 - @PhanLe1010 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 无法通过 UI 删除 backing image 备份 10068 - @chriscchien @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 卷备份详情页显示错误通知 10071 - @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 从备份列表恢复多个文件时 API 请求缺少 fromBackup 参数 10051 - @a110605 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Webhook 服务器初始化阻塞 longhorn-manager 运行 10055 - @c3y1huang @chriscchien
  • [BACKPORT][v1.7.3][BUG] CLI 预检检查忽略 NFS 安装缺失问题 9893 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 节点驱逐期间分离的卷卡在已附加状态 9810 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] 测试用例 test_node_eviction_multiple_volume 在卷分离后无法重新调度副本 9866 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] 增量恢复期间节点启停导致 DR 卷重新挂载失败并故障 9803 - @c3y1huang @roger-ryao
  • [BACKPORT][v1.7.3][BUG] 共享管理器节点关机导致 RWX PVC 无法挂载至新节点 9856 -
  • [BACKPORT][v1.7.3][BUG] RWX PVC 文件系统扩容失败 9738 - @james-munson
  • [BACKPORT][v1.7.3][BUG] 带选项的 NFS 备份目标 URL 导致备份镜像信息检查失败 9703 - @yangchiu @mantissahz
  • [BACKPORT][v1.7.3][BUG] 升级前检查 Pod 应报告失败原因 9643 - @yangchiu @james-munson

其他

  • [任务] 修复 v1.7.3 的 CVE 问题 9897 - @c3y1huang
  • [BACKPORT][v1.7.3][任务] 构建时安装最新版 grpc_health_probe 9715 - @yangchiu @c3y1huang

贡献者

  • @COLDTURNIP
  • @ChanYiLin
  • @PhanLe1010
  • @a110605
  • @c3y1huang
  • @chriscchien
  • @derekbit
  • @houhoucoop
  • @innobead
  • @james-munson
  • @mantissahz
  • @roger-ryao
  • @yangchiu

更新内容 （原始）

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

Resolved Issues in this release

Feature

  • [BACKPORT][v1.7.3][FEATURE] Add periodic HugePages (2Mi) configuration check to ensure v2 data engine compatibility 10029 - @yangchiu

Improvement

  • [BACKPORT][v1.7.3][IMPROVEMENT] Settings change validation should go back to using Volume state to determine “are all volumes detached” 10375 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][IMPROVEMENT] Add support for JSON log format configuration in Longhorn components (UI, driver) 10082 - @chriscchien
  • [BACKPORT][v1.7.3][IMPROVEMENT] Logging the reason why the instance manager pod is going to be deleted. 9887 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][IMPROVEMENT] Why is it not possible to change the replica count in v2 longhorn volume? 9806 - @chriscchien
  • [BACKPORT][v1.7.3][IMPROVEMENT] Check NFS versions in /etc/nfsmount.conf instead 9831 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][IMPROVEMENT] Add dmsetup and dmcrypt utilities check in cli 9935 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][UI][IMPROVEMENT] Improve the volume size information on UI 9964 - @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][IMPROVEMENT] Prevent Volume Resize Stuck 9914 - @c3y1huang @roger-ryao
  • [BACKPORT][v1.7.3][IMPROVEMENT][UI] Add backupBackingImage table in backup page with tabs 9970 - @houhoucoop
  • [BACKPORT][v1.7.3][IMPROVEMENT] Reject strict-local + RWX volume creation 9930 - @COLDTURNIP @yangchiu
  • [BACKPORT][v1.7.3][IMPROVEMENT] Configure the log level of other system and user managed components via longhorn manager setting 9617 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][IMPROVEMENT] Change confusing error message to warning level 9917 - @yangchiu @derekbit
  • [BACKPORT][v1.7.3][IMPROVEMENT] Building longhorn-manager takes long time 9693 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][IMPROVEMENT] Talos support for environment check in longhorn manager 9723 - @yangchiu @c3y1huang

Bug

  • [BACKPORT][v1.7.3][BUG] Data lost caused by Longhorn CSI plugin doing a wrong filesystem format action in a rare race condition 10417 - @yangchiu @PhanLe1010 @chriscchien
  • [BACKPORT][v1.7.3][BUG] kubectl drain node is blocked by unexpected orphan engine processes 10427 - @yangchiu @PhanLe1010
  • [BACKPORT][v1.7.3][BUG] Test case test_csi_mount_volume_online_expansion is failing due to unable to expand PVC 10413 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] Workload pod will not be able to move to new node when backup operation is taking a long time 10173 - @yangchiu
  • [BUG][v1.7.x] Excessive memory consumption caused by RWX volumes / ganesha.nfsd 8523 - @james-munson @chriscchien
  • [BACKPORT][v1.7.3][BUG] WebUI Volumes Disappear and Reappear 10331 - @PhanLe1010 @chriscchien @houhoucoop
  • [BACKPORT][v1.7.3][BUG] “Error get size” from “metrics_collector.(*BackupCollector).Collect” on every metric scrape 10362 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][BUG] Engine stuck in “stopped” state, prevent volume attach 9954 - @ChanYiLin @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Backup Execution Timeout setting issue in Helm chart 10326 - @james-munson @chriscchien
  • [BACKPORT][v1.7.3][BUG] Instability after power failure 10185 - @yangchiu @james-munson
  • [BACKPORT][v1.7.3][BUG] CSI plugin pod keep crashing util the backup volume appears when creation a backup via the CSI snapshotter 10024 - @mantissahz @chriscchien
  • [BACKPORT][v1.7.3][BUG] insufficient storage;precheck new replica failed after a temporary shutdown of a node 10223 - @PhanLe1010 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] longhorn-manager seems to crash rpm-DB on the host by continously calling rpm -q … 10022 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Backup progress should not add block failed to upload to successful count 9793 - @derekbit @chriscchien
  • [BACKPORT][v1.7.3][BUG][v1.8.x] Can not create backup, backup become in error state immediately 10180 - @PhanLe1010 @chriscchien
  • [BACKPORT][v1.7.3][BUG] Storage doesn’t reschedule in v1.7.2 10109 - @PhanLe1010
  • [BACKPORT][v1.7.3][BUG] Old backups are not cleaned up after timeout 9731 - @mantissahz @roger-ryao
  • [BACKPORT][v1.7.3][BUG] UnknowOS Message in Longhorn Node Condition on RHEL 9833 - @yangchiu @mantissahz @roger-ryao
  • [BACKPORT][v1.7.3][BUG] volume FailedMount - Input/output error 10005 - @PhanLe1010 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Unable to delete backing image backup through UI 10068 - @chriscchien @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Error notification appears on the volume backup details page 10071 - @houhoucoop @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Missing fromBackup Parameter in API Request When Restoring Multiple Files from Backup List 10051 - @a110605 @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Webhook servers initialization blocks longhorn-manager from running 10055 - @c3y1huang @chriscchien
  • [BACKPORT][v1.7.3][BUG] CLI check preflight glosses over absence of NFS installation. 9893 - @COLDTURNIP @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Detached Volume Stuck in Attached State During Node Eviction 9810 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] Test case test_node_eviction_multiple_volume failed to reschedule replicas after volume detached 9866 - @yangchiu @c3y1huang
  • [BACKPORT][v1.7.3][BUG] DR volume fails to reattach and faulted after node stop and start during incremental restore 9803 - @c3y1huang @roger-ryao
  • [BACKPORT][v1.7.3][BUG] Share manager is permanently stuck in stopping/error if we shutdown the node of share manager pod. This makes RWX PVC cannot attach to any new node 9856 -
  • [BACKPORT][v1.7.3][BUG] Fail to resize RWX PVC at filesystem resizing step 9738 - @james-munson
  • [BACKPORT][v1.7.3][BUG] Failed to inspect the backup backing image information if NFS backup target URL with options 9703 - @yangchiu @mantissahz
  • [BACKPORT][v1.7.3][BUG] Pre-upgrade pod should event the reason for any failures. 9643 - @yangchiu @james-munson

Misc

  • [TASK] Fix CVE issues for v1.7.3 9897 - @c3y1huang
  • [BACKPORT][v1.7.3][TASK] Install the latest grpc_health_probe at build time 9715 - @yangchiu @c3y1huang

Contributors

  • @COLDTURNIP
  • @ChanYiLin
  • @PhanLe1010
  • @a110605
  • @c3y1huang
  • @chriscchien
  • @derekbit
  • @houhoucoop
  • @innobead
  • @james-munson
  • @mantissahz
  • @roger-ryao
  • @yangchiu

下载链接