test(ecstore): cover final sweep cancel fence

# Conflicts:
#	crates/ecstore/src/core/pools.rs

.config/nextest.toml

@@ -78,6 +78,12 @@ test-group = 'embedded-test-ports'
 filter = 'package(rustfs-ecstore) & test(manual_transition_page_checkpoint_persists_durable_job_progress)'
 test-group = 'ecstore-serial-flaky'
 
+# The durable ILM decommission regressions build isolated multi-pool stores and
+# deliberately take source or target disks offline while checking fencing.
+[[profile.default.overrides]]
+filter = 'package(rustfs-ecstore) & (test(decommission_migrates_and_verifies_registered_durable_ilm_records) | test(decommission_durable_ilm_target_read_error_is_not_masked_by_peer_success) | test(decommission_durable_ilm_terminal_receipt_recovers_failed_source_cleanup) | test(decommission_durable_ilm_receipt_pagination_fails_closed_on_second_page) | test(decommission_durable_ilm_recovery_keeps_multiple_active_sources))'
+test-group = 'ecstore-serial-flaky'
+
 # Serialize the bucket-incarnation / lifecycle-fence tests. They drive
 # init_bucket_metadata_sys and bucket_metadata_sys_of, i.e. process-global
 # OnceLock state that serial_test's #[serial] cannot protect across nextest's
@@ -190,6 +196,10 @@ test-group = 'embedded-test-ports'
 filter = 'package(rustfs-ecstore) & test(manual_transition_page_checkpoint_persists_durable_job_progress)'
 test-group = 'ecstore-serial-flaky'
 
+[[profile.ci.overrides]]
+filter = 'package(rustfs-ecstore) & (test(decommission_migrates_and_verifies_registered_durable_ilm_records) | test(decommission_durable_ilm_target_read_error_is_not_masked_by_peer_success) | test(decommission_durable_ilm_terminal_receipt_recovers_failed_source_cleanup) | test(decommission_durable_ilm_receipt_pagination_fails_closed_on_second_page) | test(decommission_durable_ilm_recovery_keeps_multiple_active_sources))'
+test-group = 'ecstore-serial-flaky'
+
 # Serialize the bucket-incarnation / lifecycle-fence tests under the ci profile
 # too (see the matching default-profile override near the top). No retries.
 [[profile.ci.overrides]]

CLAUDE.md

@@ -30,8 +30,7 @@ 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 workflow steps: `.github/workflows/`; event, timeout, and required-status
-  matrix: [docs/testing/ci-gates.md](docs/testing/ci-gates.md)
+- CI gates: `.github/workflows/ci.yml` (source of truth; never copy its steps into docs)
 - 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

@@ -70,8 +70,6 @@ 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

@@ -12688,7 +12688,6 @@ dependencies = [
  "js-sys",
  "rand 0.10.2",
  "serde_core",
- "sha1_smol",
  "wasm-bindgen",
 ]
 

crates/ecstore/src/bucket/lifecycle/bucket_lifecycle_ops.rs

@@ -9136,6 +9136,7 @@ mod tests {
         assert_eq!(loaded.report.scanned, 37);
         assert_eq!(loaded.report.eligible, 11);
         assert_eq!(loaded.report.enqueued, 5);
+        assert_eq!(loaded.cursor_revision, Some(1));
         assert!(loaded.lease_expires_at_unix_nanos > 0);
         let token = loaded
             .report
@@ -9154,6 +9155,30 @@ mod tests {
         assert_eq!(admission.lease_id, loaded.lease_id);
         assert_eq!(admission.lease_expires_at_unix_nanos, loaded.lease_expires_at_unix_nanos);
 
+        let mut same_marker_report = report.clone();
+        same_marker_report.scanned += 1;
+        persist_manual_transition_page_checkpoint(
+            &checkpoint_options,
+            &same_marker_report,
+            Some("logs/page-end".to_string()),
+            Some("opaque-next-version".to_string()),
+        )
+        .await
+        .expect("same-marker version checkpoint should persist through the durable progress sink");
+        let same_marker_checkpointed = load_manual_transition_job_record(ecstore.clone(), job_id)
+            .await
+            .expect("same-marker version checkpoint should reload");
+        assert_eq!(same_marker_checkpointed.cursor_revision, Some(2));
+        let (_, version_marker) = decode_manual_transition_continuation_token(
+            same_marker_checkpointed
+                .report
+                .continuation_token
+                .as_deref()
+                .expect("same-marker version checkpoint should persist a cursor"),
+        )
+        .expect("same-marker version cursor should decode");
+        assert_eq!(version_marker.as_deref(), Some("opaque-next-version"));
+
         create_test_bucket(&ecstore, &bucket).await;
         let lifecycle_xml = format!(
             r#"<?xml version="1.0" encoding="UTF-8"?>
@@ -9210,6 +9235,7 @@ mod tests {
         assert_eq!(checkpointed.report.scanned, 1000);
         assert_eq!(checkpointed.report.eligible, 1000);
         assert_eq!(checkpointed.report.dry_run_eligible, 1000);
+        assert_eq!(checkpointed.cursor_revision, Some(3));
         let token = checkpointed
             .report
             .continuation_token

crates/ecstore/src/bucket/lifecycle/manual_transition_job.rs

@@ -24,6 +24,10 @@ use crate::bucket::lifecycle::bucket_lifecycle_ops::{
     ManualTransitionQueueSnapshot, ManualTransitionRunOptions, ManualTransitionRunReport,
 };
 use crate::bucket::lifecycle::config_boundary;
+use crate::bucket::lifecycle::durable_namespace::{
+    MANUAL_TRANSITION_JOB_NAMESPACE, MANUAL_TRANSITION_SCOPE_NAMESPACE, MANUAL_TRANSITION_TASK_NAMESPACE,
+    MANUAL_TRANSITION_WORKER_RESULT_NAMESPACE,
+};
 use crate::disk::RUSTFS_META_BUCKET;
 use crate::error::{Error, Result as EcstoreResult};
 use crate::object_api::ObjectOptions;
@@ -34,10 +38,10 @@ use crate::store::ECStore;
 pub const MANUAL_TRANSITION_JOB_SCHEMA: &str = "rustfs-manual-transition-job-v1";
 pub const MANUAL_TRANSITION_TASK_SCHEMA: &str = "rustfs-manual-transition-task-v1";
 pub const MANUAL_TRANSITION_WORKER_RESULT_SCHEMA: &str = "rustfs-manual-transition-worker-result-v1";
-pub const MANUAL_TRANSITION_JOB_RECORD_PREFIX: &str = "ilm/manual-transition/jobs";
-pub const MANUAL_TRANSITION_SCOPE_RECORD_PREFIX: &str = "ilm/manual-transition/scopes";
-pub const MANUAL_TRANSITION_TASK_PREFIX: &str = "ilm/manual-transition/tasks";
-pub const MANUAL_TRANSITION_WORKER_RESULT_PREFIX: &str = "ilm/manual-transition/results";
+pub const MANUAL_TRANSITION_JOB_RECORD_PREFIX: &str = MANUAL_TRANSITION_JOB_NAMESPACE.prefix;
+pub const MANUAL_TRANSITION_SCOPE_RECORD_PREFIX: &str = MANUAL_TRANSITION_SCOPE_NAMESPACE.prefix;
+pub const MANUAL_TRANSITION_TASK_PREFIX: &str = MANUAL_TRANSITION_TASK_NAMESPACE.prefix;
+pub const MANUAL_TRANSITION_WORKER_RESULT_PREFIX: &str = MANUAL_TRANSITION_WORKER_RESULT_NAMESPACE.prefix;
 pub const MAX_MANUAL_TRANSITION_JOB_RECORD_SIZE: usize = 64 * 1024;
 pub const MAX_MANUAL_TRANSITION_TASK_RECORD_SIZE: usize = 16 * 1024;
 pub const MAX_MANUAL_TRANSITION_WORKER_RESULT_RECORD_SIZE: usize = 8 * 1024;
@@ -195,6 +199,8 @@ pub struct ManualTransitionJobRecord {
     pub updated_at_unix_nanos: i128,
     #[serde(default, skip_serializing_if = "Option::is_none")]
     pub completed_at_unix_nanos: Option<i128>,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub cursor_revision: Option<u64>,
     pub report: ManualTransitionRunReport,
     pub queue_snapshot: ManualTransitionQueueSnapshot,
     #[serde(default, skip_serializing_if = "Option::is_none")]
@@ -224,6 +230,7 @@ impl ManualTransitionJobRecord {
             created_at_unix_nanos: now,
             updated_at_unix_nanos: now,
             completed_at_unix_nanos: None,
+            cursor_revision: Some(0),
             report: ManualTransitionRunReport {
                 bucket: bucket.to_string(),
                 prefix: options.prefix.clone(),
@@ -238,7 +245,7 @@ impl ManualTransitionJobRecord {
 
     pub fn complete(&mut self, report: ManualTransitionRunReport, queue_snapshot: ManualTransitionQueueSnapshot) {
         self.scan_completed = true;
-        self.report.merge_scan_report_preserving_worker(&report);
+        self.merge_scan_report(&report);
         self.queue_snapshot = queue_snapshot;
         self.error = None;
         self.mark_terminal_if_worker_drained();
@@ -316,7 +323,7 @@ impl ManualTransitionJobRecord {
             }
         }
         self.queue_snapshot = queue_snapshot;
-        self.updated_at_unix_nanos = OffsetDateTime::now_utc().unix_timestamp_nanos();
+        self.advance_updated_at();
         self.mark_terminal_if_worker_drained();
     }
 
@@ -359,14 +366,14 @@ impl ManualTransitionJobRecord {
         self.report.tier_failure = scan_tier_failure.saturating_add(transition_failed);
         self.report.tier_failure_by_reason = scan_tier_failure_by_reason;
         self.queue_snapshot = queue_snapshot;
-        self.updated_at_unix_nanos = OffsetDateTime::now_utc().unix_timestamp_nanos();
+        self.advance_updated_at();
         self.mark_terminal_if_worker_drained();
         true
     }
 
     pub fn mark_cancel_requested(&mut self) {
         self.cancel_requested = true;
-        self.updated_at_unix_nanos = OffsetDateTime::now_utc().unix_timestamp_nanos();
+        self.advance_updated_at();
     }
 
     pub fn claim_recovery_lease(&mut self, owner_id: impl Into<String>, queue_snapshot: ManualTransitionQueueSnapshot) {
@@ -380,7 +387,7 @@ impl ManualTransitionJobRecord {
     pub fn abandon_recovery_lease(&mut self, lease_id: Uuid) {
         if self.state == ManualTransitionJobState::Running && self.lease_id == lease_id {
             self.lease_expires_at_unix_nanos = 0;
-            self.updated_at_unix_nanos = OffsetDateTime::now_utc().unix_timestamp_nanos();
+            self.advance_updated_at();
         }
     }
 
@@ -398,7 +405,7 @@ impl ManualTransitionJobRecord {
 
     pub fn renew_lease(&mut self, queue_snapshot: ManualTransitionQueueSnapshot) {
         let now = OffsetDateTime::now_utc().unix_timestamp_nanos();
-        self.updated_at_unix_nanos = now;
+        self.updated_at_unix_nanos = self.updated_at_unix_nanos.saturating_add(1).max(now);
         self.lease_expires_at_unix_nanos = manual_transition_job_lease_expires_at(now);
         self.queue_snapshot = queue_snapshot;
     }
@@ -438,11 +445,18 @@ impl ManualTransitionJobRecord {
 
     pub fn update_running_progress(&mut self, report: ManualTransitionRunReport, queue_snapshot: ManualTransitionQueueSnapshot) {
         if self.state == ManualTransitionJobState::Running {
-            self.report.merge_scan_report_preserving_worker(&report);
+            self.merge_scan_report(&report);
             self.renew_lease(queue_snapshot);
         }
     }
 
+    fn merge_scan_report(&mut self, report: &ManualTransitionRunReport) {
+        if self.report.continuation_token != report.continuation_token {
+            self.cursor_revision = Some(self.cursor_revision.unwrap_or(0).saturating_add(1));
+        }
+        self.report.merge_scan_report_preserving_worker(report);
+    }
+
     pub fn mark_unknown_if_unowned(&mut self) {
         if self.state == ManualTransitionJobState::Running {
             self.state = ManualTransitionJobState::Unknown;
@@ -463,9 +477,13 @@ impl ManualTransitionJobRecord {
     }
 
     fn mark_updated_terminal(&mut self) {
+        self.advance_updated_at();
+        self.completed_at_unix_nanos = Some(self.updated_at_unix_nanos);
+    }
+
+    fn advance_updated_at(&mut self) {
         let now = OffsetDateTime::now_utc().unix_timestamp_nanos();
-        self.updated_at_unix_nanos = now;
-        self.completed_at_unix_nanos = Some(now);
+        self.updated_at_unix_nanos = self.updated_at_unix_nanos.saturating_add(1).max(now);
     }
 
     fn mark_terminal_if_worker_drained(&mut self) {
@@ -1109,7 +1127,8 @@ pub fn manual_transition_scope_record_object_name(scope_key: &str) -> Result<Str
 pub async fn save_manual_transition_job_record(api: Arc<ECStore>, job: &ManualTransitionJobRecord) -> EcstoreResult<()> {
     let object = manual_transition_job_record_object_name(job.job_id).map_err(manual_transition_job_store_error)?;
     let data = job.encode().map_err(manual_transition_job_store_error)?;
-    config_boundary::save_config(api, &object, data).await
+    config_boundary::save_config(api.clone(), &object, data.clone()).await?;
+    api.record_durable_ilm_decommission_progress(&object, &data).await
 }
 
 pub async fn load_manual_transition_job_record(api: Arc<ECStore>, job_id: Uuid) -> EcstoreResult<ManualTransitionJobRecord> {
@@ -1142,9 +1161,9 @@ pub async fn save_manual_transition_job_record_if_current(
     let object = manual_transition_job_record_object_name(job.job_id).map_err(manual_transition_job_store_error)?;
     let data = job.encode().map_err(manual_transition_job_store_error)?;
     config_boundary::save_config_with_opts_quiet(
-        api,
+        api.clone(),
         &object,
-        data,
+        data.clone(),
         &ObjectOptions {
             max_parity: true,
             http_preconditions: Some(HTTPPreconditions {
@@ -1154,7 +1173,8 @@ pub async fn save_manual_transition_job_record_if_current(
             ..Default::default()
         },
     )
-    .await
+    .await?;
+    api.record_durable_ilm_decommission_progress(&object, &data).await
 }
 
 /// Applies a job-record mutation with optimistic concurrency control.
@@ -1592,9 +1612,9 @@ pub async fn save_manual_transition_scope_admission_if_absent(
     let object = manual_transition_scope_record_object_name(&admission.scope_key).map_err(manual_transition_job_store_error)?;
     let data = serde_json::to_vec(admission).map_err(Error::other)?;
     config_boundary::save_config_with_opts(
-        api,
+        api.clone(),
         &object,
-        data,
+        data.clone(),
         &ObjectOptions {
             max_parity: true,
             http_preconditions: Some(HTTPPreconditions {
@@ -1604,7 +1624,8 @@ pub async fn save_manual_transition_scope_admission_if_absent(
             ..Default::default()
         },
     )
-    .await
+    .await?;
+    api.record_durable_ilm_decommission_progress(&object, &data).await
 }
 
 pub async fn load_manual_transition_scope_admission(
@@ -1642,9 +1663,9 @@ pub async fn save_manual_transition_scope_admission_if_current(
     let object = manual_transition_scope_record_object_name(&admission.scope_key).map_err(manual_transition_job_store_error)?;
     let data = serde_json::to_vec(admission).map_err(Error::other)?;
     match config_boundary::save_config_with_opts(
-        api,
+        api.clone(),
         &object,
-        data,
+        data.clone(),
         &ObjectOptions {
             max_parity: true,
             http_preconditions: Some(HTTPPreconditions {
@@ -1660,7 +1681,8 @@ pub async fn save_manual_transition_scope_admission_if_current(
             Err(Error::PreconditionFailed)
         }
         result => result,
-    }
+    }?;
+    api.record_durable_ilm_decommission_progress(&object, &data).await
 }
 
 pub async fn claim_manual_transition_scope_admission(
@@ -1953,13 +1975,15 @@ pub async fn delete_manual_transition_scope_admission_if_current(
     job_id: Uuid,
     lease_id: Uuid,
 ) -> EcstoreResult<bool> {
-    let etag = match load_manual_transition_scope_admission_with_etag(api.clone(), scope_key).await {
-        Ok((admission, etag)) if admission.job_id == job_id && admission.lease_id == lease_id => etag,
+    let (admission, etag) = match load_manual_transition_scope_admission_with_etag(api.clone(), scope_key).await {
+        Ok((admission, etag)) if admission.job_id == job_id && admission.lease_id == lease_id => (admission, etag),
         Ok(_) => return Ok(false),
         Err(Error::ConfigNotFound) => return Ok(true),
         Err(err) => return Err(err),
     };
     let object = manual_transition_scope_record_object_name(scope_key).map_err(manual_transition_job_store_error)?;
+    let data = serde_json::to_vec(&admission).map_err(Error::other)?;
+    api.record_durable_ilm_decommission_terminal(&object, &data).await?;
     match config_boundary::delete_config_if_match(api, &object, &etag).await {
         Ok(()) | Err(Error::ConfigNotFound) => Ok(true),
         Err(Error::PreconditionFailed) => Ok(false),
@@ -2577,6 +2601,25 @@ mod tests {
         assert!(decoded.report.tier_failure_by_reason.is_empty());
     }
 
+    #[test]
+    fn manual_transition_job_record_decodes_legacy_cursor_without_revision() {
+        let options = ManualTransitionRunOptions::default();
+        let record = ManualTransitionJobRecord::new(Uuid::new_v4(), "bucket", &options, TEST_OWNER);
+        let encoded = record.encode().expect("job record should encode");
+        let mut value: serde_json::Value = serde_json::from_slice(&encoded).expect("encoded job should be json");
+        value["job"]
+            .as_object_mut()
+            .expect("job should be object")
+            .remove("cursor_revision");
+        let record_bytes = serde_json::to_vec(&value["job"]).expect("legacy job should encode");
+        value["content_sha256"] = serde_json::Value::String(hex_sha256(&record_bytes, ToOwned::to_owned));
+        let legacy = serde_json::to_vec(&value).expect("legacy envelope should encode");
+
+        let decoded = ManualTransitionJobRecord::decode(record.job_id, &legacy).expect("legacy job should decode");
+
+        assert_eq!(decoded.cursor_revision, None);
+    }
+
     #[test]
     fn manual_transition_job_record_rejects_unknown_report_fields() {
         let options = ManualTransitionRunOptions::default();

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

@@ -16,6 +16,7 @@ pub mod bucket_lifecycle_audit;
 pub mod bucket_lifecycle_ops;
 mod config_boundary;
 pub mod core;
+mod durable_namespace;
 pub mod evaluator;
 pub mod manual_transition_job;
 mod metadata_boundary;
@@ -31,3 +32,8 @@ pub mod tier_free_version_recovery;
 pub mod tier_last_day_stats;
 pub mod tier_sweeper;
 pub mod transition_transaction;
+
+pub(crate) use durable_namespace::{
+    DurableIlmRecordCheckpoint, ILM_META_PREFIX, ValidatedDurableIlmRecord, classify_durable_ilm_record,
+    validate_durable_ilm_record,
+};

crates/ecstore/src/bucket/lifecycle/tier_delete_journal.rs

@@ -20,6 +20,7 @@ use tokio_util::sync::CancellationToken;
 use tracing::{debug, warn};
 
 use crate::bucket::lifecycle::config_boundary;
+use crate::bucket::lifecycle::durable_namespace::TIER_DELETE_JOURNAL_NAMESPACE;
 use crate::bucket::lifecycle::runtime_boundary;
 use crate::bucket::lifecycle::tier_sweeper::{
     Jentry, TierDeleteJournalState, TierDeleteSourceIdentity,
@@ -49,7 +50,7 @@ const TIER_DELETE_JOURNAL_VERSION: u8 = 2;
 const TIER_DELETE_JOURNAL_EXACT_VERSION: u8 = 3;
 const TIER_DELETE_JOURNAL_STATE_VERSION: u8 = 4;
 const TIER_DELETE_JOURNAL_TRANSACTION_VERSION: u8 = 5;
-pub(crate) const TIER_DELETE_JOURNAL_PREFIX: &str = "ilm/tier-delete-journal/";
+pub(crate) const TIER_DELETE_JOURNAL_PREFIX: &str = TIER_DELETE_JOURNAL_NAMESPACE.prefix;
 
 #[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
 #[serde(deny_unknown_fields)]
@@ -432,6 +433,21 @@ async fn process_committed_tier_delete_journal_entry(api: Arc<ECStore>, je: &Jen
         )
         .await?;
     }
+    let path = tier_delete_journal_object_name(je);
+    let data = encode_tier_delete_journal_entry(je).map_err(std::io::Error::other)?;
+    let target_pool_indices = api
+        .record_durable_ilm_decommission_terminal_target_pools(&path, &data)
+        .await
+        .map_err(std::io::Error::other)?;
+    if let Some(target_pool_indices) = target_pool_indices {
+        for target_pool_idx in target_pool_indices {
+            match config_boundary::delete_config(api.pools[target_pool_idx].clone(), &path).await {
+                Ok(()) | Err(Error::ConfigNotFound) => {}
+                Err(err) => return Err(std::io::Error::other(err)),
+            }
+        }
+        return Ok(());
+    }
     remove_tier_delete_journal_entry(api, je).await
 }
 

crates/ecstore/src/bucket/lifecycle/transition_transaction.rs

@@ -21,6 +21,7 @@ use tracing::{debug, warn};
 use uuid::Uuid;
 
 use crate::bucket::lifecycle::config_boundary;
+use crate::bucket::lifecycle::durable_namespace::TRANSITION_TRANSACTION_NAMESPACE;
 use crate::bucket::lifecycle::lifecycle::TRANSITION_COMPLETE;
 use crate::bucket::lifecycle::tier_sweeper::{
     delete_confirmed_transition_candidate_exact_with_lease_idempotent,
@@ -42,7 +43,7 @@ const TRANSITION_TRANSACTION_RECOVERY_INTERVAL: Duration = Duration::from_secs(6
 const TRANSITION_TRANSACTION_RECOVERY_TIMEOUT: Duration = Duration::from_secs(300);
 pub const TRANSITION_TRANSACTION_SCHEMA: &str = "rustfs-transition-transaction-v1";
 pub const TRANSITION_TRANSACTION_PREFIX: &str = "ilm/transition-transactions";
-pub const TRANSITION_TRANSACTION_RECORD_PREFIX: &str = "ilm/transition-transactions/records";
+pub const TRANSITION_TRANSACTION_RECORD_PREFIX: &str = TRANSITION_TRANSACTION_NAMESPACE.prefix;
 pub const MAX_TRANSITION_TRANSACTION_SIZE: usize = 64 * 1024;
 
 pub type Result<T> = std::result::Result<T, TransitionTransactionError>;
@@ -584,7 +585,8 @@ pub(crate) async fn save_transition_transaction_record(
     let object =
         transition_transaction_record_object_name(transaction.transaction_id).map_err(transition_transaction_store_error)?;
     let data = transaction.encode().map_err(transition_transaction_store_error)?;
-    config_boundary::save_config(api, &object, data).await
+    config_boundary::save_config(api.clone(), &object, data.clone()).await?;
+    api.record_durable_ilm_decommission_progress(&object, &data).await
 }
 
 pub(crate) async fn load_transition_transaction_record(
@@ -596,8 +598,14 @@ pub(crate) async fn load_transition_transaction_record(
     TransitionTransaction::decode(transaction_id, &data).map_err(transition_transaction_store_error)
 }
 
-pub(crate) async fn delete_transition_transaction_record(api: Arc<ECStore>, transaction_id: Uuid) -> EcstoreResult<()> {
-    let object = transition_transaction_record_object_name(transaction_id).map_err(transition_transaction_store_error)?;
+pub(crate) async fn delete_transition_transaction_record(
+    api: Arc<ECStore>,
+    transaction: &TransitionTransaction,
+) -> EcstoreResult<()> {
+    let object =
+        transition_transaction_record_object_name(transaction.transaction_id).map_err(transition_transaction_store_error)?;
+    let data = transaction.encode().map_err(transition_transaction_store_error)?;
+    api.record_durable_ilm_decommission_terminal(&object, &data).await?;
     match config_boundary::delete_config(api, &object).await {
         Ok(()) | Err(Error::ConfigNotFound) => Ok(()),
         Err(err) => Err(err),
@@ -813,7 +821,7 @@ pub async fn finalize_missing_transition_transaction_for_operator(
     if probe != TransitionOperatorProbe::Missing {
         return Err(TransitionOperatorError::CandidateNotMissing(probe));
     }
-    delete_transition_transaction_record(api, transaction_id)
+    delete_transition_transaction_record(api, &transaction)
         .await
         .map_err(TransitionOperatorError::Store)
 }
@@ -849,22 +857,22 @@ pub async fn process_transition_transaction_record(
     match transaction.state {
         TransitionTransactionState::Uploaded => {
             delete_transition_remote_candidate(api.clone(), transaction).await?;
-            delete_transition_transaction_record(api, transaction.transaction_id).await?;
+            delete_transition_transaction_record(api, transaction).await?;
             Ok(TransitionTransactionRecoveryOutcome::RemoteCandidateDeleted)
         }
         TransitionTransactionState::CleanupPending => match local_commit_matches_transaction(api.clone(), transaction).await {
             Ok(true) => {
-                delete_transition_transaction_record(api, transaction.transaction_id).await?;
+                delete_transition_transaction_record(api, transaction).await?;
                 Ok(TransitionTransactionRecoveryOutcome::RecordDeleted)
             }
             Ok(false) => {
                 delete_transition_remote_candidate(api.clone(), transaction).await?;
-                delete_transition_transaction_record(api, transaction.transaction_id).await?;
+                delete_transition_transaction_record(api, transaction).await?;
                 Ok(TransitionTransactionRecoveryOutcome::RemoteCandidateDeleted)
             }
             Err(err) if transition_source_is_missing(&err) => {
                 delete_transition_remote_candidate(api.clone(), transaction).await?;
-                delete_transition_transaction_record(api, transaction.transaction_id).await?;
+                delete_transition_transaction_record(api, transaction).await?;
                 Ok(TransitionTransactionRecoveryOutcome::RemoteCandidateDeleted)
             }
             Err(err) => Err(err),
@@ -872,7 +880,7 @@ pub async fn process_transition_transaction_record(
         TransitionTransactionState::LocalCommitStarted => {
             match local_commit_matches_transaction(api.clone(), transaction).await {
                 Ok(true) => {
-                    delete_transition_transaction_record(api, transaction.transaction_id).await?;
+                    delete_transition_transaction_record(api, transaction).await?;
                     Ok(TransitionTransactionRecoveryOutcome::RecordDeleted)
                 }
                 Ok(false) => Ok(TransitionTransactionRecoveryOutcome::Retained),
@@ -881,7 +889,7 @@ pub async fn process_transition_transaction_record(
             }
         }
         TransitionTransactionState::AbortedNoRemote | TransitionTransactionState::Committed => {
-            delete_transition_transaction_record(api, transaction.transaction_id).await?;
+            delete_transition_transaction_record(api, transaction).await?;
             Ok(TransitionTransactionRecoveryOutcome::RecordDeleted)
         }
         TransitionTransactionState::UploadOutcomeUnknown => recover_unknown_upload_outcome(api, transaction).await,
@@ -907,7 +915,7 @@ async fn recover_unknown_upload_outcome(
         .map_err(Error::other)?
     {
         TransitionCandidateProbe::Missing => {
-            delete_transition_transaction_record(api, transaction.transaction_id).await?;
+            delete_transition_transaction_record(api, transaction).await?;
             Ok(TransitionTransactionRecoveryOutcome::RecordDeleted)
         }
         TransitionCandidateProbe::UnversionedPresent => {
@@ -925,7 +933,7 @@ async fn recover_unknown_upload_outcome(
             )
             .await
             .map_err(Error::other)?;
-            delete_transition_transaction_record(api, transaction.transaction_id).await?;
+            delete_transition_transaction_record(api, transaction).await?;
             Ok(TransitionTransactionRecoveryOutcome::RemoteCandidateDeleted)
         }
         TransitionCandidateProbe::VersionedPresent(version_id) => {
@@ -958,7 +966,7 @@ async fn cleanup_recovered_unknown_upload_candidate(
         .map_err(transition_transaction_store_error)?;
     save_transition_transaction_record(api.clone(), &cleanup).await?;
     delete_transition_remote_candidate(api.clone(), &cleanup).await?;
-    delete_transition_transaction_record(api, cleanup.transaction_id).await?;
+    delete_transition_transaction_record(api, &cleanup).await?;
     Ok(TransitionTransactionRecoveryOutcome::RemoteCandidateDeleted)
 }
 

crates/ecstore/src/config/com.rs

@@ -406,6 +406,25 @@ where
     Ok(data)
 }
 
+pub(crate) async fn read_config_limited_preserve_empty<S>(api: Arc<S>, file: &str, max_bytes: usize) -> Result<Vec<u8>>
+where
+    S: EcstoreObjectIO,
+{
+    let (data, _obj) = read_config_limited_preserve_empty_with_metadata(api, file, max_bytes).await?;
+    Ok(data)
+}
+
+pub(crate) async fn read_config_limited_preserve_empty_with_metadata<S>(
+    api: Arc<S>,
+    file: &str,
+    max_bytes: usize,
+) -> Result<(Vec<u8>, ObjectInfo)>
+where
+    S: EcstoreObjectIO,
+{
+    read_config_with_metadata_inner(api, file, &ObjectOptions::default(), true, Some(max_bytes)).await
+}
+
 /// Read an existing config object without treating an empty payload as absent.
 /// Callers that validate their own payload format need to distinguish corruption
 /// from `ConfigNotFound`.

crates/ecstore/src/core/sets.rs

@@ -988,11 +988,14 @@ impl crate::storage_api_contracts::multipart::MultipartOperations for Sets {
     }
 }
 
-impl Sets {
-    pub(crate) async fn heal_format_with_fence<F>(&self, dry_run: bool, fence_lost: F) -> Result<(HealResultItem, Option<Error>)>
-    where
-        F: Fn() -> bool + Send + Sync,
-    {
+#[async_trait::async_trait]
+impl crate::storage_api_contracts::heal::HealOperations for Sets {
+    type Error = Error;
+    type HealResultItem = HealResultItem;
+    type HealOptions = HealOpts;
+
+    #[tracing::instrument(skip(self))]
+    async fn heal_format(&self, dry_run: bool) -> Result<(HealResultItem, Option<Error>)> {
         let (disks, init_errs) = init_storage_disks_with_errors(
             &self.endpoints.endpoints,
             &DiskOption {
@@ -1065,9 +1068,6 @@ impl Sets {
             // Save new formats `format.json` on unformatted disks.
             for (index, (fm, disk)) in tmp_new_formats.iter_mut().zip(disks.iter()).enumerate() {
                 if fm.is_some() && disk.is_some() {
-                    if fence_lost() {
-                        return Ok((res, Some(StorageError::SlowDown)));
-                    }
                     if let Err(err) = save_format_file(disk, fm).await {
                         if let Some(disk) = disk.as_ref() {
                             let _ = disk.close().await;
@@ -1101,18 +1101,6 @@ impl Sets {
         }
         Ok((res, None))
     }
-}
-
-#[async_trait::async_trait]
-impl crate::storage_api_contracts::heal::HealOperations for Sets {
-    type Error = Error;
-    type HealResultItem = HealResultItem;
-    type HealOptions = HealOpts;
-
-    #[tracing::instrument(skip(self))]
-    async fn heal_format(&self, dry_run: bool) -> Result<(HealResultItem, Option<Error>)> {
-        self.heal_format_with_fence(dry_run, || false).await
-    }
     #[tracing::instrument(skip(self))]
     async fn heal_bucket(&self, bucket: &str, opts: &HealOpts) -> Result<HealResultItem> {
         let mut result = HealResultItem {

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

@@ -37,7 +37,7 @@ use crate::bucket::lifecycle::{
     transition_transaction::{
         TransitionRemoteVersion, TransitionSourceIdentity, TransitionSourceVersionMode, TransitionTransaction,
         TransitionTransactionInit, TransitionTransactionState, delete_transition_transaction_record,
-        save_transition_transaction_record,
+        load_transition_transaction_record, save_transition_transaction_record,
     },
 };
 use crate::bucket::quota::reservation;
@@ -4261,7 +4261,12 @@ fn record_transition_uploaded_save_attempt(transaction: &TransitionTransaction,
 
 async fn delete_transition_transaction_if_available(api: Option<&Arc<ECStore>>, transaction_id: Uuid) -> Result<()> {
     if let Some(api) = api {
-        return delete_transition_transaction_record(api.clone(), transaction_id).await;
+        let transaction = match load_transition_transaction_record(api.clone(), transaction_id).await {
+            Ok(transaction) => transaction,
+            Err(Error::ConfigNotFound) => return Ok(()),
+            Err(err) => return Err(err),
+        };
+        return delete_transition_transaction_record(api.clone(), &transaction).await;
     }
     Ok(())
 }

crates/ecstore/src/store/heal.rs

@@ -13,12 +13,7 @@
 // limitations under the License.
 
 use super::*;
-use crate::core::pools::POOL_META_NAME;
-use crate::services::rebalance::{REBAL_META_NAME, RebalStatus};
-use crate::set_disk::get_lock_acquire_timeout;
 use crate::storage_api_contracts::heal::HealOperations as _;
-use crate::storage_api_contracts::namespace::NamespaceLocking as _;
-use rustfs_lock::NamespaceLockGuard;
 use tracing::trace;
 
 const LOG_COMPONENT_ECSTORE: &str = "ecstore";
@@ -35,119 +30,7 @@ fn invalid_heal_pool_index(pool_idx: usize, pool_count: usize) -> Error {
     )
 }
 
-#[derive(Debug, Clone, Copy)]
-enum HealFormatPoolSkip {
-    Completed,
-    Retryable,
-}
-
-fn classify_heal_format_pool(
-    pool_idx: usize,
-    pool_cmd_line: &str,
-    pool_meta: &PoolMeta,
-    rebalance_meta: Option<&RebalanceMeta>,
-) -> Option<HealFormatPoolSkip> {
-    let Some(pool) = pool_meta.pools.get(pool_idx) else {
-        return Some(HealFormatPoolSkip::Retryable);
-    };
-
-    if pool.id != pool_idx || pool_cmd_line.is_empty() || pool.cmd_line.is_empty() || pool.cmd_line != pool_cmd_line {
-        return Some(HealFormatPoolSkip::Retryable);
-    }
-
-    if let Some(decommission) = pool.decommission.as_ref() {
-        if decommission.complete {
-            return Some(HealFormatPoolSkip::Completed);
-        }
-        if decommission.failed || decommission.canceled || decommission.queued || pool_meta.is_suspended(pool_idx) {
-            return Some(HealFormatPoolSkip::Retryable);
-        }
-    }
-
-    if let Some(meta) = rebalance_meta {
-        let Some(pool_stats) = meta.pool_stats.get(pool_idx) else {
-            return Some(HealFormatPoolSkip::Retryable);
-        };
-        if pool_stats.info.stopping || (pool_stats.participating && pool_stats.info.status == RebalStatus::Started) {
-            return Some(HealFormatPoolSkip::Retryable);
-        }
-    }
-
-    None
-}
-
-fn heal_format_pool_skip_error(skip: HealFormatPoolSkip) -> Error {
-    match skip {
-        HealFormatPoolSkip::Completed => StorageError::NoHealRequired,
-        HealFormatPoolSkip::Retryable => StorageError::SlowDown,
-    }
-}
-
-fn heal_format_fence_lost_error() -> Error {
-    StorageError::SlowDown
-}
-
 impl ECStore {
-    async fn acquire_heal_format_fence(
-        &self,
-    ) -> Result<(NamespaceLockGuard, NamespaceLockGuard, PoolMeta, Option<RebalanceMeta>)> {
-        let metadata_pool = self
-            .pools
-            .first()
-            .cloned()
-            .ok_or_else(|| Error::other("heal format requires at least one storage pool"))?;
-
-        // Metadata fence order is part of the decommission/rebalance protocol:
-        // pool.bin must always be acquired before rebalance.bin.
-        let pool_lock = metadata_pool.new_ns_lock(RUSTFS_META_BUCKET, POOL_META_NAME).await?;
-        let pool_guard = pool_lock.get_write_lock(get_lock_acquire_timeout()).await?;
-        let rebalance_lock = metadata_pool.new_ns_lock(RUSTFS_META_BUCKET, REBAL_META_NAME).await?;
-        let rebalance_guard = rebalance_lock.get_write_lock(get_lock_acquire_timeout()).await?;
-
-        if pool_guard.is_lock_lost() || rebalance_guard.is_lock_lost() {
-            return Err(heal_format_fence_lost_error());
-        }
-
-        let mut pool_meta = PoolMeta::default();
-        pool_meta.load_no_lock(metadata_pool.clone()).await?;
-        if pool_meta.pools.len() != self.pools.len()
-            || pool_meta.pools.iter().enumerate().any(|(pool_idx, pool)| {
-                pool.id != pool_idx || pool.cmd_line.is_empty() || pool.cmd_line != self.pools[pool_idx].endpoints.cmd_line
-            })
-        {
-            return Err(heal_format_fence_lost_error());
-        }
-
-        let mut rebalance_meta = RebalanceMeta::new();
-        let rebalance_meta = match rebalance_meta
-            .load_with_opts(
-                metadata_pool,
-                ObjectOptions {
-                    no_lock: true,
-                    ..Default::default()
-                },
-            )
-            .await
-        {
-            Ok(()) => Some(rebalance_meta),
-            Err(Error::ConfigNotFound) => None,
-            Err(err) => return Err(err),
-        };
-
-        if rebalance_meta
-            .as_ref()
-            .is_some_and(|meta| meta.pool_stats.len() != self.pools.len())
-        {
-            return Err(heal_format_fence_lost_error());
-        }
-
-        if pool_guard.is_lock_lost() || rebalance_guard.is_lock_lost() {
-            return Err(heal_format_fence_lost_error());
-        }
-
-        Ok((pool_guard, rebalance_guard, pool_meta, rebalance_meta))
-    }
-
     fn get_pools_for_heal_object(&self, opts: &HealOpts) -> Result<Vec<Arc<Sets>>> {
         match opts.pool {
             Some(pool_idx) => Ok(vec![
@@ -169,26 +52,9 @@ impl ECStore {
         };
 
         let mut count_no_heal = 0;
-        let mut count_completed = 0;
         let mut first_error = None;
-        for (pool_idx, pool) in self.pools.iter().enumerate() {
-            let (pool_guard, rebalance_guard, pool_meta, rebalance_meta) = self.acquire_heal_format_fence().await?;
-            if pool_guard.is_lock_lost() || rebalance_guard.is_lock_lost() {
-                first_error.get_or_insert(heal_format_fence_lost_error());
-                break;
-            }
-            if let Some(skip) = classify_heal_format_pool(pool_idx, &pool.endpoints.cmd_line, &pool_meta, rebalance_meta.as_ref())
-            {
-                if matches!(skip, HealFormatPoolSkip::Completed) {
-                    count_completed += 1;
-                } else {
-                    first_error.get_or_insert(heal_format_pool_skip_error(skip));
-                }
-                continue;
-            }
-
-            let fence_lost = || pool_guard.is_lock_lost() || rebalance_guard.is_lock_lost();
-            let (mut result, err) = pool.heal_format_with_fence(dry_run, fence_lost).await?;
+        for pool in self.pools.iter() {
+            let (mut result, err) = pool.heal_format(dry_run).await?;
             if let Some(err) = err {
                 match err {
                     StorageError::NoHealRequired => {
@@ -203,18 +69,11 @@ impl ECStore {
             r.set_count += result.set_count;
             r.before.drives.append(&mut result.before.drives);
             r.after.drives.append(&mut result.after.drives);
-
-            // A lease can be lost after the final write; fail closed before
-            // reporting the pool as successfully healed.
-            if pool_guard.is_lock_lost() || rebalance_guard.is_lock_lost() {
-                first_error.get_or_insert(heal_format_fence_lost_error());
-                break;
-            }
         }
         if let Some(err) = first_error {
             return Ok((r, Some(err)));
         }
-        if count_no_heal + count_completed == self.pools.len() {
+        if count_no_heal == self.pools.len() {
             info!(
                 event = EVENT_HEAL_FORMAT_COMPLETED,
                 component = LOG_COMPONENT_ECSTORE,
@@ -443,7 +302,6 @@ mod tests {
     use crate::disk::{DeleteOptions, DiskOption, format::FormatV3, new_disk};
     use crate::layout::endpoints::{EndpointServerPools, Endpoints, PoolEndpoints};
     use crate::runtime::instance::InstanceContext;
-    use crate::services::rebalance::{RebalanceInfo, RebalanceStats};
     use crate::storage_api_contracts::bucket::{BucketOperations, MakeBucketOptions};
     use crate::storage_api_contracts::object::{ObjectIO as _, ObjectOperations};
     use crate::store::init_format::{load_format_erasure, save_format_file};
@@ -495,164 +353,6 @@ mod tests {
         }
     }
 
-    fn pool_meta_with_decommission(info: PoolDecommissionInfo) -> PoolMeta {
-        PoolMeta {
-            pools: vec![PoolStatus {
-                id: 0,
-                cmd_line: "pool-0".to_string(),
-                last_update: OffsetDateTime::UNIX_EPOCH,
-                decommission: Some(info),
-            }],
-            ..Default::default()
-        }
-    }
-
-    #[test]
-    fn heal_format_pool_state_barriers_are_classified() {
-        let active = pool_meta_with_decommission(PoolDecommissionInfo {
-            start_time: Some(OffsetDateTime::UNIX_EPOCH),
-            ..Default::default()
-        });
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &active, None),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        for info in [
-            PoolDecommissionInfo {
-                failed: true,
-                ..Default::default()
-            },
-            PoolDecommissionInfo {
-                canceled: true,
-                ..Default::default()
-            },
-        ] {
-            assert!(matches!(
-                classify_heal_format_pool(0, "pool-0", &pool_meta_with_decommission(info), None),
-                Some(HealFormatPoolSkip::Retryable)
-            ));
-        }
-
-        let completed = pool_meta_with_decommission(PoolDecommissionInfo {
-            complete: true,
-            ..Default::default()
-        });
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &completed, None),
-            Some(HealFormatPoolSkip::Completed)
-        ));
-    }
-
-    #[test]
-    fn heal_format_pool_rebalance_barriers_and_identity_are_fail_closed() {
-        let identity_meta = pool_meta_with_decommission(PoolDecommissionInfo::default());
-        let rebalance = RebalanceMeta {
-            pool_stats: vec![RebalanceStats {
-                participating: true,
-                info: RebalanceInfo {
-                    status: RebalStatus::Started,
-                    ..Default::default()
-                },
-                ..Default::default()
-            }],
-            ..Default::default()
-        };
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &identity_meta, Some(&rebalance)),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        let stopping = RebalanceMeta {
-            pool_stats: vec![RebalanceStats {
-                info: RebalanceInfo {
-                    stopping: true,
-                    ..Default::default()
-                },
-                ..Default::default()
-            }],
-            ..Default::default()
-        };
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &identity_meta, Some(&stopping)),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        let identity = pool_meta_with_decommission(PoolDecommissionInfo::default());
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-new", &identity, None),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        let identity_without_decommission = PoolMeta {
-            pools: vec![PoolStatus {
-                id: 0,
-                cmd_line: "pool-0".to_string(),
-                last_update: OffsetDateTime::UNIX_EPOCH,
-                decommission: None,
-            }],
-            ..Default::default()
-        };
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-new", &identity_without_decommission, None),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        assert!(matches!(
-            classify_heal_format_pool(0, "", &identity_meta, None),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &PoolMeta::default(), None),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-
-        let stopped = RebalanceMeta {
-            stopped_at: Some(OffsetDateTime::UNIX_EPOCH),
-            pool_stats: vec![RebalanceStats {
-                participating: true,
-                info: RebalanceInfo {
-                    status: RebalStatus::Stopped,
-                    ..Default::default()
-                },
-                ..Default::default()
-            }],
-            ..Default::default()
-        };
-        assert!(classify_heal_format_pool(0, "pool-0", &identity_meta, Some(&stopped)).is_none());
-
-        let stopping_after_stop = RebalanceMeta {
-            stopped_at: Some(OffsetDateTime::UNIX_EPOCH),
-            pool_stats: vec![RebalanceStats {
-                participating: true,
-                info: RebalanceInfo {
-                    status: RebalStatus::Started,
-                    stopping: true,
-                    ..Default::default()
-                },
-                ..Default::default()
-            }],
-            ..Default::default()
-        };
-        assert!(matches!(
-            classify_heal_format_pool(0, "pool-0", &identity_meta, Some(&stopping_after_stop)),
-            Some(HealFormatPoolSkip::Retryable)
-        ));
-    }
-
-    #[test]
-    fn skipped_heal_format_pool_is_never_reported_as_success() {
-        assert!(matches!(
-            heal_format_pool_skip_error(HealFormatPoolSkip::Retryable),
-            StorageError::SlowDown
-        ));
-        assert!(matches!(
-            heal_format_pool_skip_error(HealFormatPoolSkip::Completed),
-            StorageError::NoHealRequired
-        ));
-    }
-
     async fn multi_pool_heal_store() -> (tempfile::TempDir, Arc<ECStore>, CancellationToken) {
         let temp_dir = tempfile::tempdir().expect("multi-pool heal test directory should be created");
         let mut pool_endpoints = Vec::new();
@@ -1189,18 +889,6 @@ mod tests {
             bucket_fence_registry: std::sync::Arc::default(),
         };
 
-        let err = store
-            .handle_heal_format(false)
-            .await
-            .expect_err("missing pool metadata must fail closed before format writes");
-        assert!(matches!(err, StorageError::SlowDown));
-
-        let pool_meta = PoolMeta::new(&store.pools, &PoolMeta::default());
-        pool_meta
-            .save(store.pools.clone())
-            .await
-            .expect("pool metadata should be persisted before format heal");
-
         let (result, err) = store
             .handle_heal_format(false)
             .await
@@ -1214,22 +902,5 @@ mod tests {
             .await
             .expect("the later pool should be healed despite the first pool error");
         assert_eq!(healed.erasure.this, recoverable_format.erasure.sets[0][2]);
-
-        let mut completed_meta = PoolMeta::new(&store.pools, &PoolMeta::default());
-        for status in &mut completed_meta.pools {
-            status.decommission = Some(PoolDecommissionInfo {
-                complete: true,
-                ..Default::default()
-            });
-        }
-        completed_meta
-            .save(store.pools.clone())
-            .await
-            .expect("completed pool metadata should be persisted");
-        let (_, err) = store
-            .handle_heal_format(false)
-            .await
-            .expect("completed pools should be reported as a no-op");
-        assert!(matches!(err, Some(StorageError::NoHealRequired)));
     }
 }

crates/heal/src/heal/task/heal_erasure_set.rs

@@ -231,10 +231,6 @@ impl HealTask {
                             "Heal erasure set format repair skipped because no format heal was required"
                         );
                     } else {
-                        let error = e;
-                        if error.is_recoverable_heal() {
-                            return Err(error);
-                        }
                         error!(
                             target: "rustfs::heal::task",
                             event = EVENT_HEAL_ERASURE_SET_RESULT,
@@ -243,7 +239,7 @@ impl HealTask {
                             task_id = %self.id,
                             set_disk_id,
                             result = "format_failed",
-                            error = %error,
+                            error = %e,
                             "Heal erasure set failed"
                         );
                         {
@@ -251,7 +247,7 @@ impl HealTask {
                             progress.update_progress(4, 4, 0, 0);
                         }
                         return Err(Error::TaskExecutionFailed {
-                            message: format!("Failed to heal disk format for {set_disk_id}: {error}"),
+                            message: format!("Failed to heal disk format for {set_disk_id}: {e}"),
                         });
                     }
                 } else {
@@ -288,9 +284,6 @@ impl HealTask {
             Err(Error::TaskCancelled) => return Err(Error::TaskCancelled),
             Err(Error::TaskTimeout) => return Err(Error::TaskTimeout),
             Err(e) => {
-                if e.is_recoverable_heal() {
-                    return Err(e);
-                }
                 error!(
                     target: "rustfs::heal::task",
                     event = EVENT_HEAL_ERASURE_SET_RESULT,

crates/heal/src/heal/task/tests.rs

@@ -547,7 +547,6 @@ struct MockStorage {
     heal_object_outcome: Mutex<Option<MockHealObjectOutcome>>,
     heal_object_outcomes: Mutex<HashMap<String, VecDeque<MockHealObjectOutcome>>>,
     format_no_heal_required: Mutex<bool>,
-    format_error: Mutex<Option<Error>>,
     global_format_calls: Mutex<u32>,
     replacement_format_calls: Mutex<Vec<(usize, usize, Vec<String>)>>,
     replacement_targets_ready: Mutex<bool>,
@@ -868,9 +867,6 @@ impl HealStorageAPI for MockStorage {
 
     async fn heal_format(&self, _dry_run: bool) -> Result<(HealResultItem, Option<Error>)> {
         *self.global_format_calls.lock().unwrap() += 1;
-        if let Some(error) = self.format_error.lock().unwrap().take() {
-            return Err(error);
-        }
         let no_heal_required = *self.format_no_heal_required.lock().unwrap();
         if no_heal_required {
             Ok((HealResultItem::default(), Some(Error::Storage(EcstoreError::NoHealRequired))))
@@ -2056,30 +2052,6 @@ async fn test_erasure_set_heal_continues_after_format_no_heal_required() {
     );
 }
 
-#[tokio::test]
-async fn erasure_set_format_slowdown_is_propagated() {
-    let storage = Arc::new(MockStorage {
-        format_error: Mutex::new(Some(Error::Storage(EcstoreError::SlowDown))),
-        ..Default::default()
-    });
-    let request = HealRequest::new(
-        HealType::ErasureSet {
-            buckets: Vec::new(),
-            set_disk_id: "pool_0_set_0".to_string(),
-        },
-        HealOptions::default(),
-        HealPriority::Normal,
-    );
-    let task = HealTask::from_request(request, storage);
-
-    let error = task
-        .execute()
-        .await
-        .expect_err("format SlowDown must remain recoverable for the task manager");
-
-    assert!(matches!(error, Error::Storage(EcstoreError::SlowDown)));
-}
-
 #[tokio::test]
 async fn erasure_set_bucket_prepass_failure_stops_before_object_heal() {
     let temp = TempDir::new().expect("temporary directory should be created");

crates/test-utils/src/lib.rs

@@ -245,18 +245,6 @@ impl TestECStoreEnvBuilder {
             .await
             .expect("build test ECStore");
 
-        // The production bootstrap only persists pool.bin from the elected
-        // first cluster node.  Test stores intentionally have no cluster
-        // election, but heal-format still requires that durable fence before
-        // it can write any disk format.  Materialize the validated topology
-        // here so the shared fixture models a ready single-node store.
-        let mut pool_meta = ecstore.pool_meta.read().await.clone();
-        pool_meta.dont_save = false;
-        pool_meta
-            .save(ecstore.pools.clone())
-            .await
-            .expect("persist test pool metadata");
-
         if self.init_bucket_metadata {
             let buckets_list = ecstore
                 .list_bucket(&BucketOptions {

docs/testing/ci-gates.md

@@ -1,149 +0,0 @@
-# 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

@@ -322,7 +322,7 @@ thiserror = { workspace = true }
 tracing.workspace = true
 url = { workspace = true }
 urlencoding = { workspace = true }
-uuid = { workspace = true, features = ["v4", "v5", "fast-rng", "macro-diagnostics"] }
+uuid = { workspace = true, features = ["v4", "fast-rng", "macro-diagnostics"] }
 zip = { workspace = true }
 libc = { workspace = true }
 rand = { workspace = true, features = ["serde"] }
@@ -345,7 +345,7 @@ libsystemd.workspace = true
 libmimalloc-sys.workspace = true
 
 [dev-dependencies]
-uuid = { workspace = true, features = ["v4", "v5", "fast-rng", "macro-diagnostics"] }
+uuid = { workspace = true, features = ["v4", "fast-rng", "macro-diagnostics"] }
 serial_test = { workspace = true }
 tempfile = { workspace = true }
 aws-config = { workspace = true }

rustfs/src/admin/handlers/site_replication.rs

@@ -41,7 +41,7 @@ use crate::admin::storage_api::config::save_admin_config;
 use crate::admin::storage_api::contract::bucket::{
     BucketOperations, BucketOptions, DeleteBucketOptions, MakeBucketOptions, SRBucketDeleteOp,
 };
-use crate::admin::storage_api::error::{Error as StorageError, is_err_bucket_not_found};
+use crate::admin::storage_api::error::Error as StorageError;
 use crate::admin::storage_api::runtime::ECStore;
 use crate::admin::utils::{encode_compatible_admin_payload, read_compatible_admin_body};
 use crate::auth::constant_time_eq;
@@ -55,7 +55,6 @@ use crate::storage::storage_api::{
 use base64::Engine;
 use base64::engine::general_purpose::STANDARD as BASE64_STANDARD;
 use base64::engine::general_purpose::URL_SAFE_NO_PAD;
-use futures::StreamExt;
 use hmac::{Hmac, Mac};
 use http::header::{CONTENT_TYPE, HOST};
 use http::{HeaderMap, HeaderValue, Uri};
@@ -2097,18 +2096,6 @@ async fn remote_add_preflight_info(site: &PeerSite) -> S3Result<SiteReplicationA
             format!("invalid site replication metainfo from `{}`: {e}", site.endpoint),
         )
     })?;
-    if info.deployment_id.is_empty() {
-        // The peer will be tracked under a locally derived fallback ID
-        // (deployment_id_for_endpoint) instead of its real deployment ID.
-        warn!(
-            event = EVENT_ADMIN_SITE_REPLICATION_STATE,
-            component = LOG_COMPONENT_ADMIN,
-            subsystem = LOG_SUBSYSTEM_SITE_REPLICATION,
-            result = "peer_deployment_id_missing",
-            peer_endpoint = %site.endpoint,
-            "admin site replication state"
-        );
-    }
 
     let idp_body = send_peer_admin_get_request_with_client(
         &client,
@@ -2219,30 +2206,20 @@ fn site_replication_bootstrap_token(uri: &Uri) -> Option<String> {
     query_pairs(uri).get("bootstrapToken").cloned()
 }
 
-/// Query for a peer `make-with-versioning` bucket op. `versioningEnabled`
-/// always travels so the outbound query matches MinIO's site-replication
-/// make-bucket wire contract: MinIO's own create-bucket hook sends
-/// `versioningEnabled=true` on this op. RustFS's inbound handler
-/// force-enables versioning either way.
-fn make_with_versioning_bucket_op_path(bucket: &str, created_at: Option<&str>, lock_enabled: bool) -> String {
+fn bootstrap_bucket_make_op_path(bucket: &SRBucketInfo) -> String {
     let mut query = form_urlencoded::Serializer::new(String::new());
-    query.append_pair("bucket", bucket);
-    query.append_pair("operation", SITE_REPLICATION_BUCKET_OP_MAKE_WITH_VERSIONING);
-    query.append_pair("versioningEnabled", "true");
-    if let Some(created_at) = created_at {
-        query.append_pair("createdAt", created_at);
+    query.append_pair("bucket", &bucket.bucket);
+    query.append_pair("operation", "make-with-versioning");
+    if let Some(created_at) = bucket
+        .created_at
+        .and_then(|value| value.format(&time::format_description::well_known::Rfc3339).ok())
+    {
+        query.append_pair("createdAt", &created_at);
     }
-    if lock_enabled {
+    if bucket.object_lock_config.is_some() {
         query.append_pair("lockEnabled", "true");
     }
-    format!("{SITE_REPLICATION_PEER_BUCKET_OPS_PATH}?{}", query.finish())
-}
-
-fn bootstrap_bucket_make_op_path(bucket: &SRBucketInfo) -> String {
-    let created_at = bucket
-        .created_at
-        .and_then(|value| value.format(&time::format_description::well_known::Rfc3339).ok());
-    make_with_versioning_bucket_op_path(&bucket.bucket, created_at.as_deref(), bucket.object_lock_config.is_some())
+    format!("/rustfs/admin/v3/site-replication/peer/bucket-ops?{}", query.finish())
 }
 
 fn bootstrap_bucket_meta_item(bucket: &SRBucketInfo, item_type: &str, updated_at: Option<OffsetDateTime>) -> SRBucketMeta {
@@ -4269,7 +4246,16 @@ async fn broadcast_site_replication_make_bucket(
         .format(&time::format_description::well_known::Rfc3339)
         .unwrap_or_default();
 
-    let path = make_with_versioning_bucket_op_path(bucket, Some(&created_at), lock_enabled);
+    let path = {
+        let mut query = form_urlencoded::Serializer::new(String::new());
+        query.append_pair("bucket", bucket);
+        query.append_pair("operation", "make-with-versioning");
+        query.append_pair("createdAt", &created_at);
+        if lock_enabled {
+            query.append_pair("lockEnabled", "true");
+        }
+        format!("/rustfs/admin/v3/site-replication/peer/bucket-ops?{}", query.finish())
+    };
     let path = if let Some(token) = bootstrap_token {
         with_site_replication_bootstrap_token(&path, token)
     } else {
@@ -10220,25 +10206,13 @@ impl Operation for SiteReplicationStatusHandler {
     }
 }
 
-/// `POST /v3/site-replication/devnull` — peer link-check upload drain.
-/// MinIO streams multi-megabyte probe bodies here during site netperf link
-/// checks and expects an unbounded discard (its handler copies to io.Discard);
-/// buffering through the 1MB admin body cap turned any larger probe into a
-/// 400 and a false link failure. Stream and discard instead — no size cap.
-async fn drain_site_replication_devnull(mut input: Body) -> S3Result<()> {
-    while let Some(chunk) = input.next().await {
-        chunk.map_err(|e| s3_error!(InvalidRequest, "failed to read devnull stream: {}", e))?;
-    }
-    Ok(())
-}
-
 pub struct SiteReplicationDevNullHandler {}
 
 #[async_trait::async_trait]
 impl Operation for SiteReplicationDevNullHandler {
     async fn call(&self, req: S3Request<Body>, _params: Params<'_, '_>) -> S3Result<S3Response<(StatusCode, Body)>> {
         validate_site_replication_admin_request(&req, AdminAction::SiteReplicationOperationAction).await?;
-        drain_site_replication_devnull(req.input).await?;
+        let _ = read_plain_admin_body(req.input).await?;
         Ok(empty_response(StatusCode::NO_CONTENT))
     }
 }
@@ -10497,19 +10471,6 @@ impl Operation for SRPeerJoinHandler {
     }
 }
 
-/// Outcome of a peer-driven `purge-deleted-bucket` replay. A bucket that is
-/// already gone means the purge raced an earlier replay or a local delete —
-/// that is success — but any other failure must reach the sender like the
-/// sibling delete branches do: swallowing it answered 200 while the bucket
-/// survived on this site.
-fn purge_deleted_bucket_result(result: Result<(), StorageError>) -> S3Result<()> {
-    match result {
-        Ok(()) => Ok(()),
-        Err(err) if is_err_bucket_not_found(&err) => Ok(()),
-        Err(err) => Err(ApiError::from(err).into()),
-    }
-}
-
 pub struct SRPeerBucketOpsHandler {}
 
 #[async_trait::async_trait]
@@ -10609,18 +10570,16 @@ impl Operation for SRPeerBucketOpsHandler {
                     .map_err(ApiError::from)?;
             }
             "purge-deleted-bucket" => {
-                purge_deleted_bucket_result(
-                    store
-                        .delete_bucket(
-                            &bucket,
-                            &DeleteBucketOptions {
-                                force: true,
-                                srdelete_op: SRBucketDeleteOp::Purge,
-                                ..Default::default()
-                            },
-                        )
-                        .await,
-                )?;
+                let _ = store
+                    .delete_bucket(
+                        &bucket,
+                        &DeleteBucketOptions {
+                            force: true,
+                            srdelete_op: SRBucketDeleteOp::Purge,
+                            ..Default::default()
+                        },
+                    )
+                    .await;
             }
             _ => return Err(s3_error!(InvalidRequest, "unsupported site replication bucket operation")),
         }
@@ -13966,54 +13925,6 @@ mod tests {
         assert!(!query_flag(&uri, "missing"));
     }
 
-    /// A5 red-light: a `purge-deleted-bucket` replay must report success when
-    /// the bucket is already gone, and must propagate every other failure —
-    /// the swallowed error answered 200 while the bucket survived.
-    #[test]
-    fn test_purge_deleted_bucket_result_tolerates_only_missing_bucket() {
-        assert!(purge_deleted_bucket_result(Ok(())).is_ok());
-        assert!(purge_deleted_bucket_result(Err(StorageError::BucketNotFound("photos".to_string()))).is_ok());
-        assert!(purge_deleted_bucket_result(Err(StorageError::VolumeNotFound)).is_ok());
-        let err = purge_deleted_bucket_result(Err(StorageError::StorageFull))
-            .expect_err("non-not-found delete failures must propagate");
-        assert_ne!(*err.code(), S3ErrorCode::NoSuchBucket);
-    }
-
-    /// C5 red-light: the site-replication devnull drain must accept bodies
-    /// beyond the 1MB admin body cap — MinIO's link check streams large
-    /// probe bodies and treats a 400 as a broken link.
-    #[tokio::test]
-    async fn test_site_replication_devnull_drains_body_beyond_admin_cap() {
-        let body = Body::from(vec![0u8; MAX_ADMIN_REQUEST_BODY_SIZE + 1]);
-        drain_site_replication_devnull(body)
-            .await
-            .expect("devnull must drain bodies larger than the admin body cap");
-    }
-
-    /// A3 red-light: `versioningEnabled` must travel on every outbound
-    /// make-with-versioning bucket op so the query matches MinIO's
-    /// site-replication make-bucket wire contract (MinIO's own hook sends
-    /// `versioningEnabled=true` on this op).
-    #[test]
-    fn test_make_with_versioning_op_paths_send_versioning_enabled() {
-        let bucket = SRBucketInfo {
-            bucket: "photos".to_string(),
-            created_at: Some(OffsetDateTime::UNIX_EPOCH),
-            object_lock_config: Some(BASE64_STANDARD.encode("<ObjectLockConfiguration/>")),
-            ..Default::default()
-        };
-        let bootstrap = bootstrap_bucket_make_op_path(&bucket);
-        assert!(bootstrap.contains("operation=make-with-versioning"), "{bootstrap}");
-        assert!(bootstrap.contains("versioningEnabled=true"), "{bootstrap}");
-        assert!(bootstrap.contains("createdAt="), "{bootstrap}");
-        assert!(bootstrap.contains("lockEnabled=true"), "{bootstrap}");
-
-        // The broadcast path (create-bucket hook) shares the same builder.
-        let broadcast = make_with_versioning_bucket_op_path("photos", Some("1970-01-01T00:00:00Z"), false);
-        assert!(broadcast.contains("versioningEnabled=true"), "{broadcast}");
-        assert!(!broadcast.contains("lockEnabled"), "{broadcast}");
-    }
-
     #[tokio::test]
     #[serial]
     async fn test_add_bootstrap_scope_only_allows_expected_bucket_setup_until_guard_drops() {

rustfs/src/admin/site_replication_identity.rs

@@ -13,9 +13,9 @@
 // limitations under the License.
 
 use rustfs_madmin::{PeerInfo, SyncStatus};
-use std::collections::BTreeMap;
+use std::collections::{BTreeMap, hash_map::DefaultHasher};
+use std::hash::{Hash, Hasher};
 use url::Url;
-use uuid::Uuid;
 
 fn has_http_scheme(endpoint: &str) -> bool {
     endpoint.get(..7).is_some_and(|prefix| prefix.eq_ignore_ascii_case("http://"))
@@ -66,12 +66,10 @@ pub fn site_identity_key(endpoint: &str) -> String {
         .unwrap_or_else(|| trimmed.to_ascii_lowercase())
 }
 
-/// Fallback deployment ID for a peer that reported none. UUIDv5 over the
-/// canonical endpoint: the ID is persisted in site-replication state and
-/// broadcast to peers, so it must be identical across Rust toolchains
-/// (`DefaultHasher` is not) and across spellings of the same endpoint.
 pub fn deployment_id_for_endpoint(endpoint: &str) -> String {
-    Uuid::new_v5(&Uuid::NAMESPACE_URL, canonical_endpoint(endpoint).as_bytes()).to_string()
+    let mut hasher = DefaultHasher::new();
+    endpoint.hash(&mut hasher);
+    format!("{:016x}", hasher.finish())
 }
 
 pub fn same_identity_endpoint(left: &str, right: &str) -> bool {
@@ -176,23 +174,6 @@ mod tests {
         }
     }
 
-    /// B8 red-light: the fallback deployment ID must be a toolchain-stable
-    /// UUIDv5 over the canonical endpoint — `DefaultHasher` output is not
-    /// guaranteed stable across Rust releases, yet the ID is persisted in
-    /// site-replication state and broadcast to peers.
-    #[test]
-    fn deployment_id_for_endpoint_is_stable_uuid_v5_over_canonical_endpoint() {
-        let endpoint = "https://node-a.example.com:9000";
-        let id = deployment_id_for_endpoint(endpoint);
-        let parsed = uuid::Uuid::parse_str(&id).expect("fallback deployment ID must be a UUID");
-        assert_eq!(parsed.get_version_num(), 5, "fallback deployment ID must be UUIDv5");
-        // Deterministic for the same endpoint and for spelling variants that
-        // share a canonical form; distinct endpoints stay distinct.
-        assert_eq!(id, deployment_id_for_endpoint(endpoint));
-        assert_eq!(id, deployment_id_for_endpoint(" HTTPS://Node-A.Example.Com:9000/ "));
-        assert_ne!(id, deployment_id_for_endpoint("https://node-b.example.com:9000"));
-    }
-
     #[test]
     fn canonical_endpoint_accepts_case_insensitive_scheme() {
         assert_eq!(

rustfs/src/admin/storage_api.rs

@@ -51,7 +51,7 @@ mod ecstore_disk {
 }
 
 mod ecstore_error {
-    pub(crate) use crate::storage::storage_api::ecstore_error::{StorageError, is_err_bucket_not_found};
+    pub(crate) use crate::storage::storage_api::ecstore_error::StorageError;
 }
 
 #[allow(unused_imports)]
@@ -919,7 +919,6 @@ pub(crate) mod contract {
 }
 
 pub(crate) mod error {
-    pub(crate) use super::ecstore_error::is_err_bucket_not_found;
     pub(crate) use super::{Error, StorageError};
 }