Merge branch 'main' into fix/b5-t1-a1-replication-deny-edit

Replace the upstream xonatius/mimalloc_rust.git fork (mimalloc + libmimalloc-sys)
with the published rustfs-mimalloc (v0.5.0) and rustfs-mimalloc-sys (v0.5.0) crates
from crates.io.

The new crates are based on mimalloc V3 (v3.5.0) and provide:
- MiMalloc global allocator with safe API (collect, stats_json, process_info)
- Heap management and arena operations (heap module)
- Full FFI bindings to mimalloc V3

Changes:
- Workspace deps: mimalloc + libmimalloc-sys (git) → rustfs-mimalloc + rustfs-mimalloc-sys (crates.io)
- allocator_reclaim.rs: libmimalloc_sys::mi_collect → rustfs_mimalloc::MiMalloc::collect
- memory_observability.rs: raw FFI mi_stats_get_json → MiMalloc::stats_json()
- main.rs: heap ownership tests use Heap::contains() (V3 API)
- deny.toml: remove xonatius/mimalloc_rust.git from allow-git

Co-authored-by: heihutu <heihutu@gmail.com>

CLAUDE.md

@@ -30,7 +30,8 @@ make build-docker BUILD_OS=ubuntu22.04
 - Crate membership: `Cargo.toml` `[workspace].members`
 - Architecture, layering, crate map: [ARCHITECTURE.md](ARCHITECTURE.md)
 - Migration guardrails & readiness contracts: [docs/architecture/](docs/architecture/README.md)
-- CI gates: `.github/workflows/ci.yml` (source of truth; never copy its steps into docs)
+- CI workflow steps: `.github/workflows/`; event, timeout, and required-status
+  matrix: [docs/testing/ci-gates.md](docs/testing/ci-gates.md)
 - Test-layer taxonomy, per-layer entry commands, serial/nextest rules, flake
   policy: [docs/testing/README.md](docs/testing/README.md)
 - Tier/ILM transition debugging (xl.meta inspection, versionId tracing):

CONTRIBUTING.md

@@ -70,6 +70,8 @@ make pre-pr
 
 > For the full test-layer taxonomy (unit / ecstore black-box / e2e / s3s-e2e / S3 compatibility / chaos / fuzz / bench), each layer's entry command, the naming conventions the migration gate depends on, and the serial/nextest rules, see [docs/testing/README.md](docs/testing/README.md).
 
+> For the event, timeout, required-status, and local reproduction matrix, see [docs/testing/ci-gates.md](docs/testing/ci-gates.md).
+
 ### 🔒 Automated Pre-commit Hooks
 #### What `make pre-commit` and `make pre-pr` actually run
 

Cargo.lock

@@ -1858,9 +1858,9 @@ dependencies = [
 
 [[package]]
 name = "cc"
-version = "1.4.3"
+version = "1.4.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "509591b7bcd67f4ef775afad7662703b4935daaa6ec0e5605cfb1090b32a2b6d"
+checksum = "0ad534f4357a5264cce5019c989cf66a4f0dc4e0d1b1d15f8aacec0ff7360273"
 dependencies = [
  "find-msvc-tools",
  "jobserver",
@@ -2522,12 +2522,6 @@ dependencies = [
  "subtle",
 ]
 
-[[package]]
-name = "cty"
-version = "0.2.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b365fabc795046672053e29c954733ec3b05e4be654ab130fe8f1f94d7051f35"
-
 [[package]]
 name = "curve25519-dalek"
 version = "4.1.3"
@@ -5988,15 +5982,6 @@ version = "0.2.16"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b6d2cec3eae94f9f509c767b45932f1ada8350c4bdb85af2fcab4a3c14807981"
 
-[[package]]
-name = "libmimalloc-sys"
-version = "0.1.49"
-source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11#6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11"
-dependencies = [
- "cc",
- "cty",
-]
-
 [[package]]
 name = "libredox"
 version = "0.1.20"
@@ -6397,14 +6382,6 @@ dependencies = [
  "synstructure 0.13.2",
 ]
 
-[[package]]
-name = "mimalloc"
-version = "0.1.52"
-source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11#6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11"
-dependencies = [
- "libmimalloc-sys",
-]
-
 [[package]]
 name = "mime"
 version = "0.3.17"
@@ -9162,13 +9139,11 @@ dependencies = [
  "insta",
  "jiff",
  "libc",
- "libmimalloc-sys",
  "libsystemd",
  "matchit 0.9.2",
  "md-5 0.11.0",
  "metrics",
  "metrics-util",
- "mimalloc",
  "mime_guess",
  "opentelemetry",
  "opentelemetry_sdk",
@@ -9204,6 +9179,8 @@ dependencies = [
  "rustfs-lock",
  "rustfs-log-analyzer",
  "rustfs-madmin",
+ "rustfs-mimalloc",
+ "rustfs-mimalloc-sys",
  "rustfs-notify",
  "rustfs-object-capacity",
  "rustfs-object-data-cache",
@@ -9875,6 +9852,24 @@ dependencies = [
  "tokio",
 ]
 
+[[package]]
+name = "rustfs-mimalloc"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a406f4aa07084301d485beec873af6dccc8e3f8762da244743df92038b1db1a6"
+dependencies = [
+ "rustfs-mimalloc-sys",
+]
+
+[[package]]
+name = "rustfs-mimalloc-sys"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c3051b819175f58445d4c369a72f0ab88149f3885ba8bea2aff3be01f53fe7cd"
+dependencies = [
+ "cc",
+]
+
 [[package]]
 name = "rustfs-notify"
 version = "1.0.0-rc.3"

Cargo.toml

@@ -350,8 +350,8 @@ russh-sftp = "2.4.0"
 dav-server = "0.11.0"
 
 # Performance Analysis and Memory Profiling
-mimalloc = { version = "0.1.52", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11" }
-libmimalloc-sys = { version = "0.1.49", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11", features = ["extended"] }
+rustfs-mimalloc = { version = "0.5.0" }
+rustfs-mimalloc-sys = { version = "0.5.0" }
 hotpath = { version = "0.23.3", default-features = false }
 # Snapshot testing for output format regression detection
 insta = { version = "1.48" }

crates/data-usage/src/data_usage.rs

@@ -236,15 +236,6 @@ pub struct DataUsageInfo {
     /// without relying on synchronized clocks.
     #[serde(default, skip_serializing_if = "Option::is_none")]
     pub usage_snapshot_authoritative_baseline: Option<DataUsageSnapshotIdentity>,
-    /// Per-set freshness for an observational aggregate.  A set entry is
-    /// never sufficient to make the aggregate authoritative; it only records
-    /// which last-known-good generation contributed to the view.
-    #[serde(default, skip_serializing_if = "Vec::is_empty")]
-    pub usage_snapshot_set_states: Vec<DataUsageSnapshotSetState>,
-    /// An observational view may contain only the sets that completed this
-    /// cycle (or retained a compatible last-known-good cache).
-    #[serde(default)]
-    pub usage_snapshot_partial: bool,
     /// Deprecated kept here for backward compatibility reasons
     pub bucket_sizes: HashMap<String, u64>,
     /// Per-disk snapshot information when available
@@ -261,22 +252,6 @@ pub struct DataUsageSnapshotIdentity {
     pub scanner_epoch: Option<u64>,
 }
 
-#[derive(Clone, Debug, Default, Serialize, Deserialize, PartialEq, Eq)]
-pub struct DataUsageSnapshotSetState {
-    pub pool_index: u64,
-    pub set_index: u64,
-    #[serde(default)]
-    pub scanner_cycle: Option<u64>,
-    #[serde(default)]
-    pub scanner_epoch: Option<u64>,
-    #[serde(default, skip_serializing_if = "Option::is_none")]
-    pub scan_plan_digest: Option<[u8; 32]>,
-    #[serde(default)]
-    pub complete: bool,
-    #[serde(default)]
-    pub tombstone: bool,
-}
-
 impl DataUsageInfo {
     pub fn snapshot_identity(&self) -> DataUsageSnapshotIdentity {
         DataUsageSnapshotIdentity {
@@ -316,7 +291,7 @@ pub fn data_usage_snapshot_is_newer(candidate: &DataUsageInfo, baseline: &DataUs
 /// rollback delete/recreate fences the previous bucket incarnation too.
 pub fn observed_data_usage_is_newer(observed: &DataUsageInfo, authoritative: &DataUsageInfo) -> bool {
     observed.usage_snapshot_converged == Some(false)
-        && (observed.is_complete_bucket_usage_snapshot() || observed.is_valid_partial_snapshot())
+        && observed.is_complete_bucket_usage_snapshot()
         && observed.usage_snapshot_authoritative_baseline.as_ref() == Some(&authoritative.snapshot_identity())
         && data_usage_snapshot_is_newer(observed, authoritative)
 }
@@ -1461,39 +1436,6 @@ impl DataUsageInfo {
             && u64::try_from(self.buckets_usage.len()).ok() == Some(self.buckets_count)
     }
 
-    /// Validate provenance before an observational view can be selected for
-    /// admin display. Partial data is accepted only with unique set states,
-    /// a plan digest for every state, and at least one usable generation.
-    pub fn is_valid_partial_snapshot(&self) -> bool {
-        if !self.usage_snapshot_partial
-            || self.usage_snapshot_converged != Some(false)
-            || self.last_update.is_none()
-            || self.scanner_cycle.is_none()
-            || self.scanner_epoch.is_none()
-            || self.usage_snapshot_set_states.is_empty()
-            || u64::try_from(self.buckets_usage.len()).ok() != Some(self.buckets_count)
-        {
-            return false;
-        }
-
-        let mut previous = None;
-        let mut plan_digest = None;
-        let mut has_source = false;
-        for state in &self.usage_snapshot_set_states {
-            if state.scan_plan_digest.is_none()
-                || plan_digest.is_some_and(|digest| Some(digest) != state.scan_plan_digest)
-                || state.scanner_cycle.is_some() != state.scanner_epoch.is_some()
-                || previous.is_some_and(|(pool, set)| (pool, set) >= (state.pool_index, state.set_index))
-            {
-                return false;
-            }
-            previous = Some((state.pool_index, state.set_index));
-            plan_digest = state.scan_plan_digest;
-            has_source |= state.scanner_cycle.is_some() && !state.tombstone;
-        }
-        has_source
-    }
-
     /// Add object metadata to data usage statistics
     pub fn add_object(&mut self, object_path: &str, meta_object: &rustfs_filemeta::MetaObject) {
         // This method is kept for backward compatibility
@@ -2321,55 +2263,6 @@ mod tests {
         assert!(!observed_data_usage_is_newer(&candidate(2, 9, Some(false), true), &authoritative));
         assert!(!observed_data_usage_is_newer(&candidate(2, 11, Some(true), true), &authoritative));
         assert!(!observed_data_usage_is_newer(&candidate(2, 11, Some(false), false), &authoritative));
-
-        let mut partial = candidate(2, 11, Some(false), false);
-        partial.usage_snapshot_partial = true;
-        partial.usage_snapshot_set_states = vec![DataUsageSnapshotSetState {
-            pool_index: 0,
-            set_index: 0,
-            scanner_cycle: Some(10),
-            scanner_epoch: Some(2),
-            scan_plan_digest: Some([1; 32]),
-            complete: false,
-            tombstone: false,
-        }];
-        assert!(observed_data_usage_is_newer(&partial, &authoritative));
-    }
-
-    #[test]
-    fn mixed_topology_snapshot_is_rejected() {
-        let mut partial = DataUsageInfo {
-            last_update: Some(SystemTime::UNIX_EPOCH + Duration::from_secs(2)),
-            scanner_cycle: Some(11),
-            scanner_epoch: Some(2),
-            buckets_count: 0,
-            usage_snapshot_converged: Some(false),
-            usage_snapshot_partial: true,
-            usage_snapshot_set_states: vec![
-                DataUsageSnapshotSetState {
-                    pool_index: 0,
-                    set_index: 0,
-                    scanner_cycle: Some(11),
-                    scanner_epoch: Some(2),
-                    scan_plan_digest: Some([1; 32]),
-                    complete: true,
-                    tombstone: false,
-                },
-                DataUsageSnapshotSetState {
-                    pool_index: 1,
-                    set_index: 0,
-                    scanner_cycle: Some(10),
-                    scanner_epoch: Some(2),
-                    scan_plan_digest: Some([2; 32]),
-                    complete: false,
-                    tombstone: false,
-                },
-            ],
-            ..Default::default()
-        };
-        assert!(!partial.is_valid_partial_snapshot());
-        partial.usage_snapshot_set_states[1].scan_plan_digest = Some([1; 32]);
-        assert!(partial.is_valid_partial_snapshot());
     }
 
     #[test]

crates/ecstore/src/api/mod.rs

@@ -196,15 +196,16 @@ pub mod bucket {
             ReplicationOperation, ReplicationPoolTrait, ReplicationPriority, ReplicationQueueAdmission, ReplicationScannerBridge,
             ReplicationState, ReplicationStats, ReplicationStatusType, ReplicationStorage, ReplicationTargetValidationError,
             ReplicationType, ResyncOpts, ResyncStatusType, RuntimeReplicationTargetBacklog, TargetReplicationResyncStatus,
-            VersionPurgeStatusType, XferStats, commit_force_delete_intent, complete_force_delete_intent,
-            delete_replication_state_from_config, delete_replication_version_id, get_global_replication_pool,
-            get_global_replication_stats, get_proxy_targets, init_background_replication,
-            invalid_replication_config_status_field, persist_force_delete_intent, read_durable_mrf_backlog,
-            replication_state_to_filemeta, replication_status_to_filemeta, replication_statuses_map, replication_target_arns,
-            resync_start_conflict_id, should_remove_replication_target, should_schedule_delete_replication,
-            should_use_existing_delete_replication_info, should_use_existing_delete_replication_source,
-            unsupported_replication_config_field, validate_replication_config_structure, validate_replication_config_target_arns,
-            version_purge_status_to_filemeta,
+            VersionPurgeStatusType, XferStats, assign_site_replication_rule_priorities, commit_force_delete_intent,
+            complete_force_delete_intent, delete_replication_state_from_config, delete_replication_version_id,
+            get_global_replication_pool, get_global_replication_stats, get_proxy_targets, init_background_replication,
+            invalid_replication_config_status_field, is_site_replication_rule, merge_incoming_replication_config,
+            merge_user_replication_config, persist_force_delete_intent, read_durable_mrf_backlog, replication_state_to_filemeta,
+            replication_status_to_filemeta, replication_statuses_map, replication_target_arn_deployment_id,
+            replication_target_arns, resync_start_conflict_id, should_remove_replication_target,
+            should_schedule_delete_replication, should_use_existing_delete_replication_info,
+            should_use_existing_delete_replication_source, unsupported_replication_config_field,
+            validate_replication_config_structure, validate_replication_config_target_arns, version_purge_status_to_filemeta,
         };
     }
 

crates/ecstore/src/bucket/replication/mod.rs

@@ -47,8 +47,10 @@ pub use replication_config_boundary::{
     ObjectOpts, REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION, REMOTE_TARGET_UNSUPPORTED_FIELDS, REMOTE_TARGET_WRITABLE_FIELDS,
     REPLICATION_CAPABILITY_CONTRACT_VERSION, REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS,
     ReplicationConfigStructureError, ReplicationConfigurationExt, ReplicationTargetValidationError,
-    invalid_replication_config_status_field, replication_target_arns, should_remove_replication_target,
-    unsupported_replication_config_field, validate_replication_config_structure, validate_replication_config_target_arns,
+    assign_site_replication_rule_priorities, invalid_replication_config_status_field, is_site_replication_rule,
+    merge_incoming_replication_config, merge_user_replication_config, replication_target_arn_deployment_id,
+    replication_target_arns, should_remove_replication_target, unsupported_replication_config_field,
+    validate_replication_config_structure, validate_replication_config_target_arns,
 };
 pub(crate) use replication_filemeta_boundary::version_purge_statuses_map;
 pub use replication_filemeta_boundary::{

crates/ecstore/src/bucket/replication/replication_config_boundary.rs

@@ -16,6 +16,8 @@ pub use rustfs_replication::{
     ObjectOpts, REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION, REMOTE_TARGET_UNSUPPORTED_FIELDS, REMOTE_TARGET_WRITABLE_FIELDS,
     REPLICATION_CAPABILITY_CONTRACT_VERSION, REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS,
     ReplicationConfigStructureError, ReplicationConfigurationExt, ReplicationRuleExt, ReplicationTargetValidationError,
-    invalid_replication_config_status_field, replication_target_arns, should_remove_replication_target,
-    unsupported_replication_config_field, validate_replication_config_structure, validate_replication_config_target_arns,
+    assign_site_replication_rule_priorities, invalid_replication_config_status_field, is_site_replication_rule,
+    merge_incoming_replication_config, merge_user_replication_config, replication_target_arn_deployment_id,
+    replication_target_arns, should_remove_replication_target, unsupported_replication_config_field,
+    validate_replication_config_structure, validate_replication_config_target_arns,
 };

crates/ecstore/src/data_usage/mod.rs

@@ -73,16 +73,6 @@ struct CachedBucketUsage {
     // mutation. A strictly later generation is required before the mutation
     // evidence can be discarded.
     pending_scanner_position: Option<(u64, u64)>,
-    // Deletes are visible to admin immediately, but quota admission keeps
-    // them pending until a complete scanner generation reconciles the set.
-    // This marker intentionally remains process-local: the delete request
-    // updates this overlay before the scanner writes a durable snapshot. If
-    // the process restarts first, loading the persisted complete snapshot
-    // restores the pre-reconciliation (larger) baseline, which is
-    // conservative for quota admission. A persisted post-delete snapshot is
-    // necessarily a complete scanner reconciliation and therefore creates a
-    // fresh cache entry with no pending hold.
-    pending_negative_delta: u64,
 }
 
 type UsageMemoryCache = Arc<RwLock<HashMap<String, CachedBucketUsage>>>;
@@ -958,12 +948,7 @@ async fn load_observed_data_usage_snapshot(store: Arc<ECStore>) -> Option<DataUs
     };
 
     match parse_usage_snapshot(&data) {
-        Ok(info)
-            if info.usage_snapshot_converged == Some(false)
-                && (info.is_complete_bucket_usage_snapshot() || info.is_valid_partial_snapshot()) =>
-        {
-            Some(info)
-        }
+        Ok(info) if info.usage_snapshot_converged == Some(false) && info.is_complete_bucket_usage_snapshot() => Some(info),
         Ok(_) => {
             error!(
                 event = "data_usage_snapshot_load_failed",
@@ -1008,7 +993,7 @@ async fn load_admin_data_usage_from_backend(store: Arc<ECStore>) -> Result<DataU
 }
 
 fn discard_incomplete_bucket_usage(data_usage_info: &mut DataUsageInfo) {
-    if !data_usage_info.is_complete_bucket_usage_snapshot() && !data_usage_info.usage_snapshot_partial {
+    if !data_usage_info.is_complete_bucket_usage_snapshot() {
         data_usage_info.usage_snapshot_complete = false;
         data_usage_info.buckets_usage.clear();
         data_usage_info.bucket_sizes.clear();
@@ -1658,7 +1643,6 @@ fn cached_bucket_usage_from_backend(usage: BucketUsageInfo, updated_at: SystemTi
         dirty: false,
         stale_snapshot_pending: false,
         pending_scanner_position: None,
-        pending_negative_delta: 0,
     }
 }
 
@@ -1672,7 +1656,6 @@ fn cached_bucket_usage_now(usage: BucketUsageInfo) -> CachedBucketUsage {
         dirty: false,
         stale_snapshot_pending: false,
         pending_scanner_position: None,
-        pending_negative_delta: 0,
     }
 }
 
@@ -1825,7 +1808,6 @@ pub async fn record_bucket_object_delete_memory(bucket: &str, deleted_size: u64,
         .or_insert_with(|| cached_bucket_usage_now(BucketUsageInfo::default()));
 
     entry.usage.size = entry.usage.size.saturating_sub(deleted_size);
-    entry.pending_negative_delta = entry.pending_negative_delta.saturating_add(deleted_size);
     if removed_current_object {
         entry.usage.objects_count = entry.usage.objects_count.saturating_sub(1);
         entry.usage.versions_count = entry.usage.versions_count.saturating_sub(1);
@@ -1881,7 +1863,7 @@ pub async fn get_bucket_usage_memory(bucket: &str) -> Option<u64> {
     cache
         .get(bucket)
         .filter(|cached| cached.authoritative)
-        .map(|cached| cached.usage.size.saturating_add(cached.pending_negative_delta))
+        .map(|cached| cached.usage.size)
 }
 
 async fn update_usage_cache_if_needed() {
@@ -2961,45 +2943,6 @@ mod tests {
         assert_eq!(selected.usage_snapshot_converged, Some(true));
     }
 
-    #[test]
-    fn persisted_authoritative_stalls_but_memory_overlay_remains_visible() {
-        let authoritative = DataUsageInfo {
-            last_update: Some(SystemTime::UNIX_EPOCH),
-            scanner_epoch: Some(4),
-            scanner_cycle: Some(10),
-            usage_snapshot_complete: true,
-            ..Default::default()
-        };
-        let mut partial = authoritative.clone();
-        partial.last_update = Some(SystemTime::UNIX_EPOCH + Duration::from_secs(1));
-        partial.scanner_cycle = Some(11);
-        partial.usage_snapshot_complete = false;
-        partial.usage_snapshot_partial = true;
-        partial.usage_snapshot_converged = Some(false);
-        partial.usage_snapshot_authoritative_baseline = Some(authoritative.snapshot_identity());
-        partial.usage_snapshot_set_states = vec![rustfs_data_usage::DataUsageSnapshotSetState {
-            pool_index: 0,
-            set_index: 0,
-            scanner_cycle: Some(10),
-            scanner_epoch: Some(4),
-            scan_plan_digest: Some([1; 32]),
-            complete: false,
-            tombstone: false,
-        }];
-        partial.buckets_usage.insert(
-            "bucket".to_string(),
-            BucketUsageInfo {
-                size: 100,
-                ..Default::default()
-            },
-        );
-        partial.buckets_count = 1;
-
-        let (selected, _) = select_admin_data_usage_snapshot(authoritative, true, Some(partial));
-        assert!(selected.usage_snapshot_partial);
-        assert_eq!(selected.buckets_usage.get("bucket").map(|usage| usage.size), Some(100));
-    }
-
     #[tokio::test]
     async fn authoritative_save_cleanup_removes_observed_snapshot_best_effort() {
         let store = UsageCasStore::default();
@@ -4722,55 +4665,6 @@ mod tests {
         );
     }
 
-    #[tokio::test]
-    #[serial]
-    async fn partial_usage_is_observational_not_authoritative_for_quota() {
-        clear_usage_memory_cache_for_test().await;
-
-        let mut partial = data_usage_info_for_test("bucket-a", 10, 100, SystemTime::now());
-        partial.usage_snapshot_complete = false;
-        partial.usage_snapshot_partial = true;
-        replace_bucket_usage_memory_from_info(&partial).await;
-
-        assert_eq!(get_bucket_usage_memory("bucket-a").await, None);
-    }
-
-    #[tokio::test]
-    #[serial]
-    async fn stale_quota_uses_complete_baseline_plus_positive_deltas() {
-        clear_usage_memory_cache_for_test().await;
-
-        let baseline = data_usage_info_for_test("bucket-a", 1, 100, SystemTime::now());
-        replace_bucket_usage_memory_from_info(&baseline).await;
-        record_bucket_object_write_memory("bucket-a", None, 25).await;
-
-        assert_eq!(get_bucket_usage_memory("bucket-a").await, Some(125));
-    }
-
-    #[tokio::test]
-    #[serial]
-    async fn negative_delta_waits_for_set_reconciliation() {
-        clear_usage_memory_cache_for_test().await;
-
-        let baseline = data_usage_info_for_test("bucket-a", 1, 100, SystemTime::UNIX_EPOCH + Duration::from_secs(100));
-        replace_bucket_usage_memory_from_info(&baseline).await;
-        record_bucket_object_delete_memory("bucket-a", 25, true).await;
-
-        assert_eq!(get_bucket_usage_memory("bucket-a").await, Some(100));
-
-        // Simulate a process restart: the request-path overlay is gone, but
-        // the persisted authoritative snapshot is still the pre-reconciliation
-        // baseline. Quota must remain conservative until a complete scanner
-        // result proves the delete.
-        clear_usage_memory_cache_for_test().await;
-        replace_bucket_usage_memory_from_info(&baseline).await;
-        assert_eq!(get_bucket_usage_memory("bucket-a").await, Some(100));
-
-        let reconciled = data_usage_info_for_test("bucket-a", 0, 75, SystemTime::UNIX_EPOCH + Duration::from_secs(101));
-        replace_bucket_usage_memory_from_info(&reconciled).await;
-        assert_eq!(get_bucket_usage_memory("bucket-a").await, Some(75));
-    }
-
     #[tokio::test]
     #[serial]
     async fn memory_overlay_counts_versioned_overwrite_as_new_version() {

crates/ecstore/src/set_disk/ops/object.rs

@@ -2124,26 +2124,13 @@ impl SetDisks {
 
             let put_object_size = known_put_object_storage_size(data.size());
             let shard_file_size_raw = erasure.shard_file_size(put_object_size);
-            let is_inline_buffer =
-                storage_class_config.should_inline(shard_file_size_raw, erasure.data_shards, opts.versioned);
+            let is_inline_buffer = storage_class_config.should_inline(shard_file_size_raw, erasure.data_shards, opts.versioned);
 
             let collect_stage_timing = rustfs_io_metrics::put_stage_metrics_enabled() || issue3031_diag_enabled();
             let shard_file_size = shard_file_size_raw;
             let shard_size = erasure.shard_size();
             let write_path = classify_put_write_path(is_inline_buffer, put_object_size, fi.erasure.block_size);
             let direct_inline_commit = matches!(write_path, SmallWritePath::Inline);
-            {
-                use std::io::Write;
-                let msg = format!(
-                    "INLINE_DEBUG: bucket={} obj={} size={} shard_fs={} ds={} bs={} inline={} direct={} path={} iblock={} ver={}\n",
-                    bucket, object, put_object_size, shard_file_size_raw, erasure.data_shards, fi.erasure.block_size,
-                    is_inline_buffer, direct_inline_commit, write_path.metric_label(), storage_class_config.inline_block(), opts.versioned
-                );
-                if let Ok(mut f) = std::fs::OpenOptions::new().create(true).append(true).open("/tmp/rustfs_inline_debug.log") {
-                    let _ = f.write_all(msg.as_bytes());
-                }
-                let _ = std::io::stderr().write_all(msg.as_bytes());
-            }
             rustfs_io_metrics::record_put_object_path(write_path.metric_label());
             let writer_setup_stage_start = collect_stage_timing.then(Instant::now);
             let (mut writers, errors) = if direct_inline_commit {

crates/ecstore/src/store/object.rs

@@ -3194,7 +3194,7 @@ impl ECStore {
 
         // Default return value
         let mut del_objects = vec![DeletedObject::default(); objects.len()];
-        let mut accounting = vec![None; objects.len()];
+        let accounting = vec![None; objects.len()];
 
         let mut del_errs = Vec::with_capacity(objects.len());
         for _ in 0..objects.len() {

crates/ecstore/src/store/rebalance/support.rs

@@ -271,7 +271,7 @@ pub(super) fn resolve_latest_object_info_candidates(
             .filter(|candidate| latest_candidate_mod_time(candidate) == Some(latest_mod_time))
             .collect::<Vec<_>>();
 
-        latest_candidates.sort_by(|left, right| right.idx.cmp(&left.idx));
+        latest_candidates.sort_by_key(|candidate| std::cmp::Reverse(candidate.idx));
 
         let Some(winner) = latest_candidates.first() else {
             return Err(Error::ErasureReadQuorum);

crates/replication/src/config.rs

@@ -270,6 +270,160 @@ pub fn active_replication_rule_destination_arns(config: &ReplicationConfiguratio
     arns
 }
 
+/// Deployment id extracted from a site-replication target ARN
+/// (`arn:{rustfs|minio}:replication::<deployment-id>:<bucket>`), or `None`
+/// for an operator-authored ARN.
+pub fn replication_target_arn_deployment_id(arn: &str) -> Option<String> {
+    let parts: Vec<_> = arn.split(':').collect();
+    if parts.len() == 6
+        && parts[0] == "arn"
+        && matches!(parts[1], "rustfs" | "minio")
+        && parts[2] == "replication"
+        && !parts[4].is_empty()
+    {
+        return Some(parts[4].to_string());
+    }
+
+    None
+}
+
+/// Rule id prefix the site-replication reconciler stamps on the rules it
+/// derives (`site-repl-<peer deployment id>`).
+pub const SITE_REPLICATION_RULE_ID_PREFIX: &str = "site-repl-";
+
+/// Whether `rule` carries a site-replication rule id (`site-repl-*`). The
+/// reconciler and the peer ingestion path treat the whole namespace as theirs
+/// on a site-replication bucket; the S3 edit path must not — rule ids are not
+/// reserved, so see [`site_replication_rule_deployment_id`].
+pub fn is_site_replication_rule(rule: &ReplicationRule) -> bool {
+    rule.id
+        .as_deref()
+        .is_some_and(|id| id.starts_with(SITE_REPLICATION_RULE_ID_PREFIX))
+}
+
+/// Deployment id of the peer a reconciler-derived rule replicates to, or
+/// `None` for any other rule. The reconciler builds each rule from one peer:
+/// the id is `site-repl-<deployment id>` and the destination ARN names that
+/// same deployment id — an operator-authored `site-repl-user` rule, or a
+/// `site-repl-<peer>` id pasted onto a foreign ARN, fails the agreement check.
+/// Callers that know the current peer set must also confirm the id is one of
+/// those peers before treating the rule as reconciler-owned.
+pub fn site_replication_rule_deployment_id(rule: &ReplicationRule) -> Option<&str> {
+    let deployment_id = rule.id.as_deref()?.strip_prefix(SITE_REPLICATION_RULE_ID_PREFIX)?;
+    (!deployment_id.is_empty()
+        && replication_target_arn_deployment_id(&rule.destination.bucket).as_deref() == Some(deployment_id))
+    .then_some(deployment_id)
+}
+
+/// Whether `rule` is one the local reconciler derived for a current remote
+/// site-replication peer in `peer_deployment_ids`. With an empty peer set
+/// (site replication disabled) nothing qualifies, so a bucket outside site
+/// replication keeps the verbatim S3 put/delete semantics.
+pub fn is_reconciler_owned_site_replication_rule(rule: &ReplicationRule, peer_deployment_ids: &HashSet<String>) -> bool {
+    site_replication_rule_deployment_id(rule).is_some_and(|deployment_id| peer_deployment_ids.contains(deployment_id))
+}
+
+/// Merge an incoming replication config into the local one.
+///
+/// `site-repl-*` rules encode the *holder's* outbound direction — their
+/// destination ARN names another site — so applying an external rule set
+/// verbatim replaces the local reverse rule with one this site can never
+/// satisfy (no bucket target backs it) and replication silently stops. Only
+/// operator-authored rules travel: the site-replication peer ingestion path
+/// and the S3 put/delete-bucket-replication path both keep the local site's
+/// `site-repl-*` rules through this merge. `incoming == None` models a
+/// delete of the operator-authored rules.
+pub fn merge_incoming_replication_config(
+    incoming: Option<ReplicationConfiguration>,
+    local: Option<ReplicationConfiguration>,
+) -> Option<ReplicationConfiguration> {
+    merge_replication_config_keeping_site_rules(incoming, local, is_site_replication_rule)
+}
+
+/// [`merge_incoming_replication_config`] for the S3 put/delete-bucket-replication
+/// path (issue #1948): only rules the local reconciler derived for a current
+/// peer in `peer_deployment_ids` survive as site rules; every other stored
+/// rule — including an operator-authored `site-repl-*` id — is operator state
+/// that the request replaces or deletes. An incoming rule whose id is a
+/// current peer's `site-repl-<id>` is dropped whatever its ARN: accepting it
+/// would duplicate the reconciler rule's id.
+pub fn merge_user_replication_config(
+    incoming: Option<ReplicationConfiguration>,
+    local: Option<ReplicationConfiguration>,
+    peer_deployment_ids: &HashSet<String>,
+) -> Option<ReplicationConfiguration> {
+    let incoming = incoming.map(|mut config| {
+        config.rules.retain(|rule| {
+            !rule
+                .id
+                .as_deref()
+                .and_then(|id| id.strip_prefix(SITE_REPLICATION_RULE_ID_PREFIX))
+                .is_some_and(|deployment_id| peer_deployment_ids.contains(deployment_id))
+        });
+        config
+    });
+    merge_replication_config_keeping_site_rules(incoming, local, |rule| {
+        is_reconciler_owned_site_replication_rule(rule, peer_deployment_ids)
+    })
+}
+
+fn merge_replication_config_keeping_site_rules(
+    incoming: Option<ReplicationConfiguration>,
+    local: Option<ReplicationConfiguration>,
+    is_site_rule: impl Fn(&ReplicationRule) -> bool,
+) -> Option<ReplicationConfiguration> {
+    let incoming_role = incoming.as_ref().map(|config| config.role.clone()).unwrap_or_default();
+    // Operator rules first, then the local site rules — the same order the
+    // site-replication reconciler produces, so its no-op check matches and
+    // the bucket metadata is written once per broadcast, not twice.
+    let mut rules: Vec<ReplicationRule> = incoming
+        .into_iter()
+        .flat_map(|config| config.rules)
+        .filter(|rule| !is_site_rule(rule))
+        .collect();
+    rules.extend(local.into_iter().flat_map(|config| config.rules).filter(&is_site_rule));
+
+    if rules.is_empty() {
+        return None;
+    }
+
+    assign_site_replication_rule_priorities(&mut rules, &is_site_rule);
+
+    // A site-replication ARN in `role` is the sender's, and the reconciler's
+    // per-peer target lookup reads it — carrying it over would pin the
+    // receiver's targets to the sender's identity.
+    let role = match replication_target_arn_deployment_id(&incoming_role) {
+        Some(_) => String::new(),
+        None => incoming_role,
+    };
+
+    Some(ReplicationConfiguration { role, rules })
+}
+
+/// Give the site rules in `rules` the lowest priorities no operator rule uses,
+/// in rule order, leaving every operator rule's priority untouched. Operator
+/// priorities decide which rule wins per target, so they are part of the
+/// submitted policy; site rules are derived state and only need to be unique
+/// (`validate_replication_config_structure` rejects duplicates). The result
+/// is a pure function of the rule list, so the site-replication reconciler,
+/// the peer ingestion merge and the S3 edit merge all converge on the same
+/// bytes and the reconciler's no-op check holds.
+pub fn assign_site_replication_rule_priorities(rules: &mut [ReplicationRule], is_site_rule: impl Fn(&ReplicationRule) -> bool) {
+    let taken: HashSet<i32> = rules
+        .iter()
+        .filter(|rule| !is_site_rule(rule))
+        .map(|rule| rule.priority.unwrap_or(0))
+        .collect();
+    let mut next = 1;
+    for rule in rules.iter_mut().filter(|rule| is_site_rule(rule)) {
+        while taken.contains(&next) {
+            next += 1;
+        }
+        rule.priority = Some(next);
+        next = next.saturating_add(1);
+    }
+}
+
 pub fn replication_target_arns(config: &ReplicationConfiguration) -> HashSet<String> {
     let role = config.role.trim();
     if !role.is_empty() {
@@ -1544,4 +1698,103 @@ mod tests {
             "the child rule must win for target A while the overlapping child target B remains eligible"
         );
     }
+
+    #[test]
+    fn site_replication_rule_deployment_id_requires_id_and_arn_agreement() {
+        let reconciler_rule = replication_rule("site-repl-peer-dep", "arn:rustfs:replication::peer-dep:bucket");
+        assert_eq!(site_replication_rule_deployment_id(&reconciler_rule), Some("peer-dep"));
+
+        // A remote-target ARN carries the remote's deployment id (or a random
+        // uuid), never the operator's rule id.
+        let operator_named_rule = replication_rule("site-repl-user", "arn:minio:replication:us-east-1:2f1c-remote:bucket");
+        assert_eq!(site_replication_rule_deployment_id(&operator_named_rule), None);
+
+        let foreign_arn = replication_rule("site-repl-peer-dep", "arn:rustfs:replication::other-dep:bucket");
+        assert_eq!(site_replication_rule_deployment_id(&foreign_arn), None);
+
+        let empty_id = replication_rule("site-repl-", "arn:rustfs:replication::peer-dep:bucket");
+        assert_eq!(site_replication_rule_deployment_id(&empty_id), None);
+
+        let peers = HashSet::from(["peer-dep".to_string()]);
+        assert!(is_reconciler_owned_site_replication_rule(&reconciler_rule, &peers));
+        assert!(!is_reconciler_owned_site_replication_rule(&reconciler_rule, &HashSet::new()));
+        let removed_peer = replication_rule("site-repl-gone-dep", "arn:rustfs:replication::gone-dep:bucket");
+        assert!(!is_reconciler_owned_site_replication_rule(&removed_peer, &peers));
+    }
+
+    // The merge must not rewrite the operator's priorities: with the
+    // priority-5 rule listed first and renumbered 1 then 2, the priority-1
+    // delete-marker-disabled rule would win the replication decision.
+    #[test]
+    fn merge_keeps_operator_priorities_and_replication_decision() {
+        let user_arn = "arn:minio:replication:us-east-1:2f1c-remote:bucket";
+        let peer_arn = "arn:rustfs:replication::peer-dep:bucket";
+        let incoming = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                delete_marker_rule("dm-enabled", user_arn, "logs/", 5, true),
+                delete_marker_rule("dm-disabled", user_arn, "logs/2026/", 1, false),
+            ],
+        };
+        let mut site_rule = delete_marker_rule("site-repl-peer-dep", peer_arn, "", 7, true);
+        site_rule.prefix = None;
+        let local = structure_config(vec![site_rule]);
+        let opts = ObjectOpts {
+            name: "logs/2026/app.log".to_string(),
+            op_type: ReplicationType::Delete,
+            delete_marker: true,
+            version_id: None,
+            ..Default::default()
+        };
+        let submitted: Vec<_> = incoming.filter_target_replication_decisions(&opts);
+
+        let peers = HashSet::from(["peer-dep".to_string()]);
+        let merged = merge_user_replication_config(Some(incoming.clone()), Some(local.clone()), &peers).expect("rules");
+
+        let priorities: Vec<_> = merged
+            .rules
+            .iter()
+            .map(|rule| (rule.id.as_deref().unwrap(), rule.priority))
+            .collect();
+        assert_eq!(
+            priorities,
+            vec![
+                ("dm-enabled", Some(5)),
+                ("dm-disabled", Some(1)),
+                ("site-repl-peer-dep", Some(2))
+            ],
+            "operator priorities are kept verbatim; the site rule takes the lowest free slot"
+        );
+        assert!(validate_replication_config_structure(&merged).is_ok());
+        let mut decisions = merged.filter_target_replication_decisions(&opts);
+        decisions.retain(|(arn, _)| arn == user_arn);
+        assert_eq!(decisions, submitted, "the merged config must replicate exactly as the operator submitted");
+        assert_eq!(decisions, vec![(user_arn.to_string(), true)]);
+
+        // The peer ingestion merge follows the same rule.
+        let merged = merge_incoming_replication_config(Some(incoming), Some(local)).expect("rules");
+        let priorities: Vec<_> = merged.rules.iter().map(|rule| rule.priority).collect();
+        assert_eq!(priorities, vec![Some(5), Some(1), Some(2)]);
+    }
+
+    #[test]
+    fn site_rule_priorities_skip_every_operator_priority() {
+        let mut rules = vec![
+            delete_marker_rule("a", "arn:a", "", 2, true),
+            delete_marker_rule("site-repl-x", "arn:rustfs:replication::x:b", "", 9, true),
+            delete_marker_rule("b", "arn:a", "", 1, true),
+            delete_marker_rule("site-repl-y", "arn:rustfs:replication::y:b", "", 9, true),
+            delete_marker_rule("c", "arn:a", "", 4, true),
+        ];
+        assign_site_replication_rule_priorities(&mut rules, is_site_replication_rule);
+        let priorities: Vec<_> = rules.iter().map(|rule| rule.priority).collect();
+        assert_eq!(priorities, vec![Some(2), Some(3), Some(1), Some(5), Some(4)]);
+        assert!(validate_replication_config_structure(&structure_config(rules.clone())).is_ok());
+
+        // Idempotent, so the reconciler's pass over an already-merged config
+        // is a byte-stable no-op rather than a rewrite every period.
+        let settled = rules.clone();
+        assign_site_replication_rule_priorities(&mut rules, is_site_replication_rule);
+        assert_eq!(rules, settled);
+    }
 }

crates/replication/src/lib.rs

@@ -32,9 +32,11 @@ pub use config::{
     ObjectOpts, REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION, REMOTE_TARGET_UNSUPPORTED_FIELDS, REMOTE_TARGET_WRITABLE_FIELDS,
     REPLICATION_CAPABILITY_CONTRACT_VERSION, REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS,
     ReplicationConfigStructureError, ReplicationConfigurationExt, ReplicationTargetValidationError,
-    active_replication_rule_destination_arns, invalid_replication_config_status_field, replication_target_arns,
-    should_remove_replication_target, unsupported_replication_config_field, validate_replication_config_structure,
-    validate_replication_config_target_arns,
+    active_replication_rule_destination_arns, assign_site_replication_rule_priorities, invalid_replication_config_status_field,
+    is_reconciler_owned_site_replication_rule, is_site_replication_rule, merge_incoming_replication_config,
+    merge_user_replication_config, replication_target_arn_deployment_id, replication_target_arns,
+    should_remove_replication_target, site_replication_rule_deployment_id, unsupported_replication_config_field,
+    validate_replication_config_structure, validate_replication_config_target_arns,
 };
 pub use delete::{
     DeletedObjectReplicationInfo, delete_marker_purge_mrf_entry, delete_marker_purge_version_id,

crates/scanner/src/data_usage_define.rs

@@ -28,9 +28,8 @@ use rustfs_common::heal_channel::HealScanMode;
 use rustfs_config::ENV_SCANNER_CACHE_SAVE_TIMEOUT_SECS;
 pub use rustfs_data_usage::{
     AllTierStats, BucketTargetUsageInfo, BucketUsageInfo, DATA_USAGE_OBJECT_NAME, DATA_USAGE_OBSERVED_OBJECT_NAME,
-    DataUsageEntry, DataUsageHash, DataUsageHashMap, DataUsageInfo, DataUsageSnapshotSetState, LEGACY_DATA_USAGE_OBJECT_NAME,
-    PrefixUsageEntry, PrefixUsageQuery, PrefixUsageSummary, ReplTargetSizeSummary, SizeSummary, TierStats, hash_path,
-    prefix_usage_in_cache,
+    DataUsageEntry, DataUsageHash, DataUsageHashMap, DataUsageInfo, LEGACY_DATA_USAGE_OBJECT_NAME, PrefixUsageEntry,
+    PrefixUsageQuery, PrefixUsageSummary, ReplTargetSizeSummary, SizeSummary, TierStats, hash_path, prefix_usage_in_cache,
 };
 use rustfs_utils::path::{SLASH_SEPARATOR, path_join_buf};
 use tokio::time::{Duration, Instant, sleep, timeout};
@@ -345,18 +344,6 @@ pub struct DataUsageCacheInfo {
     pub scan_plan_digest: Option<DataUsageScanPlanDigest>,
     #[serde(default)]
     pub cache_key_format: u16,
-    /// Whether the entries retained while a set scan was incomplete come
-    /// from a prior complete set snapshot.  This is observational input only.
-    #[serde(default)]
-    pub lkg_snapshot_complete: bool,
-    #[serde(default)]
-    pub lkg_next_cycle: Option<u64>,
-    #[serde(default)]
-    pub lkg_last_update: Option<SystemTime>,
-    #[serde(default)]
-    pub lkg_leader_epoch: Option<u64>,
-    #[serde(default)]
-    pub lkg_scan_plan_digest: Option<DataUsageScanPlanDigest>,
 }
 
 impl Serialize for DataUsageCacheInfo {
@@ -366,7 +353,7 @@ impl Serialize for DataUsageCacheInfo {
     {
         // Keep this metadata map-encoded so older readers can ignore fields
         // appended by newer scanner versions during rolling upgrades.
-        let mut state = serializer.serialize_map(Some(21))?;
+        let mut state = serializer.serialize_map(Some(16))?;
         state.serialize_entry("name", &self.name)?;
         state.serialize_entry("next_cycle", &self.next_cycle)?;
         state.serialize_entry("leader_epoch", &self.leader_epoch)?;
@@ -383,11 +370,6 @@ impl Serialize for DataUsageCacheInfo {
         state.serialize_entry("snapshot_complete", &self.snapshot_complete)?;
         state.serialize_entry("scan_plan_digest", &self.scan_plan_digest)?;
         state.serialize_entry("cache_key_format", &self.cache_key_format)?;
-        state.serialize_entry("lkg_snapshot_complete", &self.lkg_snapshot_complete)?;
-        state.serialize_entry("lkg_next_cycle", &self.lkg_next_cycle)?;
-        state.serialize_entry("lkg_last_update", &self.lkg_last_update)?;
-        state.serialize_entry("lkg_leader_epoch", &self.lkg_leader_epoch)?;
-        state.serialize_entry("lkg_scan_plan_digest", &self.lkg_scan_plan_digest)?;
         state.end()
     }
 }

crates/scanner/src/scanner.rs

@@ -2274,8 +2274,9 @@ async fn final_data_usage_publication_defer_reason(
             }
         }
         ScannerCycleStatus::Deferred(reason) => Some(reason),
-        // Incomplete cycles may publish a non-authoritative observational
-        // snapshot when at least one set has a usable current/LKG view.
+        // Incomplete cycles do not publish a usage snapshot. Keep the
+        // decision permissive so existing partial-cycle handling remains
+        // unchanged if a future scanner path emits a bookkeeping update.
         ScannerCycleStatus::Incomplete => None,
     }
 }

crates/scanner/src/scanner/usage_store.rs

@@ -198,7 +198,7 @@ where
             data_usage_info.usage_snapshot_authoritative_baseline = Some(authoritative.snapshot_identity());
         }
 
-        if !data_usage_info.is_complete_bucket_usage_snapshot() && !data_usage_info.usage_snapshot_partial {
+        if !data_usage_info.is_complete_bucket_usage_snapshot() {
             error!(
                 target: "rustfs::scanner",
                 event = EVENT_SCANNER_PERSIST_STATE,

crates/scanner/src/scanner_io.rs

@@ -18,8 +18,8 @@ use crate::scanner_folder::{ScannerItem, scan_data_folder};
 use crate::sleeper::SCANNER_SLEEPER;
 use crate::{
     DATA_USAGE_CACHE_NAME, DATA_USAGE_ROOT, DataUsageCache, DataUsageCacheInfo, DataUsageCachePrepareOutcome,
-    DataUsageCacheSource, DataUsageEntry, DataUsageEntryInfo, DataUsageInfo, DataUsageScanPlanDigest, DataUsageSnapshotSetState,
-    ScannerError, SizeSummary, TierStats,
+    DataUsageCacheSource, DataUsageEntry, DataUsageEntryInfo, DataUsageInfo, DataUsageScanPlanDigest, ScannerError, SizeSummary,
+    TierStats,
 };
 use futures::future::join_all;
 use metrics::counter;
@@ -278,17 +278,6 @@ async fn publish_usage_snapshot(
     Ok(true)
 }
 
-async fn publish_observational_snapshot(
-    updates: &mpsc::Sender<DataUsageInfo>,
-    mut data_usage_info: DataUsageInfo,
-) -> Result<bool> {
-    data_usage_info.usage_snapshot_complete = false;
-    data_usage_info.usage_snapshot_partial = true;
-    data_usage_info.usage_snapshot_converged = Some(false);
-    send_data_usage_update(updates, data_usage_info).await?;
-    Ok(true)
-}
-
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]
 enum ScannerCycleActivityStatus {
     Unchanged,

crates/scanner/src/scanner_io/cache.rs

@@ -188,7 +188,7 @@ pub(super) fn completed_data_usage_info(
     }
 
     let mut total = DataUsageEntry::default();
-    let mut bucket_entries = HashMap::with_capacity(all_buckets.len());
+    let mut buckets_usage = HashMap::with_capacity(all_buckets.len());
     for bucket in all_buckets {
         let mut merged = DataUsageEntry::default();
         for result in results {
@@ -200,14 +200,10 @@ pub(super) fn completed_data_usage_info(
         if !total.checked_merge(&merged) {
             return None;
         }
-        bucket_entries.insert(bucket.clone(), merged);
+        buckets_usage.insert(bucket.clone(), checked_bucket_usage_info(&merged)?);
     }
 
     let merged_last_update = results.iter().filter_map(|result| result.info.last_update).max()?;
-    let buckets_usage = bucket_entries
-        .iter()
-        .map(|(bucket, entry)| Some((bucket.clone(), checked_bucket_usage_info(entry)?)))
-        .collect::<Option<HashMap<_, _>>>()?;
     let bucket_sizes = buckets_usage
         .iter()
         .map(|(bucket, usage)| (bucket.clone(), usage.size))
@@ -229,145 +225,6 @@ pub(super) fn completed_data_usage_info(
     Some((data_usage_info, merged_last_update))
 }
 
-/// Build a non-authoritative view from the set snapshots that completed this
-/// cycle plus compatible per-set last-known-good caches.  The caller must
-/// persist this result only on the observational object; a missing set is
-/// intentionally represented by an incomplete state and is never treated as
-/// an empty set.
-pub(super) fn observational_data_usage_info(
-    results: &[DataUsageCache],
-    expected_sources: &HashSet<DataUsageCacheSource>,
-    all_buckets: &[String],
-    expected_plan_digest: DataUsageScanPlanDigest,
-    scanner_cycle: u64,
-    leader_epoch: u64,
-) -> Option<(DataUsageInfo, SystemTime)> {
-    let mut by_source = HashMap::with_capacity(results.len());
-    for result in results {
-        let source = result.info.source?;
-        if !expected_sources.contains(&source) || by_source.insert(source, result).is_some() {
-            return None;
-        }
-    }
-
-    let mut usable = Vec::new();
-    let mut set_states = Vec::with_capacity(expected_sources.len());
-    let mut sources = expected_sources.iter().copied().collect::<Vec<_>>();
-    sources.sort_by_key(|source| (source.pool_index, source.set_index));
-    for source in sources {
-        let result = by_source.get(&source).copied();
-        let current = result.filter(|result| {
-            result.info.snapshot_complete
-                && result.info.next_cycle == scanner_cycle
-                && result.info.leader_epoch == leader_epoch
-                && result.info.scan_plan_digest == Some(expected_plan_digest)
-        });
-        let lkg = result.filter(|result| {
-            !result.info.snapshot_complete
-                && result.info.lkg_snapshot_complete
-                && result.info.lkg_scan_plan_digest == Some(expected_plan_digest)
-                && result.info.lkg_leader_epoch.is_some_and(|epoch| {
-                    epoch < leader_epoch
-                        || (epoch == leader_epoch && result.info.lkg_next_cycle.is_some_and(|cycle| cycle <= scanner_cycle))
-                })
-        });
-        let current_snapshot = current.is_some();
-        let selected = current.or(lkg);
-        if let Some(selected) = selected {
-            let (cycle, epoch, digest, last_update, complete) = if current_snapshot {
-                (
-                    Some(selected.info.next_cycle),
-                    Some(selected.info.leader_epoch),
-                    selected.info.scan_plan_digest.map(|digest| digest.0),
-                    selected.info.last_update,
-                    true,
-                )
-            } else {
-                (
-                    selected.info.lkg_next_cycle,
-                    selected.info.lkg_leader_epoch,
-                    selected.info.lkg_scan_plan_digest.map(|digest| digest.0),
-                    selected.info.lkg_last_update,
-                    false,
-                )
-            };
-            set_states.push(DataUsageSnapshotSetState {
-                pool_index: u64::try_from(source.pool_index).ok()?,
-                set_index: u64::try_from(source.set_index).ok()?,
-                scanner_cycle: cycle,
-                scanner_epoch: epoch,
-                scan_plan_digest: digest,
-                complete,
-                tombstone: false,
-            });
-            usable.push((selected, last_update));
-        } else {
-            set_states.push(DataUsageSnapshotSetState {
-                pool_index: u64::try_from(source.pool_index).ok()?,
-                set_index: u64::try_from(source.set_index).ok()?,
-                scanner_cycle: None,
-                scanner_epoch: None,
-                scan_plan_digest: Some(expected_plan_digest.0),
-                complete: false,
-                tombstone: false,
-            });
-        }
-    }
-    if usable.is_empty() {
-        return None;
-    }
-
-    let mut total = DataUsageEntry::default();
-    let mut bucket_entries = HashMap::with_capacity(all_buckets.len());
-    let mut merged_last_update = None;
-    for (result, last_update) in usable {
-        if let Some(update) = last_update {
-            merged_last_update = Some(merged_last_update.map_or(update, |current: SystemTime| current.max(update)));
-        }
-        for bucket in all_buckets {
-            let Some(entry) = result.checked_flatten(bucket) else {
-                continue;
-            };
-            let bucket_entry = bucket_entries.entry(bucket.clone()).or_insert_with(DataUsageEntry::default);
-            if !bucket_entry.checked_merge(&entry) {
-                return None;
-            }
-            if !total.checked_merge(&entry) {
-                return None;
-            }
-        }
-    }
-    let merged_last_update = merged_last_update?;
-    let buckets_usage = bucket_entries
-        .iter()
-        .map(|(bucket, entry)| Some((bucket.clone(), checked_bucket_usage_info(entry)?)))
-        .collect::<Option<HashMap<_, _>>>()?;
-    Some((
-        DataUsageInfo {
-            last_update: Some(merged_last_update),
-            scanner_cycle: Some(scanner_cycle),
-            scanner_epoch: Some(leader_epoch),
-            objects_total_count: u64::try_from(total.objects).ok()?,
-            versions_total_count: u64::try_from(total.versions).ok()?,
-            delete_markers_total_count: u64::try_from(total.delete_markers).ok()?,
-            objects_total_size: u64::try_from(total.size).ok()?,
-            tier_stats: total.all_tier_stats.filter(|tiers| !tiers.is_empty()),
-            buckets_count: u64::try_from(buckets_usage.len()).ok()?,
-            bucket_sizes: buckets_usage
-                .iter()
-                .map(|(bucket, usage)| (bucket.clone(), usage.size))
-                .collect(),
-            buckets_usage,
-            usage_snapshot_complete: false,
-            usage_snapshot_partial: true,
-            usage_snapshot_converged: Some(false),
-            usage_snapshot_set_states: set_states,
-            ..Default::default()
-        },
-        merged_last_update,
-    ))
-}
-
 pub(super) async fn send_cache_root_entry_info(
     bucket_result_tx: &mpsc::Sender<DataUsageEntryInfo>,
     cache: &DataUsageCache,

crates/scanner/src/scanner_io/io_cache.rs

@@ -40,21 +40,6 @@ impl ScannerIOCache for SetDisks {
         let set_label = self.set_index.to_string();
 
         let source = DataUsageCacheSource::new(self.pool_index, self.set_index);
-        let mut old_cache = DataUsageCache::default();
-        if let Err(e) = old_cache.load(self.clone(), DATA_USAGE_CACHE_NAME).await {
-            warn!(
-                target: "rustfs::scanner::io",
-                event = EVENT_SCANNER_CACHE_PERSIST_STATE,
-                component = LOG_COMPONENT_SCANNER,
-                subsystem = LOG_SUBSYSTEM_IO,
-                pool = self.pool_index,
-                set = self.set_index,
-                cache_name = DATA_USAGE_CACHE_NAME,
-                state = "old_cache_load_failed",
-                error = %e,
-                "Scanner old data usage cache load failed; rebuilding from bucket caches"
-            );
-        }
         if buckets.is_empty() {
             let now = SystemTime::now();
             let mut cache = DataUsageCache {
@@ -95,24 +80,6 @@ impl ScannerIOCache for SetDisks {
                 "Scanner set state found no online disks"
             );
             reset_disk_bucket_scan_gauges(&pool_label, &set_label);
-            let lkg = old_cache.info.snapshot_complete.then(|| old_cache.clone());
-            let mut incomplete_scope = lkg.clone().unwrap_or_default();
-            incomplete_scope.info.name = DATA_USAGE_ROOT.to_string();
-            incomplete_scope.info.next_cycle = want_cycle;
-            incomplete_scope.info.last_update = None;
-            incomplete_scope.info.leader_epoch = leader_epoch;
-            incomplete_scope.info.source = Some(source);
-            incomplete_scope.info.snapshot_complete = false;
-            incomplete_scope.info.scan_plan_digest = Some(scan_plan_digest);
-            incomplete_scope.info.cache_key_format = DATA_USAGE_CACHE_KEY_FORMAT;
-            if let Some(lkg) = lkg {
-                incomplete_scope.info.lkg_snapshot_complete = true;
-                incomplete_scope.info.lkg_next_cycle = Some(lkg.info.next_cycle);
-                incomplete_scope.info.lkg_last_update = lkg.info.last_update;
-                incomplete_scope.info.lkg_leader_epoch = Some(lkg.info.leader_epoch);
-                incomplete_scope.info.lkg_scan_plan_digest = lkg.info.scan_plan_digest;
-            }
-            let _ = updates.send(incomplete_scope).await;
             return Ok(());
         }
         // Preserve the original set topology across capability filtering. During
@@ -195,24 +162,6 @@ impl ScannerIOCache for SetDisks {
                 "Scanner set state found no usable namespace scanner disks"
             );
             reset_disk_bucket_scan_gauges(&pool_label, &set_label);
-            let lkg = old_cache.info.snapshot_complete.then(|| old_cache.clone());
-            let mut incomplete_scope = lkg.clone().unwrap_or_default();
-            incomplete_scope.info.name = DATA_USAGE_ROOT.to_string();
-            incomplete_scope.info.next_cycle = want_cycle;
-            incomplete_scope.info.last_update = None;
-            incomplete_scope.info.leader_epoch = leader_epoch;
-            incomplete_scope.info.source = Some(source);
-            incomplete_scope.info.snapshot_complete = false;
-            incomplete_scope.info.scan_plan_digest = Some(scan_plan_digest);
-            incomplete_scope.info.cache_key_format = DATA_USAGE_CACHE_KEY_FORMAT;
-            if let Some(lkg) = lkg {
-                incomplete_scope.info.lkg_snapshot_complete = true;
-                incomplete_scope.info.lkg_next_cycle = Some(lkg.info.next_cycle);
-                incomplete_scope.info.lkg_last_update = lkg.info.last_update;
-                incomplete_scope.info.lkg_leader_epoch = Some(lkg.info.leader_epoch);
-                incomplete_scope.info.lkg_scan_plan_digest = lkg.info.scan_plan_digest;
-            }
-            let _ = updates.send(incomplete_scope).await;
             return Ok(());
         }
         let set_disk_inventory = Arc::new(scanner_set_disk_inventory(self.as_ref()).await);
@@ -254,15 +203,22 @@ impl ScannerIOCache for SetDisks {
         record_disk_bucket_scans_active(0, &pool_label, &set_label);
         let _reset_disk_bucket_scan_gauges = DiskBucketScanGaugeReset::new(pool_label.clone(), set_label.clone());
 
-        let old_lkg = old_cache.info.snapshot_complete.then(|| {
-            (
-                old_cache.info.next_cycle,
-                old_cache.info.last_update,
-                old_cache.info.leader_epoch,
-                old_cache.info.scan_plan_digest,
-            )
-        });
-        let prepare_outcome = match old_cache.prepare_for_scan(
+        let mut old_cache = DataUsageCache::default();
+        if let Err(e) = old_cache.load(self.clone(), DATA_USAGE_CACHE_NAME).await {
+            warn!(
+                target: "rustfs::scanner::io",
+                event = EVENT_SCANNER_CACHE_PERSIST_STATE,
+                component = LOG_COMPONENT_SCANNER,
+                subsystem = LOG_SUBSYSTEM_IO,
+                pool = self.pool_index,
+                set = self.set_index,
+                cache_name = DATA_USAGE_CACHE_NAME,
+                state = "old_cache_load_failed",
+                error = %e,
+                "Scanner old data usage cache load failed; rebuilding from bucket caches"
+            );
+        }
+        match old_cache.prepare_for_scan(
             DATA_USAGE_ROOT,
             want_cycle,
             leader_epoch,
@@ -303,16 +259,7 @@ impl ScannerIOCache for SetDisks {
                 );
                 return Ok(());
             }
-            outcome => outcome,
-        };
-        if matches!(prepare_outcome, DataUsageCachePrepareOutcome::Reused)
-            && let Some((cycle, last_update, epoch, digest)) = old_lkg
-        {
-            old_cache.info.lkg_snapshot_complete = true;
-            old_cache.info.lkg_next_cycle = Some(cycle);
-            old_cache.info.lkg_last_update = last_update;
-            old_cache.info.lkg_leader_epoch = Some(epoch);
-            old_cache.info.lkg_scan_plan_digest = digest;
+            DataUsageCachePrepareOutcome::Reused | DataUsageCachePrepareOutcome::Reset => {}
         }
 
         let mut cache = DataUsageCache {
@@ -1152,29 +1099,23 @@ impl ScannerIOCache for SetDisks {
                 cache.info.next_cycle = want_cycle;
                 cache.info.last_update.get_or_insert_with(SystemTime::now);
                 cache.info.snapshot_complete = true;
-                cache.info.lkg_snapshot_complete = false;
-                cache.info.lkg_next_cycle = None;
-                cache.info.lkg_last_update = None;
-                cache.info.lkg_leader_epoch = None;
-                cache.info.lkg_scan_plan_digest = None;
                 cache.clone()
             };
             let _ = persist_and_publish_cache_snapshot(self.clone(), &updates, cache_snapshot, cache_cycle_floor.as_ref()).await;
         } else {
-            let mut incomplete_scope = cache_mutex.lock().await.clone();
-            incomplete_scope.info.name = DATA_USAGE_ROOT.to_string();
-            incomplete_scope.info.next_cycle = want_cycle;
-            incomplete_scope.info.last_update = None;
-            incomplete_scope.info.leader_epoch = leader_epoch;
-            incomplete_scope.info.source = Some(source);
-            incomplete_scope.info.snapshot_complete = false;
-            incomplete_scope.info.scan_plan_digest = Some(scan_plan_digest);
-            incomplete_scope.info.cache_key_format = DATA_USAGE_CACHE_KEY_FORMAT;
-            incomplete_scope.info.lkg_snapshot_complete = old_cache.info.lkg_snapshot_complete;
-            incomplete_scope.info.lkg_next_cycle = old_cache.info.lkg_next_cycle;
-            incomplete_scope.info.lkg_last_update = old_cache.info.lkg_last_update;
-            incomplete_scope.info.lkg_leader_epoch = old_cache.info.lkg_leader_epoch;
-            incomplete_scope.info.lkg_scan_plan_digest = old_cache.info.lkg_scan_plan_digest;
+            let incomplete_scope = DataUsageCache {
+                info: DataUsageCacheInfo {
+                    name: DATA_USAGE_ROOT.to_string(),
+                    next_cycle: want_cycle,
+                    leader_epoch,
+                    source: Some(source),
+                    snapshot_complete: false,
+                    scan_plan_digest: Some(scan_plan_digest),
+                    cache_key_format: DATA_USAGE_CACHE_KEY_FORMAT,
+                    ..Default::default()
+                },
+                cache: HashMap::new(),
+            };
             if let Err(e) = updates.send(incomplete_scope).await {
                 error!(
                     target: "rustfs::scanner::io",

crates/scanner/src/scanner_io/io_cycle.rs

@@ -234,7 +234,6 @@ impl ScannerIOCycle for ECStore {
                 let active_set_scans_clone = active_set_scans.clone();
 
                 let (tx, mut rx) = mpsc::channel::<DataUsageCache>(1);
-                let failed_scope_tx = tx.clone();
 
                 // Spawn task to receive and store results
                 let receiver_fut = tokio::spawn(async move {
@@ -315,21 +314,6 @@ impl ScannerIOCycle for ECStore {
                             state = "set_scan_failed",
                             "Scanner set scan failed; continuing cycle"
                         );
-                        let _ = failed_scope_tx
-                            .send(DataUsageCache {
-                                info: DataUsageCacheInfo {
-                                    name: DATA_USAGE_ROOT.to_string(),
-                                    next_cycle: want_cycle_clone,
-                                    leader_epoch,
-                                    source: Some(source),
-                                    snapshot_complete: false,
-                                    scan_plan_digest: Some(scan_plan_digest),
-                                    cache_key_format: DATA_USAGE_CACHE_KEY_FORMAT,
-                                    ..Default::default()
-                                },
-                                cache: HashMap::new(),
-                            })
-                            .await;
                         let mut first_err = first_err_mutex_clone.lock().await;
                         record_set_scan_failure(&mut first_err, e);
                     }
@@ -386,19 +370,6 @@ impl ScannerIOCycle for ECStore {
             budget_elapsed,
             ctx.is_cancelled(),
         );
-        let observational_usage = completed_usage
-            .is_none()
-            .then(|| {
-                observational_data_usage_info(
-                    &results,
-                    &expected_sources,
-                    &all_bucket_names,
-                    scan_plan_digest,
-                    want_cycle,
-                    leader_epoch,
-                )
-            })
-            .flatten();
         let structurally_complete_snapshot = result.is_ok() && completed_all_sets && completed_usage.is_some();
         let cycle_status = classify_nsscanner_cycle(
             structurally_complete_snapshot,
@@ -410,10 +381,6 @@ impl ScannerIOCycle for ECStore {
         );
         if let Some((data_usage_info, _)) = completed_usage {
             publish_usage_snapshot(&updates, cycle_status, data_usage_info).await?;
-        } else if !ctx.is_cancelled()
-            && let Some((data_usage_info, _)) = observational_usage
-        {
-            publish_observational_snapshot(&updates, data_usage_info).await?;
         }
         let dirty_usage_clear = should_clear_dirty_usage_snapshot(
             result.is_ok(),

crates/scanner/src/scanner_io/publish_gate_tests.rs

@@ -105,160 +105,6 @@ fn completed_data_usage_info_for_test(
     completed_data_usage_info(results, &expected_sources, all_buckets, true, budget_elapsed, cancelled)
 }
 
-fn lkg_root_cache(bucket: &str, objects: usize, source: DataUsageCacheSource) -> DataUsageCache {
-    let mut cache = completed_root_cache(bucket, objects, 10, source);
-    cache.info.snapshot_complete = false;
-    cache.info.next_cycle = 8;
-    cache.info.leader_epoch = 3;
-    cache.info.lkg_snapshot_complete = true;
-    cache.info.lkg_next_cycle = Some(7);
-    cache.info.lkg_last_update = cache.info.last_update;
-    cache.info.lkg_leader_epoch = Some(3);
-    cache.info.lkg_scan_plan_digest = Some(TEST_PLAN_DIGEST);
-    cache
-}
-
-#[test]
-fn partial_usage_is_observational_not_authoritative_for_quota() {
-    let all_buckets = vec!["bucket".to_string()];
-    let current_source = DataUsageCacheSource::new(0, 0);
-    let stalled_source = DataUsageCacheSource::new(1, 0);
-    let mut current = completed_root_cache("bucket", 2, 20, current_source);
-    current.info.next_cycle = 8;
-    current.info.leader_epoch = 3;
-    let stalled = lkg_root_cache("bucket", 1, stalled_source);
-    let expected = HashSet::from([current_source, stalled_source]);
-
-    assert!(
-        completed_data_usage_info(&[current.clone(), stalled.clone()], &expected, &all_buckets, true, false, false).is_none()
-    );
-    let (observed, _) = observational_data_usage_info(&[current, stalled], &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3)
-        .expect("a completed set should produce an observational view");
-    assert!(observed.usage_snapshot_partial);
-    assert!(!observed.usage_snapshot_complete);
-    assert_eq!(observed.usage_snapshot_converged, Some(false));
-    assert_eq!(observed.usage_snapshot_set_states.len(), 2);
-}
-
-#[test]
-fn lkg_scope_does_not_count_as_current_cycle_completion() {
-    let source = DataUsageCacheSource::new(0, 0);
-    let mut lkg = lkg_root_cache("bucket", 1, source);
-    lkg.info.last_update = None;
-    let expected = HashSet::from([source]);
-    assert!(!scanner_results_form_complete_snapshot(&[lkg], &expected));
-}
-
-#[test]
-fn stale_quota_uses_complete_baseline_plus_positive_deltas() {
-    let all_buckets = vec!["bucket".to_string()];
-    let source = DataUsageCacheSource::new(0, 0);
-    let mut current = completed_root_cache("bucket", 3, 20, source);
-    current.info.next_cycle = 8;
-    current.info.leader_epoch = 3;
-    let expected = HashSet::from([source]);
-    let (observed, _) = observational_data_usage_info(&[current], &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3)
-        .expect("complete set data is a valid observational baseline");
-    assert_eq!(observed.objects_total_size, 30);
-    assert_eq!(observed.usage_snapshot_set_states[0].complete, true);
-}
-
-#[test]
-fn negative_delta_waits_for_set_reconciliation() {
-    let all_buckets = vec!["bucket".to_string()];
-    let source = DataUsageCacheSource::new(0, 0);
-    let mut stalled = lkg_root_cache("bucket", 4, source);
-    stalled.info.lkg_scan_plan_digest = Some(DataUsageScanPlanDigest([9; 32]));
-    let expected = HashSet::from([source]);
-    assert!(observational_data_usage_info(&[stalled], &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3).is_none());
-}
-
-#[test]
-fn set_membership_add_remove_uses_generation_and_tombstone() {
-    let state = DataUsageSnapshotSetState {
-        pool_index: 1,
-        set_index: 2,
-        scanner_cycle: Some(9),
-        scanner_epoch: Some(4),
-        scan_plan_digest: Some(TEST_PLAN_DIGEST.0),
-        complete: false,
-        tombstone: true,
-    };
-    let encoded = serde_json::to_vec(&state).expect("set state should serialize");
-    let decoded: DataUsageSnapshotSetState = serde_json::from_slice(&encoded).expect("set state should deserialize");
-    assert_eq!(decoded, state);
-
-    let snapshot = DataUsageInfo {
-        last_update: Some(SystemTime::UNIX_EPOCH + Duration::from_secs(10)),
-        scanner_cycle: Some(9),
-        scanner_epoch: Some(4),
-        buckets_count: 0,
-        usage_snapshot_converged: Some(false),
-        usage_snapshot_partial: true,
-        usage_snapshot_set_states: vec![
-            DataUsageSnapshotSetState {
-                pool_index: 0,
-                set_index: 0,
-                scanner_cycle: Some(9),
-                scanner_epoch: Some(4),
-                scan_plan_digest: Some(TEST_PLAN_DIGEST.0),
-                complete: true,
-                tombstone: false,
-            },
-            state,
-        ],
-        ..Default::default()
-    };
-    assert!(snapshot.is_valid_partial_snapshot());
-}
-
-#[test]
-fn old_set_completion_cannot_overwrite_new_aggregate() {
-    let all_buckets = vec!["bucket".to_string()];
-    let source = DataUsageCacheSource::new(0, 0);
-    let mut old = completed_root_cache("bucket", 1, 20, source);
-    old.info.next_cycle = 7;
-    old.info.leader_epoch = 2;
-    let expected = HashSet::from([source]);
-    assert!(observational_data_usage_info(&[old], &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3).is_none());
-}
-
-#[test]
-fn usage_aggregate_survives_restart_and_leader_failover() {
-    let all_buckets = vec!["bucket".to_string()];
-    let source = DataUsageCacheSource::new(0, 0);
-    let mut lkg = lkg_root_cache("bucket", 5, source);
-    lkg.info.lkg_leader_epoch = Some(4);
-    lkg.info.lkg_next_cycle = Some(9);
-    let expected = HashSet::from([source]);
-    let (observed, _) = observational_data_usage_info(&[lkg], &expected, &all_buckets, TEST_PLAN_DIGEST, 10, 5)
-        .expect("compatible LKG should survive a leader change");
-    assert_eq!(observed.usage_snapshot_set_states[0].scanner_epoch, Some(4));
-    assert_eq!(observed.objects_total_size, 50);
-}
-
-#[test]
-fn usage_aggregate_cost_is_linear_in_set_count() {
-    let all_buckets = vec!["bucket".to_string()];
-    let mut results = Vec::new();
-    let mut expected = HashSet::new();
-    for index in 0..32 {
-        let source = DataUsageCacheSource::new(index, 0);
-        expected.insert(source);
-        let mut cache = completed_root_cache("bucket", 1, 20, source);
-        cache.info.next_cycle = 8;
-        cache.info.leader_epoch = 3;
-        results.push(cache);
-    }
-    let (observed, _) = observational_data_usage_info(&results, &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3)
-        .expect("all set snapshots should aggregate");
-    assert_eq!(observed.objects_total_count, 32);
-    let reversed = results.iter().rev().cloned().collect::<Vec<_>>();
-    let (reversed_observed, _) = observational_data_usage_info(&reversed, &expected, &all_buckets, TEST_PLAN_DIGEST, 8, 3)
-        .expect("reordered set snapshots should aggregate");
-    assert_eq!(observed.usage_snapshot_set_states, reversed_observed.usage_snapshot_set_states);
-}
-
 #[test]
 fn completed_data_usage_info_publishes_tier_stats_across_sets() {
     let all_buckets = vec!["bucket-a".to_string(), "bucket-b".to_string()];

deny.toml

@@ -43,9 +43,6 @@ allow-git = [
     # RustFS fork carrying presigned expiry and constant-time authentication fixes.
     # owner: rustfs-maintainers  review: 2026-10
     "https://github.com/rustfs/s3s.git",
-    # MiMalloc fork pinned for hotpath allocation counting support.
-    # owner: houseme  review: 2026-10
-    "https://github.com/xonatius/mimalloc_rust.git",
 ]
 
 [bans]

docs/testing/ci-gates.md

@@ -0,0 +1,149 @@
+# CI gate matrix
+
+This file is the source of truth for which validation runs on each event, its
+configured wall-clock budget, and whether it can block a merge. Test taxonomy,
+naming, and nextest serialization rules remain in [README.md](README.md); e2e
+membership and counts remain in
+[e2e-suite-inventory.md](e2e-suite-inventory.md).
+
+The distinction between **required** and **report-only** is load-bearing:
+a failing job blocks a merge only when its exact check name is present in the
+live `main` ruleset. A workflow name, a `merge_group` trigger, or a red PR check
+does not make a job required by itself.
+
+## Required merge checks
+
+The live `main` ruleset (`6436880`) currently requires exactly these contexts:
+
+| Required context | Producer | Validation |
+|---|---|---|
+| `CLA Check` | `.github/workflows/cla.yml` | Contributor agreement |
+| `Quick Checks` | `.github/workflows/ci.yml` | Formatting and repository guard scripts |
+| `Test and Lint` | `.github/workflows/ci.yml` | Clippy, workspace nextest excluding `e2e_test`, doctests, and migration proofs |
+
+For pull requests limited to the paths excluded by the main CI workflow,
+`.github/workflows/ci-docs-only.yml` reports `Quick Checks` and
+`Test and Lint` under the same names. It runs the real quick checks and the
+planning-document guard; it does not claim that Rust compilation or runtime
+tests ran. Despite the workflow name, these paths also include selected deploy,
+workflow, and lock files.
+
+Verify the live rule rather than trusting this snapshot before changing merge
+policy:
+
+```bash
+gh api repos/rustfs/rustfs/rulesets/6436880 \
+  --jq '.rules[] | select(.type == "required_status_checks") | .parameters'
+```
+
+The ruleset currently has `strict_required_status_checks_policy=false`.
+`Continuous Integration` accepts `merge_group` events and runs `e2e-full` for
+them, but `End-to-End Tests (full merge gate)` is not currently a required
+context. Therefore the repository is prepared to test a merge-queue SHA, but
+the workflow alone does not prove that every merge passed that lane.
+
+## Pull request and merge matrix
+
+Budgets below are job `timeout-minutes`, not typical runtimes. “Report-only”
+means the result is visible and actionable but is not in the live required
+context list.
+
+| Event | Validation | Budget | Merge status | Reproduction |
+|---|---|---:|---|---|
+| PR, non-doc change | `Quick Checks` | 10 min | Required | `make pre-commit` (broader local umbrella) |
+| PR, non-doc change | `Test and Lint` | 90 min | Required | `cargo nextest run --profile ci --all --exclude e2e_test` |
+| PR, non-doc change | `Typos` | 10 min | Report-only | `typos` |
+| PR, non-doc change | `ILM Integration (serial)` | 90 min | Report-only | Use the exact command in `.github/workflows/ci.yml` |
+| PR, non-doc change | rio-v2 / swift / sftp test-and-lint variants | 90 min each | Report-only | `cargo nextest run` with the workflow's feature set |
+| PR, non-doc change | `Build RustFS Debug Binary` | 30 min | Report-only; prerequisite for black-box lanes | `cargo build -p rustfs --bins` |
+| PR, non-doc change | `io_uring Integration (real)` | 30 min | Report-only | `cargo test -p rustfs-ecstore --lib uring_ -- --test-threads=1 --nocapture` |
+| PR, non-doc change | `End-to-End Tests` (`e2e-smoke` plus `s3s-e2e`) | 30 min | Report-only | `cargo nextest run --profile e2e-smoke -p e2e_test`; then `./scripts/e2e-run.sh ./target/debug/rustfs <data-dir>` |
+| PR, non-doc change | `S3 Implemented Tests` | 60 min | Report-only | Build `rustfs`, then run `scripts/s3-tests/run.sh` with `DEPLOY_MODE=binary`, `TEST_MODE=single`, and `MAXFAIL=0` |
+| PR, non-doc change | `S3 Lifecycle Behavior Tests` | 30 min | Report-only | Use the accelerated scanner environment in `.github/workflows/ci.yml` with `scripts/s3-tests/run.sh` |
+| PR touching dependency or workflow inputs | Cargo Deny / Workflow Pin Report / Dependency Review | 20 / 5 / 30 min | Report-only | `cargo deny check`; `scripts/security/check_workflow_pins.sh` |
+| PR touching architecture rules or architecture docs | `Architecture Migration Rules` | 10 min | Report-only | `scripts/check_architecture_migration_rules.sh` |
+| PR touching Nix or workspace manifests | `Nix Build & Check` | 60 min | Report-only | `nix flake check` |
+| PR limited to main-CI-excluded paths | companion `Quick Checks` and `Test and Lint` | 10 min each | Required | `git diff --check`; `make doc-paths-check` when documentation paths changed |
+| `merge_group` | Standard CI plus `e2e-full` | 55 min for `e2e-full` | Standard required contexts only; `e2e-full` report-only | `cargo nextest run --profile e2e-full -p e2e_test` |
+| Push to `main` | Standard CI plus `e2e-full` | 55 min for `e2e-full` | Post-merge detection | Same as `merge_group` |
+| PR touching fuzz inputs or harness paths | Build plus five 60-second fuzz smoke targets | 60 min build; 30 min per target | Report-only | `MAX_TOTAL_TIME=60 ./scripts/fuzz/run.sh` |
+| PR touching selected ecstore disk/format paths | `Rename Safety` on Windows | 60 min | Report-only | Run the four `cargo test -p rustfs-ecstore --lib <filter>` commands in `windows-filesystem.yml` on Windows |
+
+The authoritative e2e filters live in `.config/nextest.toml`; extend a profile
+instead of adding a second ad-hoc selector. Before a profile runs,
+`scripts/check_test_wiring.py` compares its exact membership to the committed
+digest so a silent test drop fails closed.
+
+## Scheduled and manual validation
+
+Scheduled lanes are independent fault domains. They do not block a pull
+request, but their workflow-local gate can fail the run and scheduled failures
+are routed to the shared failure-issue action. The scheduled-validation
+watchdog and freshness workflow separately detect incomplete runs and missing
+schedules.
+
+| Cadence (UTC unless noted) | Workflow / validation | Budget | Verdict and artifacts | Reproduction |
+|---|---|---:|---|---|
+| Daily 02:17 | Fuzz: five nightly corpus targets | 60 min build; 60 min per target | Gate; corpus/crash artifacts, scheduled failure alert | `MAX_TOTAL_TIME=<seconds> ./scripts/fuzz/run.sh` |
+| Daily 03:17 | MinIO interop (EC + SSE read parity) | 40 min | Gate; scheduled failure alert | Dispatch `minio-interop.yml` or follow its pinned Docker fixture steps |
+| Daily 04:29 | Replication / cluster-fault / protocol e2e | 45 / 90 / 90 min | Three independent gates; JUnit, membership, and server logs | `cargo nextest run --profile e2e-repl-nightly -p e2e_test`; `--profile e2e-nightly`; `-j 1 --profile e2e-protocols` |
+| Daily 06:31 | Warp performance A/B | 180 min | Regression budget gate; A/B summaries and server logs | `bash scripts/run_hotpath_warp_abba.sh --help` |
+| Daily 00:07 Asia/Shanghai (16:07 UTC previous day) | Nightly GNU build and Vault lanes | 150 / 90 / 60 min | Build, live Vault, and HA failover gates | Use the commands and pinned Vault images in `nightly-gnu.yml` |
+| Daily 03:23 | Security Audit | 20 / 5 min, plus 30 min on PR dependency review | Cargo Deny and workflow-pin gates; scheduled failure alert | `cargo deny check`; `scripts/security/check_workflow_pins.sh` |
+| Daily 23:47 | Scheduled Validation Freshness | 10 min | Fails when a critical schedule was never created or is stale | Dispatch `scheduled-validation-freshness.yml` |
+| Sunday 00:11 | Full `Continuous Integration` matrix | Per-job budgets above | Weekly variant coverage, including dormant rio-v2 binary/e2e lanes | Dispatch `ci.yml` |
+| Sunday 01:13 | Seven-platform build matrix | 150 min per platform | Build/package integrity; scheduled failure alert | Dispatch `build.yml` with an exact platform set |
+| Sunday 02:19 | Ceph s3-tests full sweep: single and real four-node, four shards each | 180 min per shard | Compatibility gate; report, JUnit, exact node IDs, and server logs | `scripts/s3-tests/run.sh` against an existing single or distributed target |
+| Sunday 06:41 | Mint | 120 min | **Report-only by design**; per-suite PASS/FAIL/NA and raw `log.json` | Reproduce the pinned Docker sequence in `mint.yml` or dispatch it |
+| Sunday 07:43 | Workspace line coverage | 120 min | Report-only trend; lcov and JSON retained 90 days | `make coverage` |
+| Monthly, day 1 06:37 | Runner Hygiene | 15 min | Validates runner ephemerality; scheduled failure alert | Dispatch `runner-hygiene.yml` |
+
+Manual `workflow_dispatch` exists for the scheduled workflows above. Manual
+runs are debugging evidence and intentionally do not open scheduled-failure
+issues. A manual performance run may explicitly allow a known regression; that
+override must not be treated as an ordinary passing baseline.
+
+## Release validation
+
+Release validation is post-merge and tag-driven; it does not substitute for a
+pull-request gate.
+
+| Event | Validation | Budget | Result |
+|---|---|---:|---|
+| Push to `main` or weekly schedule | `Build and Release` platform matrix | 150 min per platform | Build artifacts for all selected targets; no release publication on a main push |
+| Valid release or preview tag | `Build and Release` plus asset checks | 150 min per platform | Draft release, checksummed assets, and publish step |
+| Successful non-preview release-tag build | Docker image build and image scan | 60 min build; 30 min scan | Multi-architecture images plus vulnerability report |
+| Successful release-tag build | DEB/RPM packaging | 30 min per architecture | Packages and checksum files uploaded to the release |
+| Successful non-preview release-tag build | Helm template test and package | 30 min build; 30 min publish | Versioned chart and repository index |
+
+Use an exact preview tag for end-to-end release rehearsal. Manual dispatches
+are backfill/debug paths and do not prove the automatic `workflow_run` chain.
+
+## Evidence requirements
+
+A green check is useful only when it proves the intended behavior ran:
+
+- Record the exact commit SHA and run URL.
+- Separate product failure from runner prerequisites, service readiness, and
+  cancellation. Repair the precondition, then rerun the exact workload.
+- Preserve membership manifests, JUnit, raw compatibility logs, seeds, and
+  server logs where the workflow provides them.
+- For a bug fix or a new fault checker, provide sensitivity evidence: the old
+  behavior or an intentional mutation must fail the new oracle, and the fixed
+  behavior must pass it.
+- Never promote a report-only lane to required from one green run. Require at
+  least 14 days and 30 representative pull requests with at least 99% complete
+  execution, then update the ruleset and this table together.
+
+## Change checklist
+
+Update this file in the same pull request when any of these change:
+
+- workflow triggers, job names, timeouts, or nextest profile ownership;
+- required status contexts or strict/merge-queue policy;
+- scheduled cadence, alert routing, artifact contract, or local reproduction;
+- report-only versus gating semantics.
+
+Do not copy per-module test counts here. Update
+[e2e-suite-inventory.md](e2e-suite-inventory.md) and its enforced membership
+digest instead.

rustfs/Cargo.toml

@@ -336,13 +336,13 @@ opentelemetry = { workspace = true }
 tracing-opentelemetry = { workspace = true }
 # Data structures
 hashbrown = { workspace = true, features = ["serde", "rayon"] }
-mimalloc = { workspace = true }
+rustfs-mimalloc = { workspace = true }
 
 [target.'cfg(target_os = "linux")'.dependencies]
 libsystemd.workspace = true
 
 [target.'cfg(not(target_os = "windows"))'.dependencies]
-libmimalloc-sys.workspace = true
+rustfs-mimalloc-sys.workspace = true
 
 [dev-dependencies]
 uuid = { workspace = true, features = ["v4", "v5", "fast-rng", "macro-diagnostics"] }

rustfs/src/admin/handlers/site_replication.rs

@@ -31,6 +31,10 @@ use crate::admin::storage_api::bucket::metadata::{
 use crate::admin::storage_api::bucket::metadata_sys;
 use crate::admin::storage_api::bucket::quota::BucketQuota;
 use crate::admin::storage_api::bucket::replication;
+use crate::admin::storage_api::bucket::replication::{
+    assign_site_replication_rule_priorities, is_site_replication_rule, merge_incoming_replication_config,
+    replication_target_arn_deployment_id,
+};
 use crate::admin::storage_api::bucket::target::{ARN, BucketTarget, BucketTargetType, BucketTargets, Credentials};
 use crate::admin::storage_api::bucket::target_sys::BucketTargetSys;
 use crate::admin::storage_api::bucket::utils::{deserialize, serialize};
@@ -1118,6 +1122,36 @@ async fn load_site_replication_state() -> S3Result<SiteReplicationState> {
     }
 }
 
+/// Whether this deployment participates in site replication (two or more
+/// peers in the persisted state). Read by the S3 interface layer to gate
+/// replication-config edits (MinIO `ErrReplicationDenyEditError` semantics,
+/// issue #1948); a state-read failure propagates so the gate fails closed.
+pub(crate) async fn site_replication_enabled() -> S3Result<bool> {
+    Ok(load_site_replication_state().await?.enabled())
+}
+
+/// Deployment ids of the remote peers the reconciler derives a
+/// `site-repl-<id>` rule for on every bucket (the same peer filter as
+/// `build_site_replication_config`); empty when site replication is not
+/// enabled. Read by the bucket usecase so an S3 replication-config edit keeps
+/// exactly the reconciler-owned rules (issue #1948); a state-read failure
+/// propagates so the edit fails closed.
+pub(crate) async fn site_replication_remote_peer_deployment_ids() -> S3Result<HashSet<String>> {
+    let state = load_site_replication_state().await?;
+    if !state.enabled() {
+        return Ok(HashSet::new());
+    }
+    let local_peer = current_local_runtime_peer(&state);
+    Ok(state
+        .peers
+        .values()
+        .filter(|peer| {
+            peer.deployment_id != local_peer.deployment_id && !same_identity_endpoint(&peer.endpoint, &local_peer.endpoint)
+        })
+        .map(|peer| peer.deployment_id.clone())
+        .collect())
+}
+
 async fn load_site_replication_state_no_lock(store: Arc<ECStore>) -> S3Result<SiteReplicationState> {
     match read_config_no_lock(store, SITE_REPLICATION_STATE_PATH).await {
         Ok(data) => parse_site_replication_state(&data),
@@ -7762,20 +7796,6 @@ fn bucket_target_deployment_id(target: &BucketTarget) -> Option<String> {
     replication_target_arn_deployment_id(&target.arn)
 }
 
-fn replication_target_arn_deployment_id(arn: &str) -> Option<String> {
-    let parts: Vec<_> = arn.split(':').collect();
-    if parts.len() == 6
-        && parts[0] == "arn"
-        && matches!(parts[1], "rustfs" | "minio")
-        && parts[2] == "replication"
-        && !parts[4].is_empty()
-    {
-        return Some(parts[4].to_string());
-    }
-
-    None
-}
-
 fn prune_removed_site_replication_bucket_targets(
     existing: BucketTargets,
     removed_deployment_ids: &HashSet<String>,
@@ -7800,10 +7820,6 @@ fn prune_removed_site_replication_bucket_targets(
     (BucketTargets { targets }, removed)
 }
 
-fn is_site_replication_rule(rule: &ReplicationRule) -> bool {
-    rule.id.as_deref().is_some_and(|id| id.starts_with("site-repl-"))
-}
-
 /// Whether every `site-repl-*` rule on this bucket resolves to a live remote target.
 ///
 /// The rule set alone cannot answer this: a rule can be perfectly formed while the endpoint
@@ -7829,52 +7845,6 @@ async fn site_replication_targets_online(bucket: &str, replication_config_xml: &
     true
 }
 
-/// Merge a peer's replication config into the local one.
-///
-/// `site-repl-*` rules encode the *sender's* outbound direction — their destination ARN
-/// names the receiver — so applying a peer's rule set verbatim replaces the receiver's
-/// reverse rule with one pointing at itself. No bucket target can satisfy that ARN
-/// (`reconcile_site_replication_bucket_targets` skips the local peer), so the receiver
-/// silently stops replicating back: the one-directional symptom. Only operator-authored
-/// rules travel between sites; each site owns its own `site-repl-*` rules.
-fn merge_incoming_replication_config(
-    incoming: Option<ReplicationConfiguration>,
-    local: Option<ReplicationConfiguration>,
-) -> Option<ReplicationConfiguration> {
-    let incoming_role = incoming.as_ref().map(|config| config.role.clone()).unwrap_or_default();
-    // Operator rules first, then the local site rules — the same order
-    // `ensure_site_replication_bucket_replication_config_with_runtime` produces, so its
-    // no-op check matches and the bucket metadata is written once per broadcast, not twice.
-    let mut rules: Vec<ReplicationRule> = incoming
-        .into_iter()
-        .flat_map(|config| config.rules)
-        .filter(|rule| !is_site_replication_rule(rule))
-        .collect();
-    rules.extend(
-        local
-            .into_iter()
-            .flat_map(|config| config.rules)
-            .filter(is_site_replication_rule),
-    );
-
-    if rules.is_empty() {
-        return None;
-    }
-
-    for (index, rule) in rules.iter_mut().enumerate() {
-        rule.priority = Some(i32::try_from(index + 1).unwrap_or(i32::MAX));
-    }
-
-    // A site-replication ARN in `role` is the sender's, and `site_replication_target_arns_by_peer`
-    // reads it — carrying it over would pin the receiver's targets to the sender's identity.
-    let role = match replication_target_arn_deployment_id(&incoming_role) {
-        Some(_) => String::new(),
-        None => incoming_role,
-    };
-
-    Some(ReplicationConfiguration { role, rules })
-}
-
 /// Merge a peer's ILM expiry document into the local lifecycle config.
 ///
 /// Mirrors MinIO's `mergeWithCurrentLCConfig` with one hardening: incoming
@@ -8213,9 +8183,7 @@ fn prune_removed_site_replication_rules(
         return (None, removed);
     }
 
-    for (index, rule) in config.rules.iter_mut().enumerate() {
-        rule.priority = Some(i32::try_from(index + 1).unwrap_or(i32::MAX));
-    }
+    assign_site_replication_rule_priorities(&mut config.rules, is_site_replication_rule);
 
     (Some(config), removed)
 }
@@ -8389,9 +8357,10 @@ async fn ensure_site_replication_bucket_replication_config_with_runtime(
         .cloned()
         .collect();
     rules.extend(desired.rules);
-    for (index, rule) in rules.iter_mut().enumerate() {
-        rule.priority = Some(i32::try_from(index + 1).unwrap_or(i32::MAX));
-    }
+    // Operator priorities are the operator's policy; only the derived rules
+    // take free slots, by the same function as the config merges so a merged
+    // write and this pass agree byte for byte.
+    assign_site_replication_rule_priorities(&mut rules, is_site_replication_rule);
 
     // Only a site-replication ARN in `role` is ours to drop — an operator-authored role is
     // part of the bucket's S3-visible configuration, and repairing a reverse rule must not
@@ -17089,7 +17058,7 @@ mod tests {
     }
 
     #[test]
-    fn test_prune_removed_site_replication_rules_removes_site_rule_and_reorders_priorities() {
+    fn test_prune_removed_site_replication_rules_removes_site_rule_and_keeps_operator_priority() {
         let removed_deployment_ids = HashSet::from(["removed-dep".to_string()]);
         let kept_rule = build_site_replication_rule("arn:rustfs:replication::kept-dep:photos", 3, "site-repl-kept-dep");
         let removed_rule = build_site_replication_rule("arn:rustfs:replication::removed-dep:photos", 1, "site-repl-removed-dep");
@@ -17106,9 +17075,9 @@ mod tests {
         assert!(updated.role.is_empty());
         assert_eq!(updated.rules.len(), 2);
         assert_eq!(updated.rules[0].id.as_deref(), Some("user-managed-rule"));
-        assert_eq!(updated.rules[0].priority, Some(1));
+        assert_eq!(updated.rules[0].priority, Some(9), "the operator's priority is policy and stays");
         assert_eq!(updated.rules[1].id.as_deref(), Some("site-repl-kept-dep"));
-        assert_eq!(updated.rules[1].priority, Some(2));
+        assert_eq!(updated.rules[1].priority, Some(1), "the derived rule moves to the lowest free slot");
     }
 
     #[test]

rustfs/src/admin/storage_api.rs

@@ -443,6 +443,8 @@ pub(crate) mod replication {
     pub(crate) use super::ecstore_bucket::replication::{
         REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION, REMOTE_TARGET_UNSUPPORTED_FIELDS, REMOTE_TARGET_WRITABLE_FIELDS,
         REPLICATION_CAPABILITY_CONTRACT_VERSION, REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS,
+        assign_site_replication_rule_priorities, is_site_replication_rule, merge_incoming_replication_config,
+        replication_target_arn_deployment_id,
     };
     pub(crate) type BucketReplicationResyncStatus = super::ecstore_bucket::replication::BucketReplicationResyncStatus;
     pub(crate) type BucketStats = super::ecstore_bucket::replication::BucketStats;

rustfs/src/allocator_reclaim.rs

@@ -369,14 +369,8 @@ pub fn allocator_reclaim_controller_snapshot(ctx: &CancellationToken) -> Allocat
 }
 
 #[cfg(not(target_os = "windows"))]
-#[allow(unsafe_code)]
 fn collect_allocator_memory(force: bool) -> Result<(), String> {
-    // SAFETY: `mi_collect` is provided by the active global allocator backend
-    // on this target family. It is explicitly intended to reclaim retained
-    // pages/segments and does not require additional invariants from the caller.
-    unsafe {
-        libmimalloc_sys::mi_collect(force);
-    }
+    rustfs_mimalloc::MiMalloc::collect(force);
     Ok(())
 }
 

rustfs/src/app/bucket_usecase.rs

@@ -38,9 +38,9 @@ use super::storage_api::bucket_usecase::bucket::{
     metadata_sys,
     policy_sys::PolicySys,
     replication::{
-        ReplicationTargetValidationError, invalid_replication_config_status_field, replication_target_arns,
-        should_remove_replication_target, unsupported_replication_config_field, validate_replication_config_structure,
-        validate_replication_config_target_arns,
+        ReplicationTargetValidationError, invalid_replication_config_status_field, merge_user_replication_config,
+        replication_target_arns, should_remove_replication_target, unsupported_replication_config_field,
+        validate_replication_config_structure, validate_replication_config_target_arns,
     },
     target::{BucketTargetType, BucketTargets},
     utils::serialize,
@@ -623,11 +623,52 @@ async fn validate_bucket_replication_update(bucket: &str, config: &ReplicationCo
     validate_replication_config_targets(&targets, config)
 }
 
-async fn replication_targets_without_config_targets(
+/// Defense in depth for site-replication-managed buckets (issue #1948): an S3
+/// PutBucketReplication replaces the operator-authored rules but must not wipe
+/// the rules the reconciler derived for the current remote peers
+/// (`site_peer_deployment_ids`) — until its next pass (600s period) every
+/// peer link on this bucket would be silently dead. The same merge also drops
+/// incoming impostors of those rules. An empty peer set (site replication
+/// disabled) keeps the verbatim overwrite semantics: rule ids are not
+/// reserved, so an operator's own `site-repl-*` rule is ordinary state there.
+fn merge_user_replication_config_update(
+    incoming: ReplicationConfiguration,
+    existing: Option<ReplicationConfiguration>,
+    site_peer_deployment_ids: &HashSet<String>,
+) -> ReplicationConfiguration {
+    if site_peer_deployment_ids.is_empty() {
+        return incoming;
+    }
+    // `incoming` passed structure validation, so it holds at least one rule;
+    // `None` is only reachable when every incoming rule impersonates a
+    // reconciler rule, and then the stored reconciler rules are what remains.
+    merge_user_replication_config(Some(incoming.clone()), existing, site_peer_deployment_ids).unwrap_or(incoming)
+}
+
+/// Split of an S3 DeleteBucketReplication on the stored config (issue #1948):
+/// the operator-authored rules are removed, the rules the reconciler derived
+/// for the current remote peers survive (`None` means nothing survives and
+/// the config is deleted), and the returned ARNs are the ones whose bucket
+/// targets may be garbage-collected — never an ARN a surviving reconciler
+/// rule still points at.
+fn split_replication_config_for_user_delete(
+    config: ReplicationConfiguration,
+    site_peer_deployment_ids: &HashSet<String>,
+) -> (Option<ReplicationConfiguration>, HashSet<String>) {
+    let mut removable_arns = replication_target_arns(&config);
+    let remaining = merge_user_replication_config(None, Some(config), site_peer_deployment_ids);
+    if let Some(remaining) = remaining.as_ref() {
+        for rule in &remaining.rules {
+            removable_arns.remove(rule.destination.bucket.trim());
+        }
+    }
+    (remaining, removable_arns)
+}
+
+async fn replication_targets_without_arns(
     bucket: &str,
-    config: &ReplicationConfiguration,
+    target_arns: &HashSet<String>,
 ) -> S3Result<Option<(BucketTargets, usize)>> {
-    let target_arns = replication_target_arns(config);
     if target_arns.is_empty() {
         return Ok(None);
     }
@@ -638,7 +679,7 @@ async fn replication_targets_without_config_targets(
         Err(err) => return Err(ApiError::from(err).into()),
     };
 
-    let removed = remove_replication_targets_from_config_targets(&mut targets, &target_arns);
+    let removed = remove_replication_targets_from_config_targets(&mut targets, target_arns);
     if removed == 0 {
         return Ok(None);
     }
@@ -1582,9 +1623,15 @@ impl DefaultBucketUsecase {
         Ok(S3Response::new(DeleteBucketPolicyOutput {}))
     }
 
+    /// `site_peers` is the set of remote site-replication peer deployment ids
+    /// (empty when site replication is disabled). The interface layer reads it
+    /// from the persisted state and fails closed on a read error, so this
+    /// usecase stays a pure function of its inputs (layer rule: app never
+    /// imports interface).
     pub async fn execute_delete_bucket_replication(
         &self,
         req: S3Request<DeleteBucketReplicationInput>,
+        site_peers: HashSet<String>,
     ) -> S3Result<S3Response<DeleteBucketReplicationOutput>> {
         let expected_incarnation_id = bucket_config_mutation_incarnation(&req, &req.input.bucket)?;
         let request_context = req.extensions.get::<request_context::RequestContext>().cloned();
@@ -1604,15 +1651,29 @@ impl DefaultBucketUsecase {
             Err(StorageError::ConfigNotFound) => None,
             Err(err) => return Err(ApiError::from(err).into()),
         };
-        let updated_targets = if let Some(config) = replication_config.as_ref() {
-            replication_targets_without_config_targets(&bucket, config).await?
+        let (remaining_config, updated_targets) = if let Some(config) = replication_config.as_ref() {
+            let (remaining, removable_arns) = split_replication_config_for_user_delete(config.clone(), &site_peers);
+            let targets = replication_targets_without_arns(&bucket, &removable_arns).await?;
+            (remaining, targets)
         } else {
-            None
+            (None, None)
         };
 
-        delete_bucket_config_for_incarnation(&bucket, BUCKET_REPLICATION_CONFIG, expected_incarnation_id)
-            .await
-            .map_err(ApiError::from)?;
+        match remaining_config {
+            // Site-replication rules and the targets backing them survive the
+            // S3 delete (issue #1948); only the operator-authored rules go.
+            Some(remaining) => {
+                let data = serialize_config(&remaining)?;
+                update_bucket_config_for_incarnation(&bucket, BUCKET_REPLICATION_CONFIG, data, expected_incarnation_id)
+                    .await
+                    .map_err(ApiError::from)?;
+            }
+            None => {
+                delete_bucket_config_for_incarnation(&bucket, BUCKET_REPLICATION_CONFIG, expected_incarnation_id)
+                    .await
+                    .map_err(ApiError::from)?;
+            }
+        }
         if let Some((targets, removed)) = updated_targets
             && let Err(err) =
                 write_replication_targets_after_config_delete(&bucket, &targets, removed, expected_incarnation_id).await
@@ -2459,9 +2520,11 @@ impl DefaultBucketUsecase {
         Ok(S3Response::new(PutBucketCorsOutput::default()))
     }
 
+    /// See [`Self::execute_delete_bucket_replication`] for `site_peers`.
     pub async fn execute_put_bucket_replication(
         &self,
         req: S3Request<PutBucketReplicationInput>,
+        site_peers: HashSet<String>,
     ) -> S3Result<S3Response<PutBucketReplicationOutput>> {
         let expected_incarnation_id = bucket_config_mutation_incarnation(&req, &req.input.bucket)?;
         let request_context = req.extensions.get::<request_context::RequestContext>().cloned();
@@ -2485,6 +2548,13 @@ impl DefaultBucketUsecase {
 
         let targets_guard = lock_bucket_targets_metadata(&bucket).await;
         validate_bucket_replication_update(&bucket, &replication_configuration).await?;
+        let existing_config = match metadata_sys::get_replication_config(&bucket).await {
+            Ok((config, _)) => Some(config),
+            Err(StorageError::ConfigNotFound) => None,
+            Err(err) => return Err(ApiError::from(err).into()),
+        };
+        let replication_configuration =
+            merge_user_replication_config_update(replication_configuration, existing_config, &site_peers);
         let data = serialize_config(&replication_configuration)?;
         update_bucket_config_for_incarnation(&bucket, BUCKET_REPLICATION_CONFIG, data, expected_incarnation_id)
             .await
@@ -3114,6 +3184,185 @@ mod tests {
         assert!(arns.contains(destination));
     }
 
+    fn replication_rule_with_id(arn: &str, id: &str, priority: i32) -> ReplicationRule {
+        let mut rule = replication_rule_for_target(arn);
+        rule.id = Some(id.to_string());
+        rule.priority = Some(priority);
+        rule
+    }
+
+    fn site_peers(deployment_ids: &[&str]) -> HashSet<String> {
+        deployment_ids.iter().map(|id| id.to_string()).collect()
+    }
+
+    #[test]
+    fn put_replication_merge_preserves_site_replication_rules() {
+        let existing = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                replication_rule_with_id("arn:rustfs:replication::peer-dep:bucket", "site-repl-peer-dep", 1),
+                replication_rule_with_id("arn:rustfs:replication:us-east-1:old:bucket", "old-user-rule", 2),
+            ],
+        };
+        let incoming = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                replication_rule_with_id("arn:rustfs:replication:us-east-1:new:bucket", "new-user-rule", 1),
+                replication_rule_with_id("arn:rustfs:replication::forged-dep:bucket", "site-repl-peer-dep", 2),
+                replication_rule_with_id("arn:rustfs:replication::other-dep:bucket", "site-repl-other", 3),
+            ],
+        };
+
+        let merged = merge_user_replication_config_update(incoming, Some(existing), &site_peers(&["peer-dep"]));
+
+        let rules: Vec<_> = merged
+            .rules
+            .iter()
+            .map(|rule| (rule.id.as_deref().unwrap_or_default(), rule.destination.bucket.as_str()))
+            .collect();
+        assert_eq!(
+            rules,
+            vec![
+                ("new-user-rule", "arn:rustfs:replication:us-east-1:new:bucket"),
+                ("site-repl-other", "arn:rustfs:replication::other-dep:bucket"),
+                ("site-repl-peer-dep", "arn:rustfs:replication::peer-dep:bucket"),
+            ],
+            "user rules replaced, the reconciler rule for the current peer kept over the incoming impostor, \
+             a site-repl-* id that names no current peer is ordinary operator state"
+        );
+    }
+
+    // Rule ids do not reserve `site-repl-*`: outside site replication an
+    // owner's `site-repl-user` rule is ordinary state, so PUT stores it
+    // verbatim and DELETE removes it and garbage-collects its target.
+    #[test]
+    fn put_then_delete_replication_without_site_replication_treats_site_repl_id_as_user_rule() {
+        let user_arn = "arn:minio:replication:us-east-1:2f1c-remote:bucket";
+        let incoming = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![replication_rule_with_id(user_arn, "site-repl-user", 1)],
+        };
+
+        let stored = merge_user_replication_config_update(incoming.clone(), None, &HashSet::new());
+        assert_eq!(stored, incoming, "PUT on a non-site-replication bucket is verbatim");
+
+        let (remaining, removable) = split_replication_config_for_user_delete(stored, &HashSet::new());
+        assert!(remaining.is_none(), "DELETE must remove the operator's site-repl-* rule");
+        assert_eq!(removable, HashSet::from([user_arn.to_string()]));
+    }
+
+    // Under site replication only a rule the reconciler would derive — id
+    // `site-repl-<peer>` for a current peer, destination ARN naming the same
+    // peer — is reconciler-owned. Everything else is operator state.
+    #[test]
+    fn delete_replication_split_keeps_only_reconciler_derived_rules() {
+        let peer_arn = "arn:rustfs:replication::peer-dep:bucket";
+        let user_arn = "arn:minio:replication:us-east-1:2f1c-remote:bucket";
+        let config = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                replication_rule_with_id(user_arn, "site-repl-user", 1),
+                replication_rule_with_id(user_arn, "site-repl-peer-dep", 2),
+                replication_rule_with_id("arn:rustfs:replication::gone-dep:bucket", "site-repl-gone-dep", 3),
+                replication_rule_with_id(peer_arn, "site-repl-peer-dep", 4),
+            ],
+        };
+
+        let (remaining, removable) = split_replication_config_for_user_delete(config, &site_peers(&["peer-dep"]));
+
+        let remaining = remaining.expect("the reconciler-derived rule must survive");
+        assert_eq!(remaining.rules.len(), 1);
+        assert_eq!(remaining.rules[0].destination.bucket, peer_arn);
+        assert_eq!(
+            removable,
+            HashSet::from([user_arn.to_string(), "arn:rustfs:replication::gone-dep:bucket".to_string()]),
+            "targets of operator rules and of a removed peer are garbage-collected"
+        );
+    }
+
+    #[test]
+    fn put_replication_merge_returns_incoming_verbatim_without_site_rules() {
+        let existing = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![replication_rule_with_id(
+                "arn:rustfs:replication:us-east-1:old:bucket",
+                "old-user-rule",
+                7,
+            )],
+        };
+        let incoming = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![replication_rule_with_id(
+                "arn:rustfs:replication:us-east-1:new:bucket",
+                "new-user-rule",
+                5,
+            )],
+        };
+
+        let merged = merge_user_replication_config_update(incoming.clone(), Some(existing), &HashSet::new());
+
+        assert_eq!(merged.role, incoming.role);
+        assert_eq!(merged.rules, incoming.rules, "non-SR buckets keep the verbatim overwrite semantics");
+    }
+
+    #[test]
+    fn delete_replication_split_keeps_site_rules_and_their_targets() {
+        let sr_arn = "arn:rustfs:replication::peer-dep:bucket";
+        let user_arn = "arn:rustfs:replication:us-east-1:user:bucket";
+        let config = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                replication_rule_with_id(user_arn, "user-rule", 1),
+                replication_rule_with_id(sr_arn, "site-repl-peer-dep", 2),
+            ],
+        };
+
+        let (remaining, removable) = split_replication_config_for_user_delete(config, &site_peers(&["peer-dep"]));
+
+        let remaining = remaining.expect("site-replication rules must survive a user delete");
+        let ids: Vec<_> = remaining
+            .rules
+            .iter()
+            .map(|rule| rule.id.as_deref().unwrap_or_default())
+            .collect();
+        assert_eq!(ids, vec!["site-repl-peer-dep"]);
+        assert_eq!(removable, HashSet::from([user_arn.to_string()]));
+    }
+
+    #[test]
+    fn delete_replication_split_protects_targets_shared_with_site_rules() {
+        let sr_arn = "arn:rustfs:replication::peer-dep:bucket";
+        let config = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![
+                replication_rule_with_id(sr_arn, "user-rule-on-sr-target", 1),
+                replication_rule_with_id(sr_arn, "site-repl-peer-dep", 2),
+            ],
+        };
+
+        let (remaining, removable) = split_replication_config_for_user_delete(config, &site_peers(&["peer-dep"]));
+
+        assert!(remaining.is_some());
+        assert!(
+            removable.is_empty(),
+            "a target still referenced by a surviving site-replication rule must not be removed"
+        );
+    }
+
+    #[test]
+    fn delete_replication_split_removes_everything_without_site_rules() {
+        let user_arn = "arn:rustfs:replication:us-east-1:user:bucket";
+        let config = ReplicationConfiguration {
+            role: String::new(),
+            rules: vec![replication_rule_with_id(user_arn, "user-rule", 1)],
+        };
+
+        let (remaining, removable) = split_replication_config_for_user_delete(config, &site_peers(&["peer-dep"]));
+
+        assert!(remaining.is_none(), "without site-replication rules the whole config is deleted");
+        assert_eq!(removable, HashSet::from([user_arn.to_string()]));
+    }
+
     fn replication_targets_with_arn(arns: &[&str]) -> BucketTargets {
         BucketTargets {
             targets: arns
@@ -3451,7 +3700,10 @@ mod tests {
         let req = build_request(input, Method::DELETE);
         let usecase = DefaultBucketUsecase::without_context();
 
-        let err = usecase.execute_delete_bucket_replication(req).await.unwrap_err();
+        let err = usecase
+            .execute_delete_bucket_replication(req, HashSet::new())
+            .await
+            .unwrap_err();
         assert_eq!(err.code(), &S3ErrorCode::InternalError);
     }
 
@@ -4537,7 +4789,7 @@ mod tests {
         let req = build_request(input, Method::PUT);
         let usecase = DefaultBucketUsecase::without_context();
 
-        let err = usecase.execute_put_bucket_replication(req).await.unwrap_err();
+        let err = usecase.execute_put_bucket_replication(req, HashSet::new()).await.unwrap_err();
         assert_eq!(err.code(), &S3ErrorCode::InternalError);
     }
 
@@ -4555,7 +4807,7 @@ mod tests {
             .unwrap();
 
         let err = DefaultBucketUsecase::without_context()
-            .execute_put_bucket_replication(build_request(input, Method::PUT))
+            .execute_put_bucket_replication(build_request(input, Method::PUT), HashSet::new())
             .await
             .expect_err("unsupported fields must be rejected before store access");
 

rustfs/src/app/storage_api.rs

@@ -619,6 +619,8 @@ pub(crate) mod bucket {
 
         use crate::storage::storage_api::ecstore_bucket::replication as replication_contracts;
 
+        pub(crate) use replication_contracts::merge_user_replication_config;
+
         type ReplicationObjectBridge = crate::storage::storage_api::ecstore_bucket::replication::ReplicationObjectBridge;
         pub(crate) type DeleteReplicationConfigSnapshot =
             crate::storage::storage_api::ecstore_bucket::replication::DeleteReplicationConfigSnapshot;

rustfs/src/main.rs

@@ -26,22 +26,22 @@ struct MiMallocAllocator;
 unsafe impl GlobalAlloc for MiMallocAllocator {
     unsafe fn alloc(&self, layout: Layout) -> *mut u8 {
         // SAFETY: the caller upholds GlobalAlloc's contract for layout.
-        unsafe { mimalloc::MiMalloc.alloc(layout) }
+        unsafe { rustfs_mimalloc::MiMalloc.alloc(layout) }
     }
 
     unsafe fn alloc_zeroed(&self, layout: Layout) -> *mut u8 {
         // SAFETY: the caller upholds GlobalAlloc's contract for layout.
-        unsafe { mimalloc::MiMalloc.alloc_zeroed(layout) }
+        unsafe { rustfs_mimalloc::MiMalloc.alloc_zeroed(layout) }
     }
 
     unsafe fn dealloc(&self, ptr: *mut u8, layout: Layout) {
         // SAFETY: ptr and layout came from this allocator and are forwarded unchanged.
-        unsafe { mimalloc::MiMalloc.dealloc(ptr, layout) }
+        unsafe { rustfs_mimalloc::MiMalloc.dealloc(ptr, layout) }
     }
 
     unsafe fn realloc(&self, ptr: *mut u8, layout: Layout, new_size: usize) -> *mut u8 {
         // SAFETY: ptr and layout came from this allocator and are forwarded unchanged.
-        unsafe { mimalloc::MiMalloc.realloc(ptr, layout, new_size) }
+        unsafe { rustfs_mimalloc::MiMalloc.realloc(ptr, layout, new_size) }
     }
 }
 
@@ -51,7 +51,7 @@ static GLOBAL: hotpath::CountingAllocator<MiMallocAllocator> = hotpath::Counting
 
 #[cfg(not(all(feature = "hotpath", feature = "hotpath-alloc")))]
 #[global_allocator]
-static GLOBAL: mimalloc::MiMalloc = mimalloc::MiMalloc;
+static GLOBAL: rustfs_mimalloc::MiMalloc = rustfs_mimalloc::MiMalloc;
 
 fn main() {
     let _hotpath_guard = hotpath::HotpathGuardBuilder::new("main").build();
@@ -71,8 +71,9 @@ mod tests {
         allocation.extend_from_slice(&[7_u8; 64]);
 
         assert_eq!(allocation.len(), 64);
+        let heap = rustfs_mimalloc::heap::Heap::main();
         // SAFETY: the live Vec pointer is valid to inspect for heap ownership.
-        assert!(unsafe { libmimalloc_sys::mi_is_in_heap_region(allocation.as_ptr().cast()) });
+        assert!(unsafe { heap.contains(allocation.as_ptr()) });
     }
 
     #[test]
@@ -85,12 +86,13 @@ mod tests {
         let layout = Layout::from_size_align(32, 8).expect("valid test allocation layout");
         let grown_layout = Layout::from_size_align(64, 8).expect("valid grown test allocation layout");
         let allocator = super::MiMallocAllocator;
+        let heap = rustfs_mimalloc::heap::Heap::main();
 
         // SAFETY: The pointer is checked for null before use and later released
         // through the same allocator with the corresponding layout.
         let ptr = unsafe { allocator.alloc_zeroed(layout) };
         assert!(!ptr.is_null());
-        assert!(unsafe { libmimalloc_sys::mi_is_in_heap_region(ptr.cast()) });
+        assert!(unsafe { heap.contains(ptr) });
         assert!(unsafe { std::slice::from_raw_parts(ptr, 32).iter().all(|byte| *byte == 0) });
 
         // SAFETY: `ptr` was allocated by `allocator` with `layout`; on failure
@@ -102,7 +104,7 @@ mod tests {
             panic!("mimalloc realloc failed in allocator smoke test");
         }
 
-        assert!(unsafe { libmimalloc_sys::mi_is_in_heap_region(grown_ptr.cast()) });
+        assert!(unsafe { heap.contains(grown_ptr) });
         // SAFETY: `grown_ptr` was reallocated by `allocator` and is released
         // with the matching grown layout.
         unsafe { allocator.dealloc(grown_ptr, grown_layout) };

rustfs/src/memory_observability.rs

@@ -17,10 +17,7 @@ use rustfs_io_metrics::{
     record_cpu_usage, record_memory_usage, record_process_memory_split,
 };
 use serde::Serialize;
-#[cfg(any(test, not(target_os = "windows")))]
 use serde_json::Value;
-#[cfg(not(target_os = "windows"))]
-use std::ffi::CStr;
 use std::path::Path;
 use std::sync::{Arc, Mutex, OnceLock};
 use std::time::Duration;
@@ -231,7 +228,18 @@ fn read_cgroup_memory_snapshot() -> Option<CgroupMemorySnapshot> {
     read_cgroup_v2().or_else(read_cgroup_v1)
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
+fn read_allocator_memory_snapshot() -> Option<AllocatorMemorySnapshot> {
+    let json = rustfs_mimalloc::MiMalloc::stats_json();
+    if json.is_empty() {
+        return None;
+    }
+    let observation = parse_mimalloc_stats_json(&json)?;
+    Some(AllocatorMemorySnapshot {
+        backend: crate::allocator_reclaim::allocator_backend(),
+        observation,
+    })
+}
+
 fn numeric_json_value(value: &Value) -> Option<u64> {
     match value {
         Value::Number(number) => number
@@ -242,7 +250,6 @@ fn numeric_json_value(value: &Value) -> Option<u64> {
     }
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
 fn numeric_json_field(value: &Value, field: &str) -> Option<u64> {
     match value {
         Value::Object(fields) => fields
@@ -254,7 +261,6 @@ fn numeric_json_field(value: &Value, field: &str) -> Option<u64> {
     }
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
 fn mimalloc_stat_field(value: &Value, metric: &str, field: &str) -> Option<u64> {
     match value {
         Value::Object(fields) => {
@@ -271,12 +277,10 @@ fn mimalloc_stat_field(value: &Value, metric: &str, field: &str) -> Option<u64>
     }
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
 fn mimalloc_stat_current(value: &Value, metric: &str) -> Option<u64> {
     mimalloc_stat_field(value, metric, "current")
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
 fn mimalloc_stat_sum(value: &Value, metrics: &[&str], field: &str) -> Option<u64> {
     metrics
         .iter()
@@ -285,7 +289,6 @@ fn mimalloc_stat_sum(value: &Value, metrics: &[&str], field: &str) -> Option<u64
         .filter(|value| *value > 0)
 }
 
-#[cfg(any(test, not(target_os = "windows")))]
 fn parse_mimalloc_stats_json(stats_json: &str) -> Option<AllocatorMemoryObservation> {
     let value = serde_json::from_str::<Value>(stats_json).ok()?;
     let malloc_metrics = ["malloc_normal", "malloc_huge"];
@@ -312,33 +315,6 @@ fn parse_mimalloc_stats_json(stats_json: &str) -> Option<AllocatorMemoryObservat
     }
 }
 
-#[cfg(not(target_os = "windows"))]
-#[allow(unsafe_code)]
-fn read_allocator_memory_snapshot() -> Option<AllocatorMemorySnapshot> {
-    // SAFETY: `mi_stats_get_json` returns a null-terminated JSON buffer owned by
-    // mimalloc when called with a null input buffer. The mimalloc API requires
-    // freeing that buffer with `mi_free`; parsing finishes before the buffer is freed.
-    let observation = unsafe {
-        let stats_ptr = libmimalloc_sys::mi_stats_get_json(0, std::ptr::null_mut());
-        if stats_ptr.is_null() {
-            return None;
-        }
-
-        let observation = CStr::from_ptr(stats_ptr).to_str().ok().and_then(parse_mimalloc_stats_json);
-        libmimalloc_sys::mi_free(stats_ptr.cast());
-        observation?
-    };
-    Some(AllocatorMemorySnapshot {
-        backend: crate::allocator_reclaim::allocator_backend(),
-        observation,
-    })
-}
-
-#[cfg(target_os = "windows")]
-fn read_allocator_memory_snapshot() -> Option<AllocatorMemorySnapshot> {
-    None
-}
-
 fn configured_memory_observability_interval_secs() -> u64 {
     rustfs_utils::get_env_u64(ENV_MEMORY_OBSERVABILITY_INTERVAL_SECS, DEFAULT_MEMORY_OBSERVABILITY_INTERVAL_SECS).max(1)
 }
@@ -566,6 +542,13 @@ mod tests {
         assert_eq!(parse_mimalloc_stats_json(r#"{ "allocator": "unknown" }"#), None);
     }
 
+    #[test]
+    fn read_allocator_memory_snapshot_uses_mimalloc_stats_json() {
+        let snapshot = super::read_allocator_memory_snapshot();
+        #[cfg(not(target_os = "windows"))]
+        assert!(snapshot.is_some(), "allocator snapshot should be available on non-Windows");
+    }
+
     #[test]
     fn memory_observability_snapshot_reports_disabled_when_metrics_are_disabled() {
         let snapshot = build_memory_observability_status_snapshot(false, 15, false);

rustfs/src/storage/ecfs.rs

@@ -63,6 +63,69 @@ use crate::app::storage_api::object_usecase::bucket::replication::{
 };
 use crate::storage::storage_api::ecfs_consumer::StorageObjectOptions as ObjectOptions;
 
+#[cfg(test)]
+static SITE_REPLICATION_GATE_TEST_OVERRIDE: std::sync::atomic::AtomicU8 = std::sync::atomic::AtomicU8::new(0);
+#[cfg(test)]
+const SITE_REPLICATION_GATE_FORCE_DISABLED: u8 = 1;
+#[cfg(test)]
+const SITE_REPLICATION_GATE_FORCE_ENABLED: u8 = 2;
+
+async fn site_replication_gate_enabled() -> S3Result<bool> {
+    #[cfg(test)]
+    match SITE_REPLICATION_GATE_TEST_OVERRIDE.load(std::sync::atomic::Ordering::SeqCst) {
+        SITE_REPLICATION_GATE_FORCE_DISABLED => return Ok(false),
+        SITE_REPLICATION_GATE_FORCE_ENABLED => return Ok(true),
+        _ => {}
+    }
+    crate::admin::handlers::site_replication::site_replication_enabled().await
+}
+
+/// Remote site-replication peer deployment ids handed to the bucket usecase
+/// so an S3 replication-config edit keeps exactly the reconciler-owned rules
+/// (issue #1948). Read here, in the interface layer, because the usecase must
+/// not import the admin handlers (layer guard); a state-read failure
+/// propagates so the edit fails closed.
+async fn site_replication_peer_deployment_ids_for_edit() -> S3Result<std::collections::HashSet<String>> {
+    // While the gate override is in effect the test exercises the deny/allow
+    // branch, not the peer set; there is no persisted state to read.
+    #[cfg(test)]
+    if SITE_REPLICATION_GATE_TEST_OVERRIDE.load(std::sync::atomic::Ordering::SeqCst) != 0 {
+        return Ok(std::collections::HashSet::new());
+    }
+    crate::admin::handlers::site_replication::site_replication_remote_peer_deployment_ids().await
+}
+
+/// MinIO `ErrReplicationDenyEditError`.
+fn replication_deny_edit_error() -> S3Error {
+    let mut err = S3Error::with_message(
+        S3ErrorCode::Custom("XMinioReplicationDenyEdit".into()),
+        "Sub-User is not allowed to edit Replication configuration",
+    );
+    err.set_status_code(StatusCode::BAD_REQUEST);
+    err
+}
+
+/// Site-replication gate for S3 replication-config edits (issue #1948).
+///
+/// On a site-replication deployment the bucket's replication config carries
+/// the operator-managed `site-repl-*` rules that keep every peer in sync, and
+/// a successful edit is broadcast to all peers — so a user holding only
+/// bucket-scoped `s3:PutReplicationConfiguration` could rewrite or erase
+/// replication net-wide. MinIO parity (`ErrReplicationDenyEditError`): only
+/// owner credentials (root or root-parented) may edit. Runs after the policy
+/// authorization in the access layer and only on the external S3 path — the
+/// reconciler and peer bucket-meta ingestion never route through these
+/// handlers.
+async fn deny_replication_config_edit_for_non_owner<T>(req: &S3Request<T>) -> S3Result<()> {
+    if crate::storage::access::req_info_ref(req)?.is_owner {
+        return Ok(());
+    }
+    if site_replication_gate_enabled().await? {
+        return Err(replication_deny_edit_error());
+    }
+    Ok(())
+}
+
 #[derive(Debug, Clone)]
 pub struct FS {
     /// This server's late-bound application-context slot (backlog#1052 S2).
@@ -500,8 +563,10 @@ impl S3 for FS {
         &self,
         req: S3Request<DeleteBucketReplicationInput>,
     ) -> S3Result<S3Response<DeleteBucketReplicationOutput>> {
+        deny_replication_config_edit_for_non_owner(&req).await?;
+        let site_peers = site_replication_peer_deployment_ids_for_edit().await?;
         let usecase = s3_api::bucket_usecase_for(self);
-        usecase.execute_delete_bucket_replication(req).await
+        usecase.execute_delete_bucket_replication(req, site_peers).await
     }
 
     #[instrument(level = "debug", skip(self))]
@@ -1353,8 +1418,10 @@ impl S3 for FS {
         &self,
         req: S3Request<PutBucketReplicationInput>,
     ) -> S3Result<S3Response<PutBucketReplicationOutput>> {
+        deny_replication_config_edit_for_non_owner(&req).await?;
+        let site_peers = site_replication_peer_deployment_ids_for_edit().await?;
         let usecase = s3_api::bucket_usecase_for(self);
-        usecase.execute_put_bucket_replication(req).await
+        usecase.execute_put_bucket_replication(req, site_peers).await
     }
 
     async fn put_bucket_request_payment(
@@ -1919,3 +1986,103 @@ impl S3 for FS {
         Box::pin(usecase.execute_upload_part_copy(req)).await
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::{
+        FS, SITE_REPLICATION_GATE_FORCE_DISABLED, SITE_REPLICATION_GATE_FORCE_ENABLED, SITE_REPLICATION_GATE_TEST_OVERRIDE,
+    };
+    use crate::storage::access::ReqInfo;
+    use http::Method;
+    use http::StatusCode;
+    use s3s::dto::{DeleteBucketReplicationInput, PutBucketReplicationInput, ReplicationConfiguration};
+    use s3s::{S3, S3Error, S3ErrorCode, S3Request};
+    use std::sync::atomic::Ordering;
+
+    fn replication_config_edit_request<T>(input: T, is_owner: bool) -> S3Request<T> {
+        let mut req = S3Request {
+            input,
+            method: Method::PUT,
+            uri: http::Uri::from_static("/"),
+            headers: http::HeaderMap::new(),
+            extensions: http::Extensions::new(),
+            credentials: None,
+            region: None,
+            service: None,
+            trailing_headers: None,
+        };
+        req.extensions.insert(ReqInfo {
+            is_owner,
+            ..Default::default()
+        });
+        req
+    }
+
+    fn put_bucket_replication_input() -> PutBucketReplicationInput {
+        PutBucketReplicationInput {
+            bucket: "test-bucket".to_string(),
+            checksum_algorithm: None,
+            content_md5: None,
+            expected_bucket_owner: None,
+            replication_configuration: ReplicationConfiguration {
+                role: String::new(),
+                rules: Vec::new(),
+            },
+            token: None,
+        }
+    }
+
+    fn delete_bucket_replication_input() -> DeleteBucketReplicationInput {
+        DeleteBucketReplicationInput {
+            bucket: "test-bucket".to_string(),
+            expected_bucket_owner: None,
+        }
+    }
+
+    fn assert_replication_deny_edit(err: &S3Error) {
+        match err.code() {
+            S3ErrorCode::Custom(code) => assert_eq!(code, "XMinioReplicationDenyEdit"),
+            other => panic!("expected XMinioReplicationDenyEdit, got {other:?}"),
+        }
+        assert_eq!(err.status_code(), Some(StatusCode::BAD_REQUEST));
+    }
+
+    /// Single test on purpose: the branches share the process-wide gate
+    /// override, and parallel tests would race it.
+    #[tokio::test]
+    async fn replication_config_edit_gate_denies_only_non_owner_under_site_replication() {
+        let fs = FS::new();
+        SITE_REPLICATION_GATE_TEST_OVERRIDE.store(SITE_REPLICATION_GATE_FORCE_ENABLED, Ordering::SeqCst);
+
+        // Non-owner PUT/DELETE through the real S3 handlers: denied by the
+        // gate before the usecase (and thus the store) is ever touched.
+        let err = fs
+            .put_bucket_replication(replication_config_edit_request(put_bucket_replication_input(), false))
+            .await
+            .expect_err("non-owner PutBucketReplication must be denied while site replication is enabled");
+        assert_replication_deny_edit(&err);
+        let err = fs
+            .delete_bucket_replication(replication_config_edit_request(delete_bucket_replication_input(), false))
+            .await
+            .expect_err("non-owner DeleteBucketReplication must be denied while site replication is enabled");
+        assert_replication_deny_edit(&err);
+
+        // Owner passes the gate (the usecase's empty-rules structure error
+        // proves the request reached the usecase instead of the deny path).
+        let err = fs
+            .put_bucket_replication(replication_config_edit_request(put_bucket_replication_input(), true))
+            .await
+            .expect_err("owner request should pass the gate and fail later on config validation");
+        assert_eq!(err.code(), &S3ErrorCode::InvalidRequest);
+
+        // Without site replication the policy check alone still governs the edit.
+        SITE_REPLICATION_GATE_TEST_OVERRIDE.store(SITE_REPLICATION_GATE_FORCE_DISABLED, Ordering::SeqCst);
+        let err = fs
+            .put_bucket_replication(replication_config_edit_request(put_bucket_replication_input(), false))
+            .await
+            .expect_err("non-owner request should pass the gate and fail later on config validation");
+        assert_eq!(err.code(), &S3ErrorCode::InvalidRequest);
+
+        SITE_REPLICATION_GATE_TEST_OVERRIDE.store(0, Ordering::SeqCst);
+    }
+}