Merge branch 'main' into cxymds/perf-decommission-checkpoint

crates/config/src/constants/runtime.rs

@@ -57,13 +57,6 @@ pub const DEFAULT_MAX_IO_EVENTS_PER_TICK: usize = 1024;
 pub const DEFAULT_EVENT_INTERVAL: u32 = 61;
 pub const DEFAULT_RNG_SEED: Option<u64> = None; // None means random
 
-/// Dedicated blocking thread pool for fsync/fdatasync operations.
-/// When > 1, fsync operations are isolated from the main blocking pool to
-/// prevent device-bound fsync from starving read operations (pread/stat/open).
-/// Default 0 means auto (no isolation, use main runtime).
-pub const ENV_FSYNC_BLOCKING_THREADS: &str = "RUSTFS_RUNTIME_FSYNC_BLOCKING_THREADS";
-pub const DEFAULT_FSYNC_BLOCKING_THREADS: usize = 0;
-
 // Dial9 Tokio Telemetry Default values
 pub const DEFAULT_RUNTIME_DIAL9_ENABLED: bool = false; // Disabled by default
 pub const DEFAULT_RUNTIME_DIAL9_OUTPUT_DIR: &str = "/var/log/rustfs/telemetry";

crates/ecstore/src/core/pools.rs

@@ -89,6 +89,7 @@ const DECOMMISSION_STAGE_SOURCE_CLEANUP: &str = "source_cleanup";
 const DECOMMISSION_STAGE_ENTRY_FINISHED: &str = "entry_finished";
 const DECOMMISSION_PROGRESS_SAVE_INTERVAL: Duration = Duration::seconds(30);
 const DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD: usize = 1000;
+const DECOMMISSION_PROGRESS_SAVE_RETRY_BACKOFF: Duration = Duration::seconds(1);
 const DECOMMISSION_BUCKET_CONCURRENCY_ENV: &str = "RUSTFS_DECOMMISSION_BUCKET_CONCURRENCY";
 const DECOMMISSION_BUCKET_CONCURRENCY_DEFAULT_CAP: usize = 4;
 const DECOMMISSION_TARGET_CAPACITY_OVERHEAD_PERCENT: usize = 30;
@@ -638,22 +639,6 @@ fn track_decommission_current_object(meta: &mut PoolMeta, idx: usize, bucket: &s
     track_decommission_current_object_stage(meta, idx, bucket, object, "")
 }
 
-fn touch_decommission_progress(meta: &mut PoolMeta, idx: usize) -> Result<()> {
-    let pool_count = meta.pools.len();
-    ensure_valid_decommission_pool_index(pool_count, idx)?;
-
-    let Some(pool) = meta.pools.get_mut(idx) else {
-        return Err(invalid_decommission_pool_index_error(pool_count, idx));
-    };
-    let Some(info) = pool.decommission.as_mut() else {
-        return Err(decommission_metadata_not_initialized_error("touch decommission progress"));
-    };
-
-    pool.last_update = OffsetDateTime::now_utc();
-    info.mark_progress_saved();
-    Ok(())
-}
-
 fn resolve_decommission_update_after_result(result: Result<bool>) -> Result<bool> {
     result.map_err(|err| Error::other(format!("decommission metadata update failed: {err}")))
 }
@@ -1483,6 +1468,7 @@ impl TryFrom<PersistedPoolDecommissionInfo> for PoolDecommissionInfo {
             terminal_reload_attempt_at: value.terminal_reload_attempt_at,
             terminal_reload_failures: value.terminal_reload_failures,
             progress_save_item_baseline: value.items_decommissioned.saturating_add(value.items_decommission_failed),
+            progress_save_retry_after: None,
         })
     }
 }
@@ -1514,6 +1500,7 @@ impl TryFrom<LegacyPoolDecommissionInfo> for PoolDecommissionInfo {
             terminal_reload_attempt_at: None,
             terminal_reload_failures: Vec::new(),
             progress_save_item_baseline: value.items_decommissioned.saturating_add(value.items_decommission_failed),
+            progress_save_retry_after: None,
         })
     }
 }
@@ -1627,6 +1614,82 @@ impl PoolMeta {
         }
     }
 
+    fn decommission_progress_checkpoint(
+        &self,
+        idx: usize,
+        duration: Duration,
+        now: OffsetDateTime,
+    ) -> Result<Option<DecommissionProgressCheckpoint>> {
+        let pool_count = self.pools.len();
+        ensure_valid_decommission_pool_index(pool_count, idx)?;
+
+        let Some(pool) = self.pools.get(idx) else {
+            return Err(invalid_decommission_pool_index_error(pool_count, idx));
+        };
+        let Some(info) = pool.decommission.as_ref() else {
+            return Err(decommission_metadata_not_initialized_error("update decommission metadata timestamp"));
+        };
+
+        if info.progress_save_retry_after.is_some_and(|retry_after| now < retry_after) {
+            return Ok(None);
+        }
+
+        let time_threshold_reached = now.unix_timestamp() - pool.last_update.unix_timestamp() >= duration.whole_seconds();
+        let item_threshold_reached = info.items_since_last_progress_save() >= DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD;
+        if !time_threshold_reached && !item_threshold_reached {
+            return Ok(None);
+        }
+
+        Ok(Some(DecommissionProgressCheckpoint {
+            start_time: info.start_time,
+            queued: info.queued,
+            counted_items: info.counted_items(),
+            checkpoint_at: now,
+        }))
+    }
+
+    fn commit_decommission_progress_checkpoint(&mut self, idx: usize, checkpoint: DecommissionProgressCheckpoint) -> bool {
+        let Some(pool) = self.pools.get_mut(idx) else {
+            return false;
+        };
+        let Some(info) = pool.decommission.as_mut() else {
+            return false;
+        };
+
+        if info.start_time != checkpoint.start_time
+            || info.queued != checkpoint.queued
+            || !is_decommission_active(info.complete, info.failed, info.canceled)
+        {
+            return false;
+        }
+
+        info.progress_save_item_baseline = info.progress_save_item_baseline.max(checkpoint.counted_items);
+        info.progress_save_retry_after = None;
+        pool.last_update = pool.last_update.max(checkpoint.checkpoint_at);
+        true
+    }
+
+    fn defer_decommission_progress_checkpoint(
+        &mut self,
+        idx: usize,
+        checkpoint: DecommissionProgressCheckpoint,
+        retry_after: OffsetDateTime,
+    ) {
+        let Some(pool) = self.pools.get_mut(idx) else {
+            return;
+        };
+        let Some(info) = pool.decommission.as_mut() else {
+            return;
+        };
+
+        if info.start_time == checkpoint.start_time
+            && info.queued == checkpoint.queued
+            && is_decommission_active(info.complete, info.failed, info.canceled)
+        {
+            info.progress_save_retry_after = Some(retry_after);
+        }
+    }
+
     fn load_from_config_data(&mut self, data: Vec<u8>) -> Result<()> {
         if data.is_empty() {
             return Ok(());
@@ -1987,30 +2050,9 @@ impl PoolMeta {
     }
 
     pub fn update_after(&mut self, idx: usize, duration: Duration) -> Result<bool> {
-        let pool_count = self.pools.len();
-        ensure_valid_decommission_pool_index(pool_count, idx)?;
-
-        let (last_update, item_threshold_reached) = match self.pools.get(idx) {
-            Some(pool) if let Some(info) = pool.decommission.as_ref() => (
-                pool.last_update,
-                info.items_since_last_progress_save() >= DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD,
-            ),
-            Some(_) => {
-                return Err(decommission_metadata_not_initialized_error("update decommission metadata timestamp"));
-            }
-            None => return Err(invalid_decommission_pool_index_error(pool_count, idx)),
-        };
-        let now = OffsetDateTime::now_utc();
-
-        if now.unix_timestamp() - last_update.unix_timestamp() >= duration.whole_seconds() || item_threshold_reached {
-            let Some(pool) = self.pools.get_mut(idx) else {
-                return Err(invalid_decommission_pool_index_error(pool_count, idx));
-            };
-            pool.last_update = now;
-            return Ok(true);
-        }
-
-        Ok(false)
+        Ok(self
+            .decommission_progress_checkpoint(idx, duration, OffsetDateTime::now_utc())?
+            .is_some())
     }
 
     pub fn validate(&self, pools: Vec<Arc<Sets>>) -> Result<bool> {
@@ -2151,6 +2193,16 @@ pub struct PoolDecommissionInfo {
     pub terminal_reload_failures: Vec<String>,
     #[serde(skip)]
     pub progress_save_item_baseline: usize,
+    #[serde(skip)]
+    pub progress_save_retry_after: Option<OffsetDateTime>,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+struct DecommissionProgressCheckpoint {
+    start_time: Option<OffsetDateTime>,
+    queued: bool,
+    counted_items: usize,
+    checkpoint_at: OffsetDateTime,
 }
 
 impl PoolDecommissionInfo {
@@ -2185,6 +2237,7 @@ impl PoolDecommissionInfo {
 
     fn mark_progress_saved(&mut self) {
         self.progress_save_item_baseline = self.counted_items();
+        self.progress_save_retry_after = None;
     }
 
     pub fn bucket_push(&mut self, bucket: &DecomBucketInfo) {
@@ -2489,6 +2542,40 @@ impl ECStore {
         snapshot.save(self.pools.clone()).await
     }
 
+    async fn save_decommission_progress_checkpoint(&self, idx: usize) -> Result<bool> {
+        // Lock order: save gate, then the short pool metadata read/write sections. Peer
+        // reloads are intentionally performed by the caller after both locks are released.
+        let _save_guard = self.pool_meta_save_gate.lock().await;
+        let (snapshot, checkpoint) = {
+            let pool_meta = self.pool_meta.read().await;
+            let Some(checkpoint) = pool_meta.decommission_progress_checkpoint(
+                idx,
+                DECOMMISSION_PROGRESS_SAVE_INTERVAL,
+                OffsetDateTime::now_utc(),
+            )?
+            else {
+                return Ok(false);
+            };
+
+            let mut snapshot = pool_meta.clone();
+            let Some(pool) = snapshot.pools.get_mut(idx) else {
+                return Err(invalid_decommission_pool_index_error(snapshot.pools.len(), idx));
+            };
+            pool.last_update = checkpoint.checkpoint_at;
+            (snapshot, checkpoint)
+        };
+
+        if let Err(err) = snapshot.save(self.pools.clone()).await {
+            let retry_after = OffsetDateTime::now_utc() + DECOMMISSION_PROGRESS_SAVE_RETRY_BACKOFF;
+            let mut pool_meta = self.pool_meta.write().await;
+            pool_meta.defer_decommission_progress_checkpoint(idx, checkpoint, retry_after);
+            return Err(err);
+        }
+
+        let mut pool_meta = self.pool_meta.write().await;
+        Ok(pool_meta.commit_decommission_progress_checkpoint(idx, checkpoint))
+    }
+
     async fn save_current_pool_meta_for_decommission_start(
         &self,
         indices: &[usize],
@@ -2871,7 +2958,7 @@ impl ECStore {
         Ok(())
     }
 
-    async fn save_decommission_entry_progress_stage(
+    async fn track_decommission_entry_progress_stage(
         &self,
         idx: usize,
         bucket: &str,
@@ -2882,22 +2969,6 @@ impl ECStore {
             let mut pool_meta = self.pool_meta.write().await;
             track_decommission_current_object_stage(&mut pool_meta, idx, bucket, object, stage)
                 .map_err(|err| with_decommission_entry_context(stage, bucket, object, err))?;
-            touch_decommission_progress(&mut pool_meta, idx)
-                .map_err(|err| with_decommission_entry_context(stage, bucket, object, err))?;
-        }
-
-        if let Some(err) = resolve_decommission_progress_save_result(self.save_current_pool_meta().await) {
-            warn!(
-                event = EVENT_DECOMMISSION_ENTRY,
-                component = LOG_COMPONENT_ECSTORE,
-                subsystem = LOG_SUBSYSTEM_POOLS,
-                pool_index = idx,
-                bucket = %bucket,
-                object = %object,
-                stage,
-                error = ?err,
-                "Decommission progress stage save failed"
-            );
         }
 
         Ok(())
@@ -3165,7 +3236,7 @@ impl ECStore {
                 let bucket_name = bucket.clone();
                 let object_name = rd.object_info.name.clone();
 
-                self.save_decommission_entry_progress_stage(
+                self.track_decommission_entry_progress_stage(
                     idx,
                     bucket_name.as_str(),
                     object_name.as_str(),
@@ -3259,7 +3330,7 @@ impl ECStore {
             }
             decommission_cancel_signal_result(rx.is_cancelled())?;
 
-            self.save_decommission_entry_progress_stage(
+            self.track_decommission_entry_progress_stage(
                 idx,
                 bucket.as_str(),
                 entry.name.as_str(),
@@ -3267,7 +3338,7 @@ impl ECStore {
             )
             .await?;
 
-            self.save_decommission_entry_progress_stage(
+            self.track_decommission_entry_progress_stage(
                 idx,
                 bucket.as_str(),
                 entry.name.as_str(),
@@ -3334,34 +3405,42 @@ impl ECStore {
             }
         };
 
-        self.save_decommission_entry_progress_stage(idx, bucket.as_str(), entry.name.as_str(), DECOMMISSION_STAGE_ENTRY_FINISHED)
-            .await?;
+        self.track_decommission_entry_progress_stage(
+            idx,
+            bucket.as_str(),
+            entry.name.as_str(),
+            DECOMMISSION_STAGE_ENTRY_FINISHED,
+        )
+        .await?;
 
         if should_save_progress {
-            let save_result = self.save_current_pool_meta().await;
-            if let Some(err) = resolve_decommission_progress_save_result(save_result) {
-                warn!(
-                    event = EVENT_DECOMMISSION_ENTRY,
-                    component = LOG_COMPONENT_ECSTORE,
-                    subsystem = LOG_SUBSYSTEM_POOLS,
-                    pool_index = idx,
-                    bucket = %bucket,
-                    object = %entry.name,
-                    state = "progress_save_failed",
-                    error = %err,
-                    "Decommission progress save failed; continuing and will retry at the next checkpoint"
-                );
-            } else {
-                let mut pool_meta = self.pool_meta.write().await;
-                pool_meta.mark_decommission_progress_saved();
-                if let Some(notification_sys) = runtime_sources::notification_sys()
-                    && let Err(err) = resolve_decommission_entry_reload_result(
-                        notification_sys.reload_pool_meta().await,
-                        bucket.as_str(),
-                        entry.name.as_str(),
-                    )
-                {
-                    warn!("{err}");
+            match self.save_decommission_progress_checkpoint(idx).await {
+                Ok(true) => {
+                    if let Some(notification_sys) = runtime_sources::notification_sys()
+                        && let Err(err) = resolve_decommission_entry_reload_result(
+                            notification_sys.reload_pool_meta().await,
+                            bucket.as_str(),
+                            entry.name.as_str(),
+                        )
+                    {
+                        warn!("{err}");
+                    }
+                }
+                Ok(false) => {}
+                Err(err) => {
+                    if let Some(err) = resolve_decommission_progress_save_result(Err(err)) {
+                        warn!(
+                            event = EVENT_DECOMMISSION_ENTRY,
+                            component = LOG_COMPONENT_ECSTORE,
+                            subsystem = LOG_SUBSYSTEM_POOLS,
+                            pool_index = idx,
+                            bucket = %bucket,
+                            object = %entry.name,
+                            state = "progress_save_failed",
+                            error = %err,
+                            "Decommission progress save failed; continuing and will retry at the next checkpoint"
+                        );
+                    }
                 }
             }
         }
@@ -5264,11 +5343,11 @@ pub(crate) fn fallback_free_capacity_dedup(disks: &[rustfs_madmin::Disk]) -> usi
 #[cfg(test)]
 mod pools_tests {
     use super::{
-        DECOMMISSION_PROGRESS_SAVE_INTERVAL, DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD, DecomBucketInfo,
-        DecommissionStartPoolState, DecommissionTerminalState, ListCallback, PoolDecommissionInfo, PoolMeta, PoolSpaceInfo,
-        PoolStatus, apply_decommission_status_space_info, bind_decommission_cancelers, bind_missing_decommission_cancelers,
-        cancel_decommission_canceler, classify_decommission_terminal_state, count_decommission_item,
-        decommission_cancel_signal_result, decommission_item_size, decommission_meta_bucket_options,
+        DECOMMISSION_PROGRESS_SAVE_INTERVAL, DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD, DECOMMISSION_PROGRESS_SAVE_RETRY_BACKOFF,
+        DecomBucketInfo, DecommissionStartPoolState, DecommissionTerminalState, ListCallback, PoolDecommissionInfo, PoolMeta,
+        PoolSpaceInfo, PoolStatus, apply_decommission_status_space_info, bind_decommission_cancelers,
+        bind_missing_decommission_cancelers, cancel_decommission_canceler, classify_decommission_terminal_state,
+        count_decommission_item, decommission_cancel_signal_result, decommission_item_size, decommission_meta_bucket_options,
         decommission_start_pool_state, dedup_indices, default_decommission_bucket_concurrency,
         ensure_decommission_cancel_allowed, ensure_decommission_clear_allowed, ensure_decommission_listing_disks_available,
         ensure_decommission_not_rebalancing, ensure_decommission_start_allowed, ensure_decommission_start_keeps_active_pool,
@@ -5293,9 +5372,8 @@ mod pools_tests {
         should_preserve_decommission_canceled_state, should_reject_decommission_cancel_as_terminal,
         should_retry_decommission_cancel_reload, should_retry_decommission_listing, should_skip_canceled_decommission_routine,
         split_decommission_buckets, take_and_cancel_decommission_canceler, take_decommission_canceler,
-        touch_decommission_progress, track_decommission_current_object, track_decommission_current_object_stage,
-        validate_start_decommission_request, wait_decommission_listing_retry, wait_decommission_worker_drain,
-        with_decommission_entry_context,
+        track_decommission_current_object, track_decommission_current_object_stage, validate_start_decommission_request,
+        wait_decommission_listing_retry, wait_decommission_worker_drain, with_decommission_entry_context,
     };
     use crate::data_movement;
     use crate::disk::endpoint::Endpoint;
@@ -6538,7 +6616,7 @@ mod pools_tests {
     }
 
     #[test]
-    fn test_touch_decommission_progress_updates_last_update_and_save_baseline() {
+    fn test_track_decommission_stage_does_not_advance_checkpoint_state() {
         let mut meta = PoolMeta {
             pools: vec![PoolStatus {
                 id: 0,
@@ -6553,11 +6631,13 @@ mod pools_tests {
             ..Default::default()
         };
 
-        touch_decommission_progress(&mut meta, 0).expect("valid decommission progress should be touched");
+        track_decommission_current_object_stage(&mut meta, 0, "bucket", "object", "migrate_object")
+            .expect("valid decommission progress should be tracked");
 
-        assert!(meta.pools[0].last_update > OffsetDateTime::UNIX_EPOCH);
+        assert_eq!(meta.pools[0].last_update, OffsetDateTime::UNIX_EPOCH);
         let info = meta.pools[0].decommission.as_ref().expect("decommission info should exist");
-        assert_eq!(info.items_since_last_progress_save(), 0);
+        assert_eq!(info.items_since_last_progress_save(), 5);
+        assert_eq!(info.stage, "migrate_object");
     }
 
     #[test]
@@ -6632,6 +6712,134 @@ mod pools_tests {
         assert_eq!(info.items_since_last_progress_save(), 1);
     }
 
+    #[test]
+    fn test_pool_meta_update_after_does_not_advance_last_update_before_save() {
+        let last_update = OffsetDateTime::UNIX_EPOCH;
+        let mut meta = PoolMeta {
+            pools: vec![PoolStatus {
+                id: 0,
+                cmd_line: "pool-0".to_string(),
+                last_update,
+                decommission: Some(PoolDecommissionInfo {
+                    start_time: Some(last_update),
+                    items_decommissioned: DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD,
+                    ..Default::default()
+                }),
+            }],
+            ..Default::default()
+        };
+
+        assert!(
+            meta.update_after(0, DECOMMISSION_PROGRESS_SAVE_INTERVAL)
+                .expect("item threshold should request a checkpoint")
+        );
+        assert_eq!(meta.pools[0].last_update, last_update);
+    }
+
+    #[test]
+    fn test_decommission_progress_checkpoint_commits_exact_snapshot_watermark() {
+        let start_time = OffsetDateTime::UNIX_EPOCH;
+        let checkpoint_at = start_time + Duration::seconds(30);
+        let mut meta = PoolMeta {
+            pools: vec![PoolStatus {
+                id: 0,
+                cmd_line: "pool-0".to_string(),
+                last_update: start_time,
+                decommission: Some(PoolDecommissionInfo {
+                    start_time: Some(start_time),
+                    items_decommissioned: DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD,
+                    ..Default::default()
+                }),
+            }],
+            ..Default::default()
+        };
+
+        let checkpoint = meta
+            .decommission_progress_checkpoint(0, DECOMMISSION_PROGRESS_SAVE_INTERVAL, checkpoint_at)
+            .expect("valid decommission state should produce a checkpoint")
+            .expect("item threshold should produce a checkpoint");
+        meta.count_item(0, 1, false);
+
+        assert!(meta.commit_decommission_progress_checkpoint(0, checkpoint));
+        let info = meta.pools[0].decommission.as_ref().expect("decommission info should exist");
+        assert_eq!(info.progress_save_item_baseline, checkpoint.counted_items);
+        assert_eq!(info.items_since_last_progress_save(), 1);
+        assert_eq!(meta.pools[0].last_update, checkpoint_at);
+    }
+
+    #[test]
+    fn test_decommission_progress_checkpoint_backoff_does_not_advance_baseline() {
+        let start_time = OffsetDateTime::UNIX_EPOCH;
+        let checkpoint_at = start_time + Duration::seconds(30);
+        let retry_after = checkpoint_at + DECOMMISSION_PROGRESS_SAVE_RETRY_BACKOFF;
+        let mut meta = PoolMeta {
+            pools: vec![PoolStatus {
+                id: 0,
+                cmd_line: "pool-0".to_string(),
+                last_update: start_time,
+                decommission: Some(PoolDecommissionInfo {
+                    start_time: Some(start_time),
+                    items_decommissioned: DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD,
+                    ..Default::default()
+                }),
+            }],
+            ..Default::default()
+        };
+
+        let checkpoint = meta
+            .decommission_progress_checkpoint(0, DECOMMISSION_PROGRESS_SAVE_INTERVAL, checkpoint_at)
+            .expect("valid decommission state should produce a checkpoint")
+            .expect("item threshold should produce a checkpoint");
+        meta.defer_decommission_progress_checkpoint(0, checkpoint, retry_after);
+
+        assert!(
+            meta.decommission_progress_checkpoint(0, DECOMMISSION_PROGRESS_SAVE_INTERVAL, checkpoint_at)
+                .expect("retry backoff check should succeed")
+                .is_none()
+        );
+        assert_eq!(meta.pools[0].last_update, start_time);
+        assert_eq!(
+            meta.pools[0]
+                .decommission
+                .as_ref()
+                .expect("decommission info should exist")
+                .progress_save_item_baseline,
+            0
+        );
+    }
+
+    #[test]
+    fn test_decommission_progress_checkpoint_count_scales_with_threshold() {
+        let start_time = OffsetDateTime::UNIX_EPOCH;
+        let checkpoint_at = start_time;
+        let mut meta = PoolMeta {
+            pools: vec![PoolStatus {
+                id: 0,
+                cmd_line: "pool-0".to_string(),
+                last_update: start_time,
+                decommission: Some(PoolDecommissionInfo {
+                    start_time: Some(start_time),
+                    ..Default::default()
+                }),
+            }],
+            ..Default::default()
+        };
+        let mut checkpoint_count = 0;
+
+        for _ in 0..(DECOMMISSION_PROGRESS_SAVE_ITEM_THRESHOLD * 10) {
+            meta.count_item(0, 1, false);
+            if let Some(checkpoint) = meta
+                .decommission_progress_checkpoint(0, DECOMMISSION_PROGRESS_SAVE_INTERVAL, checkpoint_at)
+                .expect("valid decommission state should produce a checkpoint")
+            {
+                checkpoint_count += 1;
+                assert!(meta.commit_decommission_progress_checkpoint(0, checkpoint));
+            }
+        }
+
+        assert_eq!(checkpoint_count, 10);
+    }
+
     #[test]
     fn test_ensure_decommission_not_rebalancing_rejects_running_rebalance() {
         let err = ensure_decommission_not_rebalancing(true).expect_err("rebalance running should be rejected");

crates/ecstore/src/disk/os.rs

@@ -315,7 +315,7 @@ pub async fn fsync_dir(dir: impl AsRef<Path>) -> io::Result<()> {
     #[cfg(unix)]
     {
         let dir = dir.as_ref().to_path_buf();
-        fsync_spawn_blocking(move || fsync_dir_std(dir)).await?
+        tokio::task::spawn_blocking(move || fsync_dir_std(dir)).await?
     }
 
     #[cfg(not(unix))]
@@ -683,7 +683,7 @@ async fn fsync_open_dst_dir_group(group: &DstDirFsyncGroup) -> io::Result<()> {
     #[cfg(test)]
     let dir = group.dir.clone();
     let dir_file = group.dir_file.clone();
-    fsync_spawn_blocking(move || {
+    tokio::task::spawn_blocking(move || {
         #[cfg(test)]
         {
             if let Some(kind) = fsync_dir_recorder::take_grouped_failure(&dir) {
@@ -1080,44 +1080,6 @@ const TEST_GLOBAL_FILE_SYNCS: usize = 64;
 
 static FILE_SYNC_PERMITS: LazyLock<Semaphore> = LazyLock::new(|| Semaphore::new(global_file_sync_limit()));
 static DISK_FILE_SYNC_LIMITERS: LazyLock<Mutex<HashMap<PathBuf, Weak<Semaphore>>>> = LazyLock::new(|| Mutex::new(HashMap::new()));
-
-/// Dedicated tokio runtime for fsync/fdatasync blocking operations. When
-/// configured with >1 threads, isolates device-bound fsync from the main
-/// blocking pool so reads (pread/stat/open) are not starved. `None` means
-/// fall back to the main runtime (zero behavior change).
-static FSYNC_RUNTIME: LazyLock<Option<tokio::runtime::Runtime>> = LazyLock::new(|| {
-    let threads =
-        rustfs_utils::get_env_usize(rustfs_config::ENV_FSYNC_BLOCKING_THREADS, rustfs_config::DEFAULT_FSYNC_BLOCKING_THREADS);
-    if threads <= 1 {
-        return None;
-    }
-    let mut builder = tokio::runtime::Builder::new_multi_thread();
-    builder
-        .worker_threads(num_cpus::get().min(8))
-        .max_blocking_threads(threads)
-        .thread_name("rustfs-fsync")
-        .thread_stack_size(512 * 1024)
-        .enable_all();
-    match builder.build() {
-        Ok(rt) => {
-            tracing::info!(threads, "fsync dedicated blocking pool enabled");
-            Some(rt)
-        }
-        Err(err) => {
-            tracing::warn!(%err, "failed to build fsync runtime, falling back to main pool");
-            None
-        }
-    }
-});
-
-/// Spawn a blocking task on the fsync-dedicated runtime if configured,
-/// otherwise fall back to the main tokio blocking pool.
-fn fsync_spawn_blocking<T: Send + 'static>(f: impl FnOnce() -> T + Send + 'static) -> tokio::task::JoinHandle<T> {
-    match FSYNC_RUNTIME.as_ref() {
-        Some(rt) => rt.spawn_blocking(f),
-        None => tokio::task::spawn_blocking(f),
-    }
-}
 static DISK_VOLUME_MUTATION_LOCKS: LazyLock<Mutex<HashMap<PathBuf, Weak<RwLock<()>>>>> =
     LazyLock::new(|| Mutex::new(HashMap::new()));
 type NamespaceMutationLock = AsyncMutex<()>;
@@ -1255,7 +1217,7 @@ where
     F: FnOnce() -> io::Result<T> + Send + 'static,
 {
     let (disk_permit, global_permit) = acquire_file_sync_permits(disk_permits).await?;
-    let result = fsync_spawn_blocking(move || {
+    let result = tokio::task::spawn_blocking(move || {
         let _disk_permit = disk_permit;
         work()
     })
@@ -2184,7 +2146,7 @@ async fn run_blocking_namespace_file_sync_operation_with_global<T: Send + 'stati
         wait_started,
     );
     let disk_permit = admission.disk_permit.clone();
-    let result = fsync_spawn_blocking(move || {
+    let result = tokio::task::spawn_blocking(move || {
         let _lease = lease;
         let _disk_permit = disk_permit;
         operation()

crates/heal/src/heal/erasure_healer.rs

@@ -373,11 +373,6 @@ impl ErasureSetHealer {
         set_disk_id: &str,
         buckets: &[String],
     ) -> Result<(ResumeManager, CheckpointManager)> {
-        if self.replacement_task_id.is_none() && CheckpointManager::is_blocked(&self.disk, task_id).await {
-            return Err(Error::TaskExecutionFailed {
-                message: format!("Resume task {task_id} has a blocked checkpoint"),
-            });
-        }
         // check if resume state exists
         let has_resume_state = if self.replacement_task_id.is_some() {
             ResumeManager::has_replacement_intent(&self.disk, task_id).await

crates/heal/src/heal/resume.rs

@@ -51,7 +51,6 @@ const RESUME_STATE_FILE: &str = "ahm_resume_state.json";
 const REPLACEMENT_INTENT_FILE: &str = "ahm_replacement_intent.json";
 const RESUME_PROGRESS_FILE: &str = "ahm_progress.json";
 pub(super) const RESUME_CHECKPOINT_FILE: &str = "ahm_checkpoint.json";
-pub(super) const RESUME_CHECKPOINT_BLOCKED_FILE: &str = "ahm_checkpoint.blocked";
 const REPLACEMENT_COMPLETION_PROOF_FILE: &str = "ahm_replacement_completion_proof.json";
 const REPLACEMENT_RECOVERY_DIR: &str = "ahm-replacement";
 const REPLACEMENT_INTENT_SEAL_FILE: &str = "ahm_replacement_intent_seal";

crates/heal/src/heal/resume/checkpoint.rs

@@ -18,14 +18,13 @@ use std::collections::HashSet;
 use std::path::Path;
 use std::sync::{Arc, Mutex};
 use std::time::{SystemTime, UNIX_EPOCH};
-use tokio::sync::{Mutex as AsyncMutex, RwLock};
+use tokio::sync::RwLock;
 use tracing::{debug, warn};
 
-use super::super::storage_api::owner::{EcstoreConditionalFileUpdate, EcstoreDiskAPI, EcstoreDiskBytes};
-use super::super::{BUCKET_META_PREFIX, DiskStore, HealDiskExt, RUSTFS_META_BUCKET};
+use super::super::{BUCKET_META_PREFIX, DiskStore, HealDiskExt as _, RUSTFS_META_BUCKET};
 use super::{
-    LOG_COMPONENT_HEAL, LOG_SUBSYSTEM_RESUME, PersistThrottle, RESUME_CHECKPOINT_BLOCKED_FILE, RESUME_CHECKPOINT_FILE,
-    delete_resume_file, path_to_str, validate_resume_task_id,
+    LOG_COMPONENT_HEAL, LOG_SUBSYSTEM_RESUME, PersistThrottle, RESUME_CHECKPOINT_FILE, delete_resume_file, path_to_str,
+    validate_resume_task_id,
 };
 
 const EVENT_HEAL_CHECKPOINT_STATE: &str = "heal_checkpoint_state";
@@ -117,108 +116,17 @@ pub struct CheckpointManager {
     disk: DiskStore,
     checkpoint: Arc<RwLock<ResumeCheckpoint>>,
     throttle: Mutex<PersistThrottle>,
-    save_lock: AsyncMutex<()>,
-    last_saved: Mutex<Option<EcstoreDiskBytes>>,
 }
 
 impl CheckpointManager {
-    fn blocked_path(task_id: &str) -> std::path::PathBuf {
-        Path::new(BUCKET_META_PREFIX).join(format!("{task_id}_{RESUME_CHECKPOINT_BLOCKED_FILE}"))
-    }
-
-    /// Return whether a checkpoint was permanently isolated after a malformed
-    /// or unsupported snapshot was observed.
-    pub(crate) async fn is_blocked(disk: &DiskStore, task_id: &str) -> bool {
-        if validate_resume_task_id(task_id).is_err() {
-            return false;
-        }
-        let blocked_path = Self::blocked_path(task_id);
-        let Ok(path) = path_to_str(&blocked_path) else {
-            return false;
-        };
-        match HealDiskExt::read_all(disk.as_ref(), RUSTFS_META_BUCKET, path).await {
-            Ok(_) => true,
-            Err(crate::heal::DiskError::FileNotFound) => false,
-            Err(_) => true,
-        }
-    }
-
-    /// Validate the checkpoint while enumerating resumable state. This reads
-    /// the checkpoint once and also isolates malformed or unsupported data.
-    pub(crate) async fn is_resumable(disk: &DiskStore, task_id: &str) -> bool {
-        if validate_resume_task_id(task_id).is_err() || Self::is_blocked(disk, task_id).await {
-            return false;
-        }
-        let file_path = Path::new(BUCKET_META_PREFIX).join(format!("{task_id}_{RESUME_CHECKPOINT_FILE}"));
-        let Ok(path) = path_to_str(&file_path) else {
-            return false;
-        };
-        match HealDiskExt::read_all(disk.as_ref(), RUSTFS_META_BUCKET, path).await {
-            Ok(bytes) if bytes.is_empty() => true,
-            Ok(bytes) => Self::load_from_data(disk.clone(), task_id, bytes.to_vec()).await.is_ok(),
-            Err(crate::heal::DiskError::FileNotFound) => true,
-            Err(_) => false,
-        }
-    }
-
-    async fn block_invalid_snapshot(disk: &DiskStore, task_id: &str) {
-        // This marker is intentionally version-agnostic: an unsupported reader
-        // must stop selector retries until an operator cleans up the snapshot.
-        let blocked_path = Self::blocked_path(task_id);
-        let Ok(path) = path_to_str(&blocked_path) else {
-            return;
-        };
-        let result = EcstoreDiskAPI::compare_and_update_file(
-            disk.as_ref(),
-            RUSTFS_META_BUCKET,
-            path,
-            None,
-            Some(EcstoreDiskBytes::from_static(b"blocked")),
-        )
-        .await;
-        match result {
-            Ok(EcstoreConditionalFileUpdate::Updated | EcstoreConditionalFileUpdate::Mismatch) => {}
-            Ok(EcstoreConditionalFileUpdate::Missing) => warn!(
-                target: "rustfs::heal::resume",
-                event = EVENT_HEAL_CHECKPOINT_STATE,
-                component = LOG_COMPONENT_HEAL,
-                subsystem = LOG_SUBSYSTEM_RESUME,
-                task_id,
-                state = "blocked_marker_write_failed",
-                error = "marker target disappeared",
-                "Heal checkpoint could not persist its blocked marker"
-            ),
-            Err(error) => warn!(
-                target: "rustfs::heal::resume",
-                event = EVENT_HEAL_CHECKPOINT_STATE,
-                component = LOG_COMPONENT_HEAL,
-                subsystem = LOG_SUBSYSTEM_RESUME,
-                task_id,
-                state = "blocked_marker_write_failed",
-                error = %error,
-                "Heal checkpoint could not persist its blocked marker"
-            ),
-        }
-    }
-
     /// create new checkpoint manager
     pub async fn new(disk: DiskStore, task_id: String) -> Result<Self> {
         validate_resume_task_id(&task_id)?;
-        let checkpoint_volume = format!("{RUSTFS_META_BUCKET}/{BUCKET_META_PREFIX}");
-        if let Err(error) = EcstoreDiskAPI::make_volume(disk.as_ref(), &checkpoint_volume).await
-            && error != crate::heal::DiskError::VolumeExists
-        {
-            return Err(Error::TaskExecutionFailed {
-                message: format!("Failed to create checkpoint volume: {error}"),
-            });
-        }
         let checkpoint = ResumeCheckpoint::new(task_id);
         let manager = Self {
             disk,
             checkpoint: Arc::new(RwLock::new(checkpoint)),
             throttle: Mutex::new(PersistThrottle::new()),
-            save_lock: AsyncMutex::new(()),
-            last_saved: Mutex::new(None),
         };
 
         // save initial checkpoint
@@ -232,7 +140,6 @@ impl CheckpointManager {
                 error = %e,
                 "Heal checkpoint persistence failed"
             );
-            return Err(e);
         }
         Ok(manager)
     }
@@ -241,22 +148,11 @@ impl CheckpointManager {
     pub async fn load_from_disk(disk: DiskStore, task_id: &str) -> Result<Self> {
         validate_resume_task_id(task_id)?;
         let checkpoint_data = Self::read_checkpoint_file(&disk, task_id).await?;
-        Self::load_from_data(disk, task_id, checkpoint_data).await
-    }
-
-    async fn load_from_data(disk: DiskStore, task_id: &str, checkpoint_data: Vec<u8>) -> Result<Self> {
-        validate_resume_task_id(task_id)?;
-        let mut checkpoint: ResumeCheckpoint = match serde_json::from_slice(&checkpoint_data) {
-            Ok(checkpoint) => checkpoint,
-            Err(error) => {
-                Self::block_invalid_snapshot(&disk, task_id).await;
-                return Err(Error::TaskExecutionFailed {
-                    message: format!("Failed to deserialize checkpoint: {error}"),
-                });
-            }
-        };
+        let mut checkpoint: ResumeCheckpoint =
+            serde_json::from_slice(&checkpoint_data).map_err(|e| Error::TaskExecutionFailed {
+                message: format!("Failed to deserialize checkpoint: {e}"),
+            })?;
         if checkpoint.task_id != task_id {
-            Self::block_invalid_snapshot(&disk, task_id).await;
             return Err(Error::TaskExecutionFailed {
                 message: "Resume checkpoint task id does not match filename".to_string(),
             });
@@ -267,7 +163,6 @@ impl CheckpointManager {
         // identities. Discard the stale sets and position, then stamp the
         // current schema so the scan restarts cleanly.
         if checkpoint.schema_version > CURRENT_CHECKPOINT_SCHEMA {
-            Self::block_invalid_snapshot(&disk, task_id).await;
             return Err(Error::TaskExecutionFailed {
                 message: format!(
                     "Checkpoint schema {} is newer than supported schema {CURRENT_CHECKPOINT_SCHEMA}",
@@ -299,8 +194,6 @@ impl CheckpointManager {
             disk,
             checkpoint: Arc::new(RwLock::new(checkpoint)),
             throttle: Mutex::new(PersistThrottle::new()),
-            save_lock: AsyncMutex::new(()),
-            last_saved: Mutex::new(Some(EcstoreDiskBytes::from(checkpoint_data))),
         })
     }
 
@@ -311,7 +204,7 @@ impl CheckpointManager {
         }
         let file_path = Path::new(BUCKET_META_PREFIX).join(format!("{task_id}_{RESUME_CHECKPOINT_FILE}"));
         match path_to_str(&file_path) {
-            Ok(path_str) => match HealDiskExt::read_all(disk.as_ref(), RUSTFS_META_BUCKET, path_str).await {
+            Ok(path_str) => match disk.read_all(RUSTFS_META_BUCKET, path_str).await {
                 Ok(data) => !data.is_empty(),
                 Err(_) => false,
             },
@@ -399,7 +292,6 @@ impl CheckpointManager {
 
         let checkpoint_file = Path::new(BUCKET_META_PREFIX).join(format!("{task_id}_{RESUME_CHECKPOINT_FILE}"));
         delete_resume_file(&self.disk, &checkpoint_file).await?;
-        delete_resume_file(&self.disk, &Self::blocked_path(&task_id)).await?;
 
         debug!(
             target: "rustfs::heal::resume",
@@ -415,126 +307,21 @@ impl CheckpointManager {
 
     /// save checkpoint to disk
     async fn save_checkpoint(&self) -> Result<()> {
-        // Serialize saves and take the snapshot only after acquiring the lock:
-        // a slower writer must not publish a snapshot taken before a newer one.
-        let _save_guard = self.save_lock.lock().await;
-        let checkpoint = self.checkpoint.read().await.clone();
+        let checkpoint = self.checkpoint.read().await;
         validate_resume_task_id(&checkpoint.task_id)?;
-        let checkpoint_data =
-            EcstoreDiskBytes::from(serde_json::to_vec(&checkpoint).map_err(|e| Error::TaskExecutionFailed {
-                message: format!("Failed to serialize checkpoint: {e}"),
-            })?);
+        let checkpoint_data = serde_json::to_vec(&*checkpoint).map_err(|e| Error::TaskExecutionFailed {
+            message: format!("Failed to serialize checkpoint: {e}"),
+        })?;
 
         let file_path = Path::new(BUCKET_META_PREFIX).join(format!("{}_{}", checkpoint.task_id, RESUME_CHECKPOINT_FILE));
 
         let path_str = path_to_str(&file_path)?;
-        let last_saved = self
-            .last_saved
-            .lock()
-            .map_err(|_| Error::TaskExecutionFailed {
-                message: "Checkpoint save state lock is poisoned; refusing to save".to_string(),
-            })?
-            .clone();
-        let update = EcstoreDiskAPI::compare_and_update_file(
-            self.disk.as_ref(),
-            RUSTFS_META_BUCKET,
-            path_str,
-            last_saved.clone(),
-            Some(checkpoint_data.clone()),
-        )
-        .await
-        .map_err(|e| Error::TaskExecutionFailed {
-            message: format!("Failed to save checkpoint: {e}"),
-        })?;
-
-        let expected = match update {
-            EcstoreConditionalFileUpdate::Updated => None,
-            EcstoreConditionalFileUpdate::Missing => {
-                return Err(Error::TaskExecutionFailed {
-                    message: "Checkpoint was removed after this manager saved it; refusing to recreate it".to_string(),
-                });
-            }
-            EcstoreConditionalFileUpdate::Mismatch => {
-                // A healthy manager normally completes the CAS above without
-                // another read or JSON parse. Inspect only after a mismatch so
-                // corruption and future schemas cannot be overwritten blindly.
-                let existing = match HealDiskExt::read_all(self.disk.as_ref(), RUSTFS_META_BUCKET, path_str).await {
-                    Ok(existing) => existing,
-                    Err(crate::heal::DiskError::FileNotFound) => {
-                        return Err(Error::TaskExecutionFailed {
-                            message: "Checkpoint was removed after this manager saved it; refusing to recreate it".to_string(),
-                        });
-                    }
-                    Err(error) => {
-                        return Err(Error::TaskExecutionFailed {
-                            message: format!("Failed to inspect checkpoint after CAS mismatch: {error}"),
-                        });
-                    }
-                };
-
-                if existing.is_empty() && last_saved.is_none() {
-                    Some(existing)
-                } else {
-                    let current: ResumeCheckpoint = match serde_json::from_slice(&existing) {
-                        Ok(current) => current,
-                        Err(error) => {
-                            Self::block_invalid_snapshot(&self.disk, &checkpoint.task_id).await;
-                            return Err(Error::TaskExecutionFailed {
-                                message: format!("Existing checkpoint is corrupt: {error}"),
-                            });
-                        }
-                    };
-                    if current.task_id != checkpoint.task_id {
-                        Self::block_invalid_snapshot(&self.disk, &checkpoint.task_id).await;
-                        return Err(Error::TaskExecutionFailed {
-                            message: "Existing checkpoint task id does not match filename".to_string(),
-                        });
-                    }
-                    if current.schema_version > CURRENT_CHECKPOINT_SCHEMA {
-                        Self::block_invalid_snapshot(&self.disk, &checkpoint.task_id).await;
-                        return Err(Error::TaskExecutionFailed {
-                            message: format!(
-                                "Existing checkpoint schema {} is newer than supported schema {CURRENT_CHECKPOINT_SCHEMA}",
-                                current.schema_version
-                            ),
-                        });
-                    }
-                    if last_saved.as_ref().is_none_or(|saved| saved.as_ref() != existing.as_ref()) {
-                        return Err(Error::TaskExecutionFailed {
-                            message: "Checkpoint changed since this manager loaded it; refusing to overwrite newer progress"
-                                .to_string(),
-                        });
-                    }
-                    Some(existing)
-                }
-            }
-        };
-
-        if let Some(expected) = expected {
-            match EcstoreDiskAPI::compare_and_update_file(
-                self.disk.as_ref(),
-                RUSTFS_META_BUCKET,
-                path_str,
-                Some(expected),
-                Some(checkpoint_data.clone()),
-            )
+        self.disk
+            .write_all(RUSTFS_META_BUCKET, path_str, checkpoint_data.into())
             .await
             .map_err(|e| Error::TaskExecutionFailed {
-                message: format!("Failed to save checkpoint after CAS mismatch: {e}"),
-            })? {
-                EcstoreConditionalFileUpdate::Updated => {}
-                EcstoreConditionalFileUpdate::Missing | EcstoreConditionalFileUpdate::Mismatch => {
-                    return Err(Error::TaskExecutionFailed {
-                        message: "Checkpoint changed while saving; refusing to overwrite newer progress".to_string(),
-                    });
-                }
-            }
-        }
-
-        let mut last_saved = self.last_saved.lock().map_err(|_| Error::TaskExecutionFailed {
-            message: "Checkpoint save state lock is poisoned after save".to_string(),
-        })?;
-        *last_saved = Some(checkpoint_data);
+                message: format!("Failed to save checkpoint: {e}"),
+            })?;
 
         debug!(
             target: "rustfs::heal::resume",
@@ -554,7 +341,7 @@ impl CheckpointManager {
         let file_path = Path::new(BUCKET_META_PREFIX).join(format!("{task_id}_{RESUME_CHECKPOINT_FILE}"));
 
         let path_str = path_to_str(&file_path)?;
-        HealDiskExt::read_all(disk.as_ref(), RUSTFS_META_BUCKET, path_str)
+        disk.read_all(RUSTFS_META_BUCKET, path_str)
             .await
             .map(|bytes| bytes.to_vec())
             .map_err(|e| Error::TaskExecutionFailed {

crates/heal/src/heal/resume/tests.rs

@@ -1675,264 +1675,6 @@ async fn future_resume_and_checkpoint_schemas_are_rejected() {
     temp_dir.close().expect("remove schema test directory");
 }
 
-#[tokio::test]
-async fn checkpoint_save_does_not_replace_a_non_empty_truncated_snapshot() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let manager = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("create checkpoint manager");
-    let checkpoint_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_FILE}");
-    let truncated = b"{\"schema_version\":5,\"task_id\":";
-    disk.write_all(RUSTFS_META_BUCKET, &checkpoint_path, truncated.as_slice().into())
-        .await
-        .expect("write truncated checkpoint fixture");
-
-    let error = manager
-        .update_position(2, 7)
-        .await
-        .expect_err("a truncated checkpoint must fail closed during save");
-    assert!(error.to_string().contains("Existing checkpoint is corrupt"));
-    assert_eq!(
-        disk.read_all(RUSTFS_META_BUCKET, &checkpoint_path)
-            .await
-            .expect("read truncated checkpoint fixture"),
-        truncated.as_slice()
-    );
-    assert!(CheckpointManager::is_blocked(&disk, &task_id).await);
-    temp_dir.close().expect("remove checkpoint save test directory");
-}
-
-#[tokio::test]
-async fn checkpoint_save_does_not_replace_a_future_schema_snapshot() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let manager = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("create checkpoint manager");
-    let checkpoint_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_FILE}");
-    let mut future = ResumeCheckpoint::new(task_id.clone());
-    future.schema_version = CURRENT_CHECKPOINT_SCHEMA + 1;
-    let future_bytes = serde_json::to_vec(&future).expect("serialize future checkpoint fixture");
-    disk.write_all(RUSTFS_META_BUCKET, &checkpoint_path, future_bytes.clone().into())
-        .await
-        .expect("write future checkpoint fixture");
-
-    let error = manager
-        .update_position(2, 7)
-        .await
-        .expect_err("a future schema must fail closed during save");
-    assert!(error.to_string().contains("Existing checkpoint schema"));
-    assert_eq!(
-        disk.read_all(RUSTFS_META_BUCKET, &checkpoint_path)
-            .await
-            .expect("read future checkpoint fixture"),
-        future_bytes
-    );
-    assert!(CheckpointManager::is_blocked(&disk, &task_id).await);
-    temp_dir.close().expect("remove future schema test directory");
-}
-
-#[tokio::test]
-async fn new_checkpoint_manager_rebuilds_an_empty_snapshot() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let checkpoint_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_FILE}");
-    disk.write_all(RUSTFS_META_BUCKET, &checkpoint_path, EcstoreDiskBytes::new())
-        .await
-        .expect("write empty checkpoint fixture");
-
-    let manager = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("a new manager must rebuild an empty checkpoint");
-    manager
-        .update_position(3, 11)
-        .await
-        .expect("rebuilt checkpoint must remain writable");
-    assert!(CheckpointManager::has_checkpoint(&disk, &task_id).await);
-    temp_dir.close().expect("remove empty checkpoint test directory");
-}
-
-#[tokio::test]
-async fn deleted_checkpoint_is_not_recreated_by_an_old_manager() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let manager = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("create checkpoint manager");
-    manager.cleanup().await.expect("delete checkpoint fixture");
-
-    let error = manager
-        .update_position(1, 2)
-        .await
-        .expect_err("an old manager must not resurrect a deleted checkpoint");
-    assert!(error.to_string().contains("removed after this manager saved it"));
-    assert!(!CheckpointManager::has_checkpoint(&disk, &task_id).await);
-    temp_dir.close().expect("remove deleted checkpoint test directory");
-}
-
-#[tokio::test]
-async fn an_empty_blocked_marker_still_blocks_resume_selection() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let manager = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("create checkpoint manager");
-    let blocked_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_BLOCKED_FILE}");
-    disk.write_all(RUSTFS_META_BUCKET, &blocked_path, EcstoreDiskBytes::new())
-        .await
-        .expect("write empty blocked marker fixture");
-
-    assert!(CheckpointManager::is_blocked(&disk, &task_id).await);
-    assert!(!CheckpointManager::is_resumable(&disk, &task_id).await);
-    // Recovery requires replacing/cleaning the snapshot, then removing the
-    // marker; ordinary selector retries are intentionally not an unlock path.
-    manager.cleanup().await.expect("clean blocked checkpoint");
-    assert!(!CheckpointManager::is_blocked(&disk, &task_id).await);
-    temp_dir.close().expect("remove empty blocked marker test directory");
-}
-
-#[tokio::test]
-async fn resumable_selector_skips_healthy_tasks_with_blocked_markers() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let tasks = [
-        (ResumeUtils::generate_task_id(), EcstoreDiskBytes::new()),
-        (ResumeUtils::generate_task_id(), EcstoreDiskBytes::from_static(b"blocked")),
-    ];
-    for (task_id, marker) in &tasks {
-        ResumeManager::new(
-            disk.clone(),
-            task_id.clone(),
-            "erasure_set".to_string(),
-            "pool_0_set_0".to_string(),
-            vec!["bucket".to_string()],
-        )
-        .await
-        .expect("create healthy resume state");
-        CheckpointManager::new(disk.clone(), task_id.clone())
-            .await
-            .expect("create healthy checkpoint");
-        let checkpoint_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_FILE}");
-        let checkpoint_bytes = disk
-            .read_all(RUSTFS_META_BUCKET, &checkpoint_path)
-            .await
-            .expect("read healthy checkpoint before blocking");
-        let marker_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_BLOCKED_FILE}");
-        disk.write_all(RUSTFS_META_BUCKET, &marker_path, marker.clone())
-            .await
-            .expect("write blocked marker");
-
-        assert!(
-            ResumeUtils::get_resumable_tasks(&disk)
-                .await
-                .expect("filter blocked healthy task")
-                .is_empty()
-        );
-        assert_eq!(
-            disk.read_all(RUSTFS_META_BUCKET, &checkpoint_path)
-                .await
-                .expect("read healthy checkpoint after blocking"),
-            checkpoint_bytes
-        );
-    }
-    temp_dir.close().expect("remove blocked selector test directory");
-}
-
-#[tokio::test]
-async fn stale_checkpoint_manager_cannot_overwrite_newer_progress() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let task_id = ResumeUtils::generate_task_id();
-    let first = CheckpointManager::new(disk.clone(), task_id.clone())
-        .await
-        .expect("create first checkpoint manager");
-    let second = CheckpointManager::load_from_disk(disk.clone(), &task_id)
-        .await
-        .expect("load second checkpoint manager");
-
-    second
-        .update_position(4, 20)
-        .await
-        .expect("persist newer checkpoint progress");
-    let error = first
-        .update_position(1, 3)
-        .await
-        .expect_err("stale checkpoint manager must not overwrite newer progress");
-    assert!(error.to_string().contains("newer progress"));
-
-    let persisted = CheckpointManager::load_from_disk(disk.clone(), &task_id)
-        .await
-        .expect("load newer checkpoint progress")
-        .get_checkpoint()
-        .await;
-    assert_eq!(persisted.current_bucket_index, 4);
-    assert_eq!(persisted.current_object_index, 20);
-    temp_dir.close().expect("remove stale manager test directory");
-}
-
-#[tokio::test]
-async fn resumable_selector_isolates_future_and_corrupt_checkpoints() {
-    let (temp_dir, disk) = schema_test_disk().await;
-    let future_task = ResumeUtils::generate_task_id();
-    let corrupt_task = ResumeUtils::generate_task_id();
-    for task_id in [&future_task, &corrupt_task] {
-        ResumeManager::new(
-            disk.clone(),
-            task_id.to_string(),
-            "erasure_set".to_string(),
-            "pool_0_set_0".to_string(),
-            vec!["bucket".to_string()],
-        )
-        .await
-        .expect("create resumable state fixture");
-    }
-
-    let future_path = format!("{BUCKET_META_PREFIX}/{future_task}_{RESUME_CHECKPOINT_FILE}");
-    let mut future = ResumeCheckpoint::new(future_task.clone());
-    future.schema_version = CURRENT_CHECKPOINT_SCHEMA + 1;
-    let future_bytes = serde_json::to_vec(&future).expect("serialize future checkpoint fixture");
-    disk.write_all(RUSTFS_META_BUCKET, &future_path, future_bytes.clone().into())
-        .await
-        .expect("write future checkpoint fixture");
-    let corrupt_path = format!("{BUCKET_META_PREFIX}/{corrupt_task}_{RESUME_CHECKPOINT_FILE}");
-    let corrupt_bytes = b"{truncated";
-    disk.write_all(RUSTFS_META_BUCKET, &corrupt_path, corrupt_bytes.as_slice().into())
-        .await
-        .expect("write corrupt checkpoint fixture");
-
-    assert!(
-        ResumeUtils::get_resumable_tasks(&disk)
-            .await
-            .expect("filter malformed resumable tasks")
-            .is_empty()
-    );
-    assert!(
-        ResumeUtils::get_resumable_tasks(&disk)
-            .await
-            .expect("filter blocked resumable tasks")
-            .is_empty()
-    );
-    for (task_id, path, bytes) in [
-        (&future_task, future_path, future_bytes),
-        (&corrupt_task, corrupt_path, corrupt_bytes.to_vec()),
-    ] {
-        assert_eq!(
-            disk.read_all(RUSTFS_META_BUCKET, &path)
-                .await
-                .expect("read isolated checkpoint bytes"),
-            bytes
-        );
-        let blocked_path = format!("{BUCKET_META_PREFIX}/{task_id}_{RESUME_CHECKPOINT_BLOCKED_FILE}");
-        assert!(
-            !disk
-                .read_all(RUSTFS_META_BUCKET, &blocked_path)
-                .await
-                .expect("read checkpoint blocked marker")
-                .is_empty()
-        );
-    }
-    temp_dir.close().expect("remove selector isolation test directory");
-}
-
 #[test]
 fn test_persist_throttle_batches_until_threshold() {
     let mut throttle = PersistThrottle::new();

crates/heal/src/heal/resume/utils.rs

@@ -21,7 +21,7 @@ use uuid::Uuid;
 use super::super::{BUCKET_META_PREFIX, DiskError, DiskStore, HealDiskExt as _, RUSTFS_META_BUCKET};
 use super::replacement::{ReplacementPhase, ReplacementRecoveryRecord};
 use super::{
-    CheckpointManager, EVENT_HEAL_RESUME_STATE, LOG_COMPONENT_HEAL, LOG_SUBSYSTEM_RESUME, REPLACEMENT_COMPLETION_PROOF_FILE,
+    EVENT_HEAL_RESUME_STATE, LOG_COMPONENT_HEAL, LOG_SUBSYSTEM_RESUME, REPLACEMENT_COMPLETION_PROOF_FILE,
     REPLACEMENT_INTENT_FILE, RESUME_STATE_FILE, ResumeManager, ResumeStateFile, is_replacement_intent, path_to_str,
     replacement_recovery_corruption_for_state_load, replacement_recovery_dir, validate_resume_task_id,
 };
@@ -67,7 +67,6 @@ impl ResumeUtils {
                 // Extract task ID from filename: {task_id}_ahm_resume_state.json
                 if let Some(task_id) = entry.strip_suffix(&format!("_{RESUME_STATE_FILE}"))
                     && validate_resume_task_id(task_id).is_ok()
-                    && CheckpointManager::is_resumable(disk, task_id).await
                 {
                     task_ids.push(task_id.to_string());
                 }