refactor(storage): narrow metadata bounds (#3343)

crates/ecstore/src/bucket/migration.rs

@@ -17,7 +17,9 @@
 use crate::bucket::metadata::BUCKET_METADATA_FILE;
 use crate::bucket::replication::{decode_resync_file, encode_resync_file};
 use crate::disk::{BUCKET_META_PREFIX, MIGRATING_META_BUCKET, RUSTFS_META_BUCKET};
-use crate::store_api::{BucketOptions, ObjectOptions, PutObjReader, StorageAPI};
+use crate::store_api::{
+    BucketOperations, BucketOptions, ListOperations, ObjectIO, ObjectOperations, ObjectOptions, PutObjReader,
+};
 use http::HeaderMap;
 use rustfs_policy::auth::UserIdentity;
 use rustfs_policy::policy::PolicyDoc;
@@ -176,7 +178,10 @@ fn normalize_bucket_meta_blob(path: &str, data: &[u8]) -> std::result::Result<Op
 /// Uses list_bucket (from disk volumes) to get bucket names, since list_objects_v2 on the legacy
 /// meta bucket may not work (legacy format differs from object layer expectations).
 /// Skips buckets that already exist in RustFS (idempotent).
-pub async fn try_migrate_bucket_metadata<S: StorageAPI>(store: Arc<S>) {
+pub async fn try_migrate_bucket_metadata<S>(store: Arc<S>)
+where
+    S: BucketOperations + ObjectIO + ObjectOperations,
+{
     let buckets_list = match store
         .list_bucket(&BucketOptions {
             no_metadata: true,
@@ -218,13 +223,10 @@ pub async fn try_migrate_bucket_metadata<S: StorageAPI>(store: Arc<S>) {
     }
 }
 
-async fn migrate_one_if_missing<S: StorageAPI>(
-    store: Arc<S>,
-    opts: &ObjectOptions,
-    headers: &HeaderMap,
-    path: &str,
-    label: &str,
-) {
+async fn migrate_one_if_missing<S>(store: Arc<S>, opts: &ObjectOptions, headers: &HeaderMap, path: &str, label: &str)
+where
+    S: ObjectIO + ObjectOperations,
+{
     if store
         .get_object_info(RUSTFS_META_BUCKET, path, &ObjectOptions::default())
         .await
@@ -275,7 +277,10 @@ async fn migrate_one_if_missing<S: StorageAPI>(
 /// Lists all objects under the IAM prefix in the source, copies each to the target if not present.
 /// Skips objects that already exist in RustFS (idempotent).
 /// If list_objects_v2 on the legacy bucket fails (e.g. format differs), migration is skipped.
-pub async fn try_migrate_iam_config<S: StorageAPI>(store: Arc<S>) {
+pub async fn try_migrate_iam_config<S>(store: Arc<S>)
+where
+    S: ListOperations + ObjectIO + ObjectOperations,
+{
     let opts = ObjectOptions {
         max_parity: true,
         no_lock: true,

crates/ecstore/src/config/com.rs

@@ -16,7 +16,7 @@ use crate::config::{Config, KVS, audit, notify, oidc, set_global_storage_class,
 use crate::disk::{MIGRATING_META_BUCKET, RUSTFS_META_BUCKET};
 use crate::error::{Error, Result};
 use crate::global::is_first_cluster_node_local;
-use crate::store_api::{ObjectInfo, ObjectOptions, PutObjReader, StorageAPI};
+use crate::store_api::{ObjectIO, ObjectInfo, ObjectOperations, ObjectOptions, PutObjReader};
 use http::HeaderMap;
 use rustfs_config::audit::{
     AUDIT_AMQP_KEYS, AUDIT_AMQP_SUB_SYS, AUDIT_KAFKA_KEYS, AUDIT_KAFKA_SUB_SYS, AUDIT_MQTT_KEYS, AUDIT_MQTT_SUB_SYS,
@@ -181,12 +181,12 @@ fn audit_target_descriptors() -> [TargetConfigDescriptor; 9] {
 }
 
 #[instrument(skip(api))]
-pub async fn read_config<S: StorageAPI>(api: Arc<S>, file: &str) -> Result<Vec<u8>> {
+pub async fn read_config<S: ObjectIO>(api: Arc<S>, file: &str) -> Result<Vec<u8>> {
     let (data, _obj) = read_config_with_metadata(api, file, &ObjectOptions::default()).await?;
     Ok(data)
 }
 
-pub async fn read_config_no_lock<S: StorageAPI>(api: Arc<S>, file: &str) -> Result<Vec<u8>> {
+pub async fn read_config_no_lock<S: ObjectIO>(api: Arc<S>, file: &str) -> Result<Vec<u8>> {
     let (data, _obj) = read_config_with_metadata(
         api,
         file,
@@ -199,7 +199,7 @@ pub async fn read_config_no_lock<S: StorageAPI>(api: Arc<S>, file: &str) -> Resu
     Ok(data)
 }
 
-pub async fn read_config_with_metadata<S: StorageAPI>(
+pub async fn read_config_with_metadata<S: ObjectIO>(
     api: Arc<S>,
     file: &str,
     opts: &ObjectOptions,
@@ -227,7 +227,7 @@ pub async fn read_config_with_metadata<S: StorageAPI>(
 }
 
 #[instrument(skip(api, data))]
-pub async fn save_config<S: StorageAPI>(api: Arc<S>, file: &str, data: Vec<u8>) -> Result<()> {
+pub async fn save_config<S: ObjectIO>(api: Arc<S>, file: &str, data: Vec<u8>) -> Result<()> {
     save_config_with_opts(
         api,
         file,
@@ -241,7 +241,7 @@ pub async fn save_config<S: StorageAPI>(api: Arc<S>, file: &str, data: Vec<u8>)
 }
 
 #[instrument(skip(api))]
-pub async fn delete_config<S: StorageAPI>(api: Arc<S>, file: &str) -> Result<()> {
+pub async fn delete_config<S: ObjectOperations>(api: Arc<S>, file: &str) -> Result<()> {
     match api
         .delete_object(
             RUSTFS_META_BUCKET,
@@ -265,7 +265,7 @@ pub async fn delete_config<S: StorageAPI>(api: Arc<S>, file: &str) -> Result<()>
     }
 }
 
-pub async fn save_config_with_opts<S: StorageAPI>(api: Arc<S>, file: &str, data: Vec<u8>, opts: &ObjectOptions) -> Result<()> {
+pub async fn save_config_with_opts<S: ObjectIO>(api: Arc<S>, file: &str, data: Vec<u8>, opts: &ObjectOptions) -> Result<()> {
     let mut put_data = PutObjReader::from_vec(data);
     if let Err(err) = api.put_object(RUSTFS_META_BUCKET, file, &mut put_data, opts).await {
         error!("save_config_with_opts: err: {:?}, file: {}", err, file);
@@ -280,7 +280,7 @@ fn new_server_config() -> Config {
 
 async fn new_and_save_server_config<S>(api: Arc<S>) -> Result<Config>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: ObjectIO + StorageAdminApi,
 {
     let mut cfg = new_server_config();
     lookup_configs(&mut cfg, api.clone()).await;
@@ -982,7 +982,10 @@ fn is_object_not_found(err: &Error) -> bool {
     *err == Error::FileNotFound || matches!(err, Error::ObjectNotFound(_, _) | Error::BucketNotFound(_))
 }
 
-pub async fn try_migrate_server_config<S: StorageAPI>(api: Arc<S>) {
+pub async fn try_migrate_server_config<S>(api: Arc<S>)
+where
+    S: ObjectIO + ObjectOperations,
+{
     let config_file = get_config_file();
     match api
         .get_object_info(
@@ -1065,7 +1068,7 @@ pub async fn try_migrate_server_config<S: StorageAPI>(api: Arc<S>) {
 /// Handle the situation where the configuration file does not exist, create and save a new configuration
 async fn handle_missing_config<S>(api: Arc<S>, context: &str) -> Result<Config>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: ObjectIO + StorageAdminApi,
 {
     warn!("Configuration not found ({}): Start initializing new configuration", context);
     let cfg = if is_first_cluster_node_local().await {
@@ -1087,7 +1090,7 @@ fn handle_config_read_error(err: Error, file_path: &str) -> Result<Config> {
 
 pub async fn read_config_without_migrate<S>(api: Arc<S>) -> Result<Config>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: ObjectIO + StorageAdminApi,
 {
     let config_file = get_config_file();
 
@@ -1101,7 +1104,7 @@ where
 
 async fn read_server_config<S>(api: Arc<S>, data: &[u8]) -> Result<Config>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: ObjectIO + StorageAdminApi,
 {
     // If the provided data is empty, try to read from the file again
     if data.is_empty() {
@@ -1125,7 +1128,7 @@ where
     Ok(cfg.merge())
 }
 
-pub async fn save_server_config<S: StorageAPI>(api: Arc<S>, cfg: &Config) -> Result<()> {
+pub async fn save_server_config<S: ObjectIO>(api: Arc<S>, cfg: &Config) -> Result<()> {
     let config_file = get_config_file();
     let existing = match read_config(api.clone(), &config_file).await {
         Ok(v) => Some(v),
@@ -1156,7 +1159,7 @@ pub async fn save_server_config<S: StorageAPI>(api: Arc<S>, cfg: &Config) -> Res
 
 pub async fn lookup_configs<S>(cfg: &mut Config, api: Arc<S>)
 where
-    S: StorageAPI + StorageAdminApi,
+    S: StorageAdminApi,
 {
     // TODO: from etcd
     if let Err(err) = apply_dynamic_config(cfg, api).await {
@@ -1166,7 +1169,7 @@ where
 
 async fn apply_dynamic_config<S>(cfg: &mut Config, api: Arc<S>) -> Result<()>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: StorageAdminApi,
 {
     for key in SUB_SYSTEMS_DYNAMIC.iter() {
         apply_dynamic_config_for_sub_sys(cfg, api.clone(), key).await?;
@@ -1177,7 +1180,7 @@ where
 
 async fn apply_dynamic_config_for_sub_sys<S>(cfg: &mut Config, api: Arc<S>, subsys: &str) -> Result<()>
 where
-    S: StorageAPI + StorageAdminApi,
+    S: StorageAdminApi,
 {
     let set_drive_counts = StorageAdminApi::set_drive_counts(api.as_ref());
     if subsys == STORAGE_CLASS_SUB_SYS {

crates/ecstore/src/rebalance.rs

@@ -566,11 +566,11 @@ impl RebalanceMeta {
     pub fn new() -> Self {
         Self::default()
     }
-    pub async fn load<S: StorageAPI>(&mut self, store: Arc<S>) -> Result<()> {
+    pub async fn load<S: ObjectIO>(&mut self, store: Arc<S>) -> Result<()> {
         self.load_with_opts(store, ObjectOptions::default()).await
     }
 
-    pub async fn load_with_opts<S: StorageAPI>(&mut self, store: Arc<S>, opts: ObjectOptions) -> Result<()> {
+    pub async fn load_with_opts<S: ObjectIO>(&mut self, store: Arc<S>, opts: ObjectOptions) -> Result<()> {
         let (data, _) = read_config_with_metadata(store, REBAL_META_NAME, &opts).await?;
         if data.is_empty() {
             info!("rebalanceMeta load_with_opts: no data");
@@ -599,11 +599,11 @@ impl RebalanceMeta {
         Ok(())
     }
 
-    pub async fn save<S: StorageAPI>(&self, store: Arc<S>) -> Result<()> {
+    pub async fn save<S: ObjectIO>(&self, store: Arc<S>) -> Result<()> {
         self.save_with_opts(store, ObjectOptions::default()).await
     }
 
-    pub async fn save_with_opts<S: StorageAPI>(&self, store: Arc<S>, opts: ObjectOptions) -> Result<()> {
+    pub async fn save_with_opts<S: ObjectIO>(&self, store: Arc<S>, opts: ObjectOptions) -> Result<()> {
         if self.pool_stats.is_empty() {
             info!("rebalanceMeta save_with_opts: no pool stats");
             return Ok(());

crates/ecstore/src/tier/tier.rs

@@ -46,12 +46,11 @@ use crate::tier::{
     warm_backend::{check_warm_backend, new_warm_backend},
 };
 use crate::{
-    StorageAPI,
     config::com::{CONFIG_PREFIX, read_config},
     disk::{MIGRATING_META_BUCKET, RUSTFS_META_BUCKET},
     global::is_first_cluster_node_local,
     store::ECStore,
-    store_api::{ObjectIO as _, ObjectOptions, PutObjReader},
+    store_api::{ObjectIO, ObjectOperations, ObjectOptions, PutObjReader},
 };
 use rustfs_rio::HashReader;
 use rustfs_utils::path::{SLASH_SEPARATOR, path_join};
@@ -1033,14 +1032,14 @@ impl TierConfigMgr {
         self.save_tiering_config(api).await
     }
 
-    pub async fn save_tiering_config<S: StorageAPI>(&self, api: Arc<S>) -> std::result::Result<(), std::io::Error> {
+    pub async fn save_tiering_config<S: ObjectIO>(&self, api: Arc<S>) -> std::result::Result<(), std::io::Error> {
         let data = encode_external_tiering_config_blob(self)?;
         let config_file = tier_config_path(TIER_CONFIG_FILE);
 
         self.save_config(api, &config_file, data).await
     }
 
-    pub async fn save_config<S: StorageAPI>(
+    pub async fn save_config<S: ObjectIO>(
         &self,
         api: Arc<S>,
         file: &str,
@@ -1058,7 +1057,7 @@ impl TierConfigMgr {
         .await
     }
 
-    pub async fn save_config_with_opts<S: StorageAPI>(
+    pub async fn save_config_with_opts<S: ObjectIO>(
         &self,
         api: Arc<S>,
         file: &str,
@@ -1100,7 +1099,7 @@ impl TierConfigMgr {
     }
 }
 
-async fn new_and_save_tiering_config<S: StorageAPI>(api: Arc<S>) -> Result<TierConfigMgr> {
+async fn new_and_save_tiering_config<S: ObjectIO>(api: Arc<S>) -> Result<TierConfigMgr> {
     let mut cfg = TierConfigMgr {
         driver_cache: HashMap::new(),
         tiers: HashMap::new(),
@@ -1159,7 +1158,7 @@ async fn load_tier_config(api: Arc<ECStore>) -> std::result::Result<TierConfigMg
     }
 }
 
-async fn read_tier_config_from_bucket<S: StorageAPI>(
+async fn read_tier_config_from_bucket<S: ObjectIO>(
     api: Arc<S>,
     bucket: &str,
     path: &str,
@@ -1177,7 +1176,7 @@ async fn read_tier_config_from_bucket<S: StorageAPI>(
     Ok(Some(data))
 }
 
-async fn write_tier_config_to_rustfs<S: StorageAPI>(api: Arc<S>, path: &str, data: Bytes) -> io::Result<()> {
+async fn write_tier_config_to_rustfs<S: ObjectIO>(api: Arc<S>, path: &str, data: Bytes) -> io::Result<()> {
     let mut put_data = PutObjReader::from_vec(data.to_vec());
     api.put_object(
         RUSTFS_META_BUCKET,
@@ -1193,7 +1192,10 @@ async fn write_tier_config_to_rustfs<S: StorageAPI>(api: Arc<S>, path: &str, dat
     Ok(())
 }
 
-pub async fn try_migrate_tiering_config<S: StorageAPI>(api: Arc<S>) {
+pub async fn try_migrate_tiering_config<S>(api: Arc<S>)
+where
+    S: ObjectIO + ObjectOperations,
+{
     let target_path = tier_config_path(TIER_CONFIG_FILE);
     if api
         .get_object_info(

docs/architecture/migration-progress.md

@@ -5,15 +5,15 @@ Status values: `[ ]` not started, `[~]` in progress, `[x]` complete, `[!]` block
 ## Current Context
 
 - Issue: [`rustfs/backlog#660`](https://github.com/rustfs/backlog/issues/660)
-- Branch: `overtrue/arch-storage-api-next-contracts`
-- Baseline: `origin/main` at `87968275a776362e9b3734eac5bd9e3233d6cbe9`
+- Branch: `overtrue/arch-storage-api-bound-narrowing`
+- Baseline: `origin/main` at `6b86e3875c4912f3a7c27915b0bbd43399402847`
 - PR type for this branch: `dependency-migration`
 - Runtime behavior changes: none.
-- Rust code changes: remove duplicate admin-read methods from the old
-  `StorageAPI` trait after all admin inventory consumers have moved to
-  `StorageAdminApi`.
+- Rust code changes: narrow metadata helper generic bounds away from full
+  `StorageAPI` where helpers only need object I/O, object metadata/delete,
+  bucket listing, object listing, or `StorageAdminApi`.
 - CI/script changes: none.
-- Docs changes: record API-007 completion and the current admin surface cleanup
+- Docs changes: record API-008 completion and the current bound-narrowing
   context, verification evidence, and expert review outcomes.
 
 ## Phase 0 Tasks
@@ -218,25 +218,33 @@ Status values: `[ ]` not started, `[~]` in progress, `[x]` complete, `[!]` block
     drive-count, or disk-set inventory when the inventory-facing
     `StorageAdminApi` contract represents the same read-only operation.
 
-- [~] `API-008` Remove duplicate old-path admin surfaces.
-  - Current branch slice: remove admin-read methods from the old `StorageAPI`
-    trait and its ECStore/Sets/SetDisks test implementations after API-007
-    migrated their consumers.
+- [x] `API-008` Remove duplicate old-path admin surfaces.
+  - Completed slice: `rustfs/rustfs#3340` removed duplicate admin-read methods
+    from the old `StorageAPI` trait and its ECStore/Sets/SetDisks/test
+    implementations after API-007 migrated their consumers.
   - Acceptance: old `StorageAPI` keeps storage operation traits and
     `new_ns_lock`, while admin inventory surfaces live only on
     `StorageAdminApi`.
-  - Must preserve: object API trait bounds, namespace lock behavior, bucket,
-    object, list, multipart, and heal operations; config read/write byte shape;
-    dynamic storage-class lookup; object/rebalance object selection paths;
-    scanner cache persistence; replication/tier persistence; and storage hot
-    paths.
-  - Risk defense: do not remove `StorageAPI` itself, do not move object/config
-    persistence contracts, and only replace previous `SetDisks::get_disks(0, 0)`
-    trait calls with the same internal `disk_inventory()` data source that the
-    old impl returned.
-  - Verification: pending focused compile/tests, migration guards, Rust risk
-    scan, and required quality/architecture, migration-preservation, and
-    testing/verification review.
+
+- [~] `API-009` Narrow metadata helper storage bounds.
+  - Current branch slice: narrow server config, tier config, rebalance
+    metadata, and startup metadata migration helper bounds away from full
+    `StorageAPI` when the helper only needs `ObjectIO`, `ObjectOperations`,
+    `BucketOperations`, `ListOperations`, or `StorageAdminApi`.
+  - Acceptance: metadata helper contracts express the actual operation group
+    they need, while callers and persistence behavior remain unchanged.
+  - Must preserve: config/tier/rebalance serialized byte shapes, read/write
+    object names, legacy bucket/IAM migration normalization and skip behavior,
+    lock behavior around rebalance metadata merge, dynamic storage-class lookup,
+    object APIs, scanner cache persistence, heal repair, replication
+    persistence, and storage hot paths.
+  - Risk defense: do not move traits to `rustfs-storage-api`, do not remove
+    `StorageAPI`, do not alter helper bodies or storage options, and keep
+    `save_rebalance_meta_with_merge` on full `StorageAPI` because it still needs
+    `new_ns_lock`.
+  - Verification: focused compile/tests, migration guards, Rust risk scan, and
+    required quality/architecture, migration-preservation, and
+    testing/verification review passed before push.
 
 ## Phase 8 Background Controller Tasks
 
@@ -261,28 +269,26 @@ Status values: `[ ]` not started, `[~]` in progress, `[x]` complete, `[!]` block
 
 ## Next PRs
 
-1. `consumer-migration`: migrate the remaining readiness/admin/capacity
-   consumers to the inventory-facing admin contract one group at a time.
-2. `dependency-migration`: remove duplicate old-path admin surfaces only after
-   consumer migration proves equivalent behavior.
-3. `api-extraction`: move only the pure server-config model into
+1. `dependency-migration`: narrow metadata/helper bounds away from full
+   `StorageAPI` where the helper only needs a specific operation trait.
+2. `api-extraction`: move only the pure server-config model into
    rustfs-config as CFG-003.
-4. `api-extraction`: keep the old rustfs_ecstore::config::* path with
+3. `api-extraction`: keep the old rustfs_ecstore::config::* path with
    RUSTFS_COMPAT_TODO(CFG-004) and cleanup-register coverage.
-5. `consumer-migration`: migrate external consumers one group at a time only
+4. `consumer-migration`: migrate external consumers one group at a time only
    after the model path and compatibility shim are stable.
-6. `security-change`: make Local KMS unsafe defaults explicit development
+5. `security-change`: make Local KMS unsafe defaults explicit development
    opt-ins or production failures in KMSD-002.
-7. `security-change`: make Vault unsafe defaults explicit development opt-ins
+6. `security-change`: make Vault unsafe defaults explicit development opt-ins
    or production failures in KMSD-003.
 
 ## Pre-Push Review Log
 
 | Expert | Status | Notes |
 |---|---|---|
-| Quality/architecture | pass | Confirmed the old `StorageAPI` now keeps only operation traits and `new_ns_lock`, admin inventory surfaces remain on `StorageAdminApi`, dependency direction is clean, and scope/naming stay limited to API-008. |
-| Migration preservation | pass | Confirmed object/bucket/list/multipart/heal contracts, namespace locks, config byte paths, dynamic storage-class behavior, rebalance selection, scanner cache, heal repair, replication/tier/config persistence, and storage hot paths are preserved. |
-| Testing/verification | pass | Initially requested wider rebalance coverage; after `cargo test -p rustfs-ecstore rebalance --lib` passed, confirmed the focused matrix is sufficient for this larger dependency-migration slice while skipping full pre-commit per current instruction. |
+| Quality/architecture | pass | Confirmed the Rust diff only changes imports/generic bounds, the new operation-group bounds match actual calls, `StorageAPI` and `new_ns_lock` remain where needed, and no over-abstraction, naming regression, dependency reversal, or hot-path intrusion was found. |
+| Migration preservation | pass | Confirmed config/tier/rebalance byte shapes, object names, buckets, options, bucket/IAM normalization and skip behavior, dynamic storage-class lookup, and rebalance metadata merge locking remain unchanged. |
+| Testing/verification | pass | Confirmed the focused compile, config/tier/rebalance/bucket migration tests, migration guards, diff hygiene, and added-line Rust risk scan are sufficient; full pre-commit can be skipped under the current larger-granularity instruction. |
 
 ## Verification Notes
 
@@ -292,42 +298,40 @@ Passed:
 - `cargo fmt --all --check`.
 - `cargo check -p rustfs-storage-api -p rustfs-ecstore -p rustfs --lib`.
 - `cargo test -p rustfs-ecstore config --lib`; 59 passed.
-- `cargo test -p rustfs-ecstore store::rebalance --lib`; 19 passed.
+- `cargo test -p rustfs-ecstore tier --lib`; 27 passed.
 - `cargo test -p rustfs-ecstore rebalance --lib`; 198 passed.
-- `cargo test -p rustfs-ecstore set_disk --lib`; 86 passed.
+- `cargo test -p rustfs-ecstore bucket::migration --lib`; 2 passed.
 - `./scripts/check_architecture_migration_rules.sh`.
 - `./scripts/check_layer_dependencies.sh`.
 - `./scripts/check_metrics_migration_refs.sh`.
 - `./scripts/check_unsafe_code_allowances.sh`.
 - `git diff --check`.
-- Rust code-quality scan on changed `.rs` files, plus added-line scan for
-  unwrap/expect, numeric casts, `Result<_, String>`, `Box<dyn Error>`,
-  println/eprintln, and `Ordering::Relaxed`.
+- Rust code-quality scan on changed `.rs` files: broad full-file matches are
+  pre-existing test/business-code patterns; added-line scan found no new
+  unwrap/expect, numeric cast, `Result<_, String>`, `Box<dyn Error>`,
+  println/eprintln, or `Ordering::Relaxed`.
 
 Notes:
 - Full pre-commit is intentionally skipped when the focused tests and guards
   pass, per the current instruction to increase PR granularity.
-- The broad changed-file quality scan reports pre-existing unwrap/cast matches
-  and `Result<String>` business return values in touched ECStore files; the
-  added-line scan found no new risky code patterns.
-- Old `StorageAPI` remains as the object/bucket/list/multipart/heal namespace
-  lock contract; only duplicate admin inventory methods are removed.
-- Config dynamic storage-class lookup now reads drive counts through
-  `StorageAdminApi`.
-- Object/rebalance object selection paths, scanner cache persistence, heal
-  object repair, object APIs, replication/config/tier persistence paths, and
-  storage hot paths must remain unchanged.
+- This slice changes generic bounds and imports only; helper bodies, storage
+  options, object names, and metadata normalization logic are unchanged.
+- `save_rebalance_meta_with_merge` intentionally still requires full
+  `StorageAPI` because it uses `new_ns_lock`.
 - No temporary compatibility shim was added.
 
 ## Handoff Notes
 
-- Keep this API-008 slice as a `dependency-migration` PR that only removes
-  duplicate admin inventory methods from old `StorageAPI`.
+- Keep this API-009 slice as a `dependency-migration` PR that only narrows
+  metadata/helper generic bounds for config, tier, rebalance, and startup
+  metadata migration helpers.
 - Do not remove `StorageAPI` itself, object operation traits, or
   `new_ns_lock` in this PR.
-- Do not migrate scanner cache writes, heal object repair, object APIs,
-  replication, config/tier persistence, or storage hot-path consumers in this
-  PR.
-- Do not make the old `StorageAPI` trait inherit `StorageAdminApi`.
+- Do not move traits into `rustfs-storage-api` or introduce new compatibility
+  shims in this PR.
+- Do not alter config/tier/rebalance serialized byte shapes, legacy bucket/IAM
+  migration normalization, object names, storage options, scanner cache writes,
+  heal object repair, object APIs, replication persistence, or storage hot-path
+  consumers in this PR.
 - Do not add temporary compatibility code unless a matching
   `RUSTFS_COMPAT_TODO(<task-id>)` marker and cleanup-register entry are added.