fix(log-analyzer): track storage probe failures (#7434)

* fix(log-analyzer): track storage probe failures

* fix(error): merge equivalent api message branches

Co-Authored-By: heihutu <heihutu@gmail.com>

Co-Authored-By: zhi22915 <qiuzgang@gmail.com>

---------

Co-authored-by: houseme <housemecn@gmail.com>
Co-authored-by: zhi22915 <qiuzgang@gmail.com>

crates/log-analyzer/src/rules/seed/disk.rs

@@ -68,14 +68,17 @@ pub(super) fn rules() -> Vec<Rule> {
             )
         },
         Rule {
-            anchors: strings(["reporting peer disks offline after consecutive storage_info failures"]),
+            anchors: strings(["Storage inventory probe failed; current drive health is unknown"]),
             ..base(
                 "peer-disks-offline",
                 P2Degraded,
                 "disk",
-                "peer 磁盘被整体判定离线",
-                contains("reporting peer disks offline after consecutive storage_info failures"),
-                "对某 peer 连续 storage_info 失败,判定其磁盘整体离线。",
+                "peer 存储清单探测失败",
+                any([
+                    contains("Storage inventory probe failed; current drive health is unknown"),
+                    contains("reporting peer disks offline after consecutive storage_info failures"),
+                ]),
+                "某 peer 的 storage_info 探测失败,当前磁盘健康状态未知。",
                 "检查该 peer 节点存活与 RPC 端口可达。",
             )
         },

crates/log-analyzer/src/rules/seed/tests.rs

@@ -110,7 +110,7 @@ fn every_rule_has_a_positive_sample() {
         ("remote-peer-faulty", msg("Remote peer health check failed for node2: marking as faulty")),
         (
             "peer-disks-offline",
-            msg("reporting peer disks offline after consecutive storage_info failures"),
+            msg("Storage inventory probe failed; current drive health is unknown"),
         ),
         ("drive-faulty-error", msg("remote drive is faulty")),
         (
@@ -318,6 +318,10 @@ fn smoke_samples_hit_exact_rule_sets() {
         &["disk-marked-faulty"],
     );
     exact(&msg("erasure write quorum (required=8, achieved=5)"), &["ec-write-quorum"]);
+    exact(
+        &msg("reporting peer disks offline after consecutive storage_info failures"),
+        &["peer-disks-offline"],
+    );
     exact(
         &Sample {
             message: "Metacache listing quorum failed",

crates/scanner/src/scanner.rs

@@ -988,6 +988,17 @@ pub async fn init_scanner_with_recovery(
         return None;
     }
     Some(tokio::spawn(async move {
+        if let Err(error) = run_scanner_usage_recovery_intents_for_startup(ctx.clone(), storeapi.clone()).await {
+            warn!(
+                target: "rustfs::scanner",
+                event = EVENT_SCANNER_PERSIST_STATE,
+                component = LOG_COMPONENT_SCANNER,
+                subsystem = LOG_SUBSYSTEM_RUNTIME,
+                state = "recovery_intent_startup_discovery_failed",
+                error = %error,
+                "Disabled scanner recovery intent startup discovery failed"
+            );
+        }
         if let Err(error) = resume_scanner_cycle_cleanup(ctx, storeapi).await {
             warn!(
                 target: "rustfs::scanner",

crates/scanner/src/scanner/tests/recovery_control.rs

@@ -457,7 +457,7 @@ async fn scanner_recovery_intent_startup_rejects_corrupt_pending_record() {
 
 #[tokio::test]
 #[serial]
-async fn scanner_recovery_intent_disabled_startup_preserves_non_terminal_intent() {
+async fn scanner_recovery_intent_disabled_startup_executes_persisted_non_terminal_intent() {
     let (_dir, store) = setup_scanner_cycle_store().await;
     let record = match accept_scanner_usage_recovery_intent(
         store.clone(),
@@ -473,12 +473,12 @@ async fn scanner_recovery_intent_disabled_startup_preserves_non_terminal_intent(
     let restarted = restart_scanner_cycle_store_from(&store).await;
     run_disabled_startup(CancellationToken::new(), restarted.clone()).await;
 
-    let preserved = get_scanner_usage_recovery_intent(restarted, &record.intent_id)
+    let completed = get_scanner_usage_recovery_intent(restarted, &record.intent_id)
         .await
-        .expect("startup-skipped intent should read")
-        .expect("startup-skipped intent should remain durable");
-    assert_eq!(preserved.state, "accepted");
-    assert_eq!(preserved.intent_id, record.intent_id);
+        .expect("startup-executed intent should read")
+        .expect("startup-executed intent should remain durable");
+    assert_eq!(completed.state, "completed");
+    assert_eq!(completed.intent_id, record.intent_id);
 }
 
 #[tokio::test]