feat: add check replica command to identify broken snapshot chains

cmd/local/subcmd/check.go

@@ -9,6 +9,7 @@ import (
 
 	"github.com/longhorn/cli/pkg/consts"
 	local "github.com/longhorn/cli/pkg/local/preflight"
+	localreplica "github.com/longhorn/cli/pkg/local/replica"
 	"github.com/longhorn/cli/pkg/types"
 	"github.com/longhorn/cli/pkg/utils"
 )
@@ -22,6 +23,61 @@ func NewCmdCheck(globalOpts *types.GlobalCmdOptions) *cobra.Command {
 	utils.SetGlobalOptionsLocal(cmd, globalOpts)
 
 	cmd.AddCommand(newCmdCheckPreflight(globalOpts))
+	cmd.AddCommand(newCmdCheckReplica(globalOpts))
+
+	return cmd
+}
+
+func newCmdCheckReplica(globalOpts *types.GlobalCmdOptions) *cobra.Command {
+	var localChecker = localreplica.Checker{}
+
+	cmd := &cobra.Command{
+		Use:   consts.SubCmdReplica,
+		Short: "Check Longhorn replica integrity",
+		Long: `This command checks the integrity of the snapshot chains in the Longhorn replica data directories.
+It identifies broken snapshot chains, for example snapshots referencing a missing parent, disk files without metadata files, and metadata files without disk files.
+The results are presented by the replica data directory names, not the actual Custom Resource (CR) names.
+
+By default, this command checks all Longhorn replicas in the data directory.
+You can narrow down the results by using the following options:
+- --name: Specify the Longhorn replica data directory name to check a specific replica.
+- --volume-name: Filter replicas by the volume they belong to.`,
+
+		PreRun: func(cmd *cobra.Command, args []string) {
+			localChecker.LogLevel = globalOpts.LogLevel
+
+			err := localChecker.Init()
+			if err != nil {
+				utils.CheckErr(errors.Wrap(err, "Failed to initialize replica checker"))
+			}
+		},
+
+		Run: func(cmd *cobra.Command, args []string) {
+			err := localChecker.Run()
+			if err != nil {
+				utils.CheckErr(errors.Wrap(err, "Failed to run replica checker"))
+			}
+
+			logrus.Info("Successfully checked replica integrity")
+		},
+
+		PostRun: func(cmd *cobra.Command, args []string) {
+			err := localChecker.Output()
+			if err != nil {
+				utils.CheckErr(errors.Wrap(err, "Failed to output replica checker collection"))
+			}
+
+			logrus.Info("Successfully output replica checker collection")
+		},
+	}
+
+	utils.SetGlobalOptionsLocal(cmd, globalOpts)
+
+	cmd.Flags().StringVar(&localChecker.CurrentNodeID, consts.CmdOptNodeId, os.Getenv(consts.EnvCurrentNodeID), "Current node ID.")
+	cmd.Flags().StringVarP(&localChecker.OutputFilePath, consts.CmdOptOutputFile, "o", os.Getenv(consts.EnvOutputFilePath), "Output the result to a file, default to stdout.")
+	cmd.Flags().StringVar(&localChecker.ReplicaName, consts.CmdOptName, os.Getenv(consts.EnvLonghornReplicaName), "Specify the name of the replica to check.")
+	cmd.Flags().StringVar(&localChecker.VolumeName, consts.CmdOptLonghornVolumeName, os.Getenv(consts.EnvLonghornVolumeName), "Specify the name of the volume to check its replicas.")
+	cmd.Flags().StringVar(&localChecker.LonghornDataDirectory, consts.CmdOptLonghornDataDirectory, os.Getenv(consts.EnvLonghornDataDirectory), "Specify the Longhorn data directory. If not provided, the default will be attempted, or it will fall back to the directory of longhorn-disk.cfg.")
 
 	return cmd
 }

cmd/remote/subcmd/check.go

@@ -7,6 +7,7 @@ import (
 
 	"github.com/longhorn/cli/pkg/consts"
 	"github.com/longhorn/cli/pkg/remote/preflight"
+	"github.com/longhorn/cli/pkg/remote/replica"
 	"github.com/longhorn/cli/pkg/types"
 	"github.com/longhorn/cli/pkg/utils"
 )
@@ -20,6 +21,88 @@ func NewCmdCheck(globalOpts *types.GlobalCmdOptions) *cobra.Command {
 	utils.SetGlobalOptionsRemote(cmd, globalOpts)
 
 	cmd.AddCommand(newCmdCheckPreflight(globalOpts))
+	cmd.AddCommand(newCmdCheckReplica(globalOpts))
+
+	return cmd
+}
+
+func newCmdCheckReplica(globalOpts *types.GlobalCmdOptions) *cobra.Command {
+	var replicaChecker = replica.Checker{}
+
+	cmd := &cobra.Command{
+		Use:   consts.SubCmdReplica,
+		Short: "Check Longhorn replica integrity",
+		Long: `This command checks the integrity of the snapshot chains in the Longhorn replica data directories on each node.
+It identifies broken snapshot chains, for example snapshots referencing a missing parent, disk files without metadata files, and metadata files without disk files.
+The results are presented by the replica data directory names, not the actual Custom Resource (CR) names.
+
+By default, this command checks all Longhorn replicas in the data directory.
+You can narrow down the results by using the following options:
+- --name: Specify the Longhorn replica data directory name to check a specific replica.
+- --volume-name: Filter replicas by the volume they belong to.`,
+		Example: `$ longhornctl check replica
+INFO[2024-07-16T17:23:47+08:00] Initializing replica checker
+INFO[2024-07-16T17:23:47+08:00] Cleaning up replica checker
+INFO[2024-07-16T17:23:47+08:00] Running replica checker
+INFO[2024-07-16T17:23:51+08:00] Retrieved replica check results:
+ replicas:
+    pvc-48a6457d-585e-423b-b530-bbc68a5f948a-0e2603a7:
+        - node: ip-10-0-2-123
+          directory: /var/lib/longhorn/replicas/pvc-48a6457d-585e-423b-b530-bbc68a5f948a-0e2603a7
+          volumeName: pvc-48a6457d-585e-423b-b530-bbc68a5f948a
+          snapshotChain:
+            - volume-head-001.img
+            - volume-snap-40b3b028-b3b3-4a35-a806-8bea77f27c00.img
+          errors:
+            - 'broken snapshot chain: disk volume-snap-40b3b028-b3b3-4a35-a806-8bea77f27c00.img references parent volume-snap-6f244bbe-2857-46e4-92e2-eb1e16a63ba1.img, but the parent metadata file is missing'
+INFO[2024-07-16T17:23:51+08:00] Cleaning up replica checker
+INFO[2024-07-16T17:23:51+08:00] Completed replica checker`,
+
+		PreRun: func(cmd *cobra.Command, args []string) {
+			replicaChecker.Image = globalOpts.Image
+			replicaChecker.ImageRegistry = globalOpts.ImageRegistry
+			replicaChecker.ImagePullSecret = globalOpts.ImagePullSecret
+			replicaChecker.KubeConfigPath = globalOpts.KubeConfigPath
+			replicaChecker.NodeSelector = globalOpts.NodeSelector
+			replicaChecker.Tolerations = globalOpts.Tolerations
+			replicaChecker.Namespace = globalOpts.Namespace
+
+			logrus.Info("Initializing replica checker")
+			if err := replicaChecker.Init(); err != nil {
+				utils.CheckErr(errors.Wrap(err, "Failed to initialize replica checker"))
+			}
+
+			logrus.Info("Cleaning up replica checker")
+			if err := replicaChecker.Cleanup(); err != nil {
+				utils.CheckErr(errors.Wrapf(err, "Failed to cleanup replica checker"))
+			}
+		},
+
+		Run: func(cmd *cobra.Command, args []string) {
+			logrus.Info("Running replica checker")
+			output, err := replicaChecker.Run()
+			if err != nil {
+				utils.CheckErr(errors.Wrap(err, "Failed to run replica checker"))
+			}
+
+			logrus.Infof("Retrieved replica check results:\n %v", output)
+		},
+
+		PostRun: func(cmd *cobra.Command, args []string) {
+			logrus.Info("Cleaning up replica checker")
+			if err := replicaChecker.Cleanup(); err != nil {
+				utils.CheckErr(errors.Wrapf(err, "Failed to cleanup replica checker"))
+			}
+
+			logrus.Info("Completed replica checker")
+		},
+	}
+
+	utils.SetGlobalOptionsRemote(cmd, globalOpts)
+
+	cmd.Flags().StringVar(&replicaChecker.ReplicaName, consts.CmdOptName, "", "Specify the name of the replica to check.")
+	cmd.Flags().StringVar(&replicaChecker.VolumeName, consts.CmdOptLonghornVolumeName, "", "Specify the name of the volume to check its replicas.")
+	cmd.Flags().StringVar(&replicaChecker.LonghornDataDirectory, consts.CmdOptLonghornDataDirectory, "/var/lib/longhorn", "Specify the Longhorn data directory. If not provided, the default will be attempted, or it will fall back to the directory of longhorn-disk.cfg.")
 
 	return cmd
 }

docs/longhornctl_check.md

@@ -25,5 +25,6 @@ Longhorn checking operations
 
 * [longhornctl](longhornctl.md)	 - Command-line interface for Longhorn.
 * [longhornctl check preflight](longhornctl_check_preflight.md)	 - Run a preflight check for Longhorn
+* [longhornctl check replica](longhornctl_check_replica.md)	 - Check Longhorn replica integrity
 
-###### Auto generated by spf13/cobra on 6-Jul-2026
+###### Auto generated by spf13/cobra on 20-Jul-2026

docs/longhornctl_check_replica.md

@@ -0,0 +1,68 @@
+## longhornctl check replica
+
+Check Longhorn replica integrity
+
+### Synopsis
+
+This command checks the integrity of the snapshot chains in the Longhorn replica data directories on each node.
+It identifies broken snapshot chains, for example snapshots referencing a missing parent, disk files without metadata files, and metadata files without disk files.
+The results are presented by the replica data directory names, not the actual Custom Resource (CR) names.
+
+By default, this command checks all Longhorn replicas in the data directory.
+You can narrow down the results by using the following options:
+- --name: Specify the Longhorn replica data directory name to check a specific replica.
+- --volume-name: Filter replicas by the volume they belong to.
+
+```
+longhornctl check replica [flags]
+```
+
+### Examples
+
+```
+$ longhornctl check replica
+INFO[2024-07-16T17:23:47+08:00] Initializing replica checker
+INFO[2024-07-16T17:23:47+08:00] Cleaning up replica checker
+INFO[2024-07-16T17:23:47+08:00] Running replica checker
+INFO[2024-07-16T17:23:51+08:00] Retrieved replica check results:
+ replicas:
+    pvc-48a6457d-585e-423b-b530-bbc68a5f948a-0e2603a7:
+        - node: ip-10-0-2-123
+          directory: /var/lib/longhorn/replicas/pvc-48a6457d-585e-423b-b530-bbc68a5f948a-0e2603a7
+          volumeName: pvc-48a6457d-585e-423b-b530-bbc68a5f948a
+          snapshotChain:
+            - volume-head-001.img
+            - volume-snap-40b3b028-b3b3-4a35-a806-8bea77f27c00.img
+          errors:
+            - 'broken snapshot chain: disk volume-snap-40b3b028-b3b3-4a35-a806-8bea77f27c00.img references parent volume-snap-6f244bbe-2857-46e4-92e2-eb1e16a63ba1.img, but the parent metadata file is missing'
+INFO[2024-07-16T17:23:51+08:00] Cleaning up replica checker
+INFO[2024-07-16T17:23:51+08:00] Completed replica checker
+```
+
+### Options
+
+```
+      --data-dir string            Specify the Longhorn data directory. If not provided, the default will be attempted, or it will fall back to the directory of longhorn-disk.cfg. (default "/var/lib/longhorn")
+  -h, --help                       help for replica
+      --image string               Image containing longhornctl-local (default "longhornio/longhorn-cli:v1.13.0-dev")
+      --image-pull-secret string   Secret with registry credentials for pulling images
+      --image-registry string      Registry to apply to all images (CLI, engine, pause, BCI, etc.), replacing any registry already specified in those images.
+      --kubeconfig string          Kubernetes config (kubeconfig) path
+  -l, --log-level string           Log level (default "info")
+      --name string                Specify the name of the replica to check.
+      --namespace string           The namespace to run DaemonSet pods. (default "longhorn-system")
+      --node-selector string       Comma-separated list of key=value pairs to match against node labels, selecting the nodes the DaemonSet will run on (e.g. env=prod,zone=us-west).
+      --volume-name string         Specify the name of the volume to check its replicas.
+```
+
+### Options inherited from parent commands
+
+```
+      --tolerations string   Semicolon-separated list of tolerations for DaemonSet pods (e.g. key=value:NoSchedule;:NoExecute).
+```
+
+### SEE ALSO
+
+* [longhornctl check](longhornctl_check.md)	 - Longhorn checking operations
+
+###### Auto generated by spf13/cobra on 20-Jul-2026

pkg/consts/replica.go

@@ -1,6 +1,7 @@
 package consts
 
 const (
+	AppNameReplicaChecker  = "longhorn-replica-checker"
 	AppNameReplicaExporter = "longhorn-replica-exporter"
 	AppNameReplicaGetter   = "longhorn-replica-getter"
 )