Merge branch 'main' into overtrue/require-mc-e2e

* test(kms): bound Vault failover progress wait

* ci(nightly): honor manual dispatch ref

* test(kms): preserve Vault worker failures

* test(kms): validate Vault circuit recovery

.config/nextest.toml

@@ -325,8 +325,8 @@ slow-timeout = { period = "60s", terminate-after = 2, grace-period = "10s" }
 #
 # Wired by .github/workflows/e2e-replication-nightly.yml (schedule +
 # workflow_dispatch), which builds the rustfs binary once, installs awscurl so
-# the STS dual-node test actually exercises its path (it skips gracefully with
-# a visible log line when awscurl is absent), and routes scheduled failures
+# the STS dual-node test actually exercises its path (the test fails when
+# awscurl is absent), and routes scheduled failures
 # through .github/actions/schedule-failure-issue (ci-8). Explicit division of
 # labor with e2e-full: these tests run only in the consolidated nightly
 # workflow, not in the merge/main lane.

.github/workflows/ci.yml

@@ -681,6 +681,19 @@ jobs:
           cache-save-if: 'false'
           install-build-packaging-tools: 'false'
 
+      - name: Set up Python
+        uses: actions/setup-python@ece7cb06caefa5fff74198d8649806c4678c61a1 # v6.3.0
+        with:
+          python-version: "3.12"
+
+      - name: Install awscurl
+        run: |
+          python3 -m pip install --user --upgrade pip "awscurl==0.44"
+          echo "AWSCURL_PATH=$HOME/.local/bin/awscurl" >> "$GITHUB_ENV"
+
+      - name: Verify awscurl
+        run: test -x "$AWSCURL_PATH"
+
       # Download after the cache restore so the freshly built binary from the
       # build job always wins over anything restored into target/debug.
       - name: Download debug binary
@@ -803,6 +816,20 @@ jobs:
       - name: Verify awscurl
         run: test -x "$AWSCURL_PATH"
 
+      - name: Install mc
+        env:
+          MC_VERSION: RELEASE.2025-08-13T08-35-41Z
+          MC_SHA256: 01f866e9c5f9b87c2b09116fa5d7c06695b106242d829a8bb32990c00312e891
+        run: |
+          MC_BINARY="mc.linux-amd64.${MC_VERSION}"
+          curl -fsSLo "$RUNNER_TEMP/mc" "https://github.com/minio/mc/releases/download/${MC_VERSION}/${MC_BINARY}"
+          echo "${MC_SHA256}  $RUNNER_TEMP/mc" | sha256sum --check --status
+          chmod +x "$RUNNER_TEMP/mc"
+          echo "$RUNNER_TEMP" >> "$GITHUB_PATH"
+
+      - name: Verify mc
+        run: mc --version
+
       - name: Install Vault
         run: |
           VAULT_VERSION="1.17.6"

.github/workflows/e2e-replication-nightly.yml

@@ -75,11 +75,7 @@ jobs:
           cache-save-if: ${{ github.ref == 'refs/heads/main' }}
           install-build-packaging-tools: 'false'
 
-      # awscurl lets the STS dual-node test actually exercise its path. Without
-      # it the test skips gracefully with a visible log line
-      # (`awscurl_available()` in crates/e2e_test/src/common.rs), so the lane
-      # still passes — installing it just upgrades that one test from skip to
-      # real coverage.
+      # The STS dual-node test requires awscurl and fails if it is unavailable.
       - name: Set up Python
         uses: actions/setup-python@ece7cb06caefa5fff74198d8649806c4678c61a1 # v6.3.0
         with:
@@ -87,7 +83,7 @@ jobs:
 
       - name: Install awscurl
         run: |
-          python3 -m pip install --user --upgrade pip awscurl
+          python3 -m pip install --user --upgrade pip "awscurl==0.44"
           echo "AWSCURL_PATH=$HOME/.local/bin/awscurl" >> "$GITHUB_ENV"
 
       - name: Verify awscurl

.github/workflows/nightly-gnu.yml

@@ -39,11 +39,10 @@ jobs:
     env:
       FORCE_JAVASCRIPT_ACTIONS_TO_NODE24: "true"
     steps:
-      - name: Checkout main branch
+      - name: Checkout repository
         uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7
         with:
           persist-credentials: false
-          ref: main
 
       - name: Setup Rust environment
         uses: ./.github/actions/setup
@@ -89,11 +88,10 @@ jobs:
       # either casing.
       NO_PROXY: 127.0.0.1,localhost
     steps:
-      - name: Checkout main branch
+      - name: Checkout repository
         uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7
         with:
           persist-credentials: false
-          ref: main
 
       - name: Setup Rust environment
         uses: ./.github/actions/setup
@@ -178,11 +176,10 @@ jobs:
       FORCE_JAVASCRIPT_ACTIONS_TO_NODE24: "true"
       NO_PROXY: 127.0.0.1,localhost
     steps:
-      - name: Checkout main branch
+      - name: Checkout repository
         uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7
         with:
           persist-credentials: false
-          ref: main
 
       - name: Setup Rust environment
         uses: ./.github/actions/setup

crates/e2e_test/README.md

@@ -123,7 +123,7 @@ via `create_s3_client(idx)` / `create_all_clients()`. See
 | `find_available_port` | Random free port (isolation primitive) |
 | `rustfs_binary_path` / `_with_features` | Locate/build the binary; honors `RUSTFS_BUILD_FEATURES` |
 | `requested_rustfs_build_features` / `rustfs_build_feature_enabled` | Feature-gate a test to what the binary was built with |
-| `awscurl_available` + `execute_awscurl` / `awscurl_post` / `_get` / `_put` / `_delete` / `awscurl_post_sts_form_urlencoded` | Admin/STS API calls via `awscurl` (skip gracefully when absent) |
+| `execute_awscurl` / `awscurl_post` / `_get` / `_put` / `_delete` / `awscurl_post_sts_form_urlencoded` | Admin/STS API calls via `awscurl`; missing binaries are test failures |
 | `replication_fast_env` | Env vars that shrink replication timers (from repl-4); pass to `start_rustfs_server_with_env` |
 | `local_http_client` / `init_logging` | Loopback HTTP client; idempotent tracing init |
 | `RustFSTestClusterEnvironment` (`new`/`start`/`start_node`/`stop_node`/`create_all_clients`) | Multi-node harness |
@@ -189,7 +189,7 @@ cargo nextest run --profile e2e-smoke -p e2e_test
 cargo nextest run --profile e2e-full -p e2e_test
 # Cluster fault nightly lane
 cargo nextest run --profile e2e-nightly -p e2e_test
-# Replication nightly lane; install awscurl so STS paths do not skip
+# Replication nightly lane; awscurl is required for STS paths
 cargo nextest run --profile e2e-repl-nightly -p e2e_test
 # Fixed-port protocol nightly lane
 RUSTFS_BUILD_FEATURES=ftps,webdav,sftp \
@@ -221,9 +221,8 @@ The `s3s-e2e` CI job selects a random `RUSTFS_TEST_PORT` (see the `e2e-tests`
 job) to dodge this; local single-node tests already use random ports, so a
 lingering orphan is usually the cause of a spurious bind failure.
 
-**`awscurl` not found.** `awscurl`-dependent tests skip gracefully with a
-visible log line (`awscurl_available()`); install `awscurl` to actually run
-them.
+**`awscurl` not found.** `awscurl`-dependent tests fail closed with a process
+spawn error. Install the pinned CI version before running their profiles.
 
 ## Related
 
@@ -258,10 +257,9 @@ A test module may join the smoke filter only if every test in it is:
 2. **Single-node** — spawns its own server via
    `RustFSTestEnvironment`/`start_rustfs_server` on a random port with an
    isolated temp dir. No `RustFSTestClusterEnvironment`, no fixed ports.
-3. **Dependency-free** — no pre-started server at `localhost:9000`, no Vault,
-   no fixed protocol ports. Tools that may be absent on the runner (e.g.
-   `awscurl`) are acceptable only when the test skips gracefully with a
-   visible log line (see `bucket_policy_check_test.rs`).
+3. **Hermetic dependencies** — no pre-started server at `localhost:9000`, no
+   Vault, and no fixed protocol ports. Any required CLI must be pinned and
+   installed by the workflow; a missing CLI must fail the test.
 4. **Not `#[ignore]`** — ignored tests are activation work (backlog#1149
    ci-13 / backlog#1148 ilm-3), not smoke candidates.
 

crates/e2e_test/src/bucket_policy_check_test.rs

@@ -52,10 +52,6 @@ fn create_user_client(env: &RustFSTestEnvironment, access_key: &str, secret_key:
 #[tokio::test]
 async fn test_bucket_policy_authenticated_user() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !crate::common::awscurl_available() {
-        info!("Skipping test_bucket_policy_authenticated_user because awscurl is not available");
-        return Ok(());
-    }
     info!("Starting test_bucket_policy_authenticated_user...");
 
     let mut env = RustFSTestEnvironment::new().await?;

crates/e2e_test/src/common.rs

@@ -494,17 +494,6 @@ fn awscurl_binary_path() -> PathBuf {
         .unwrap_or_else(|| PathBuf::from("awscurl"))
 }
 
-pub fn awscurl_available() -> bool {
-    let path = awscurl_binary_path();
-    if path.components().count() > 1 || path.is_absolute() {
-        return path.is_file();
-    }
-
-    std::env::var_os("PATH")
-        .map(|paths| std::env::split_paths(&paths).any(|dir| dir.join(&path).is_file()))
-        .unwrap_or(false)
-}
-
 // Global initialization
 static INIT: Once = Once::new();
 

crates/e2e_test/src/existing_object_tag_policy_test.rs

@@ -16,9 +16,7 @@
 //! session policy** (`Policy` parameter) via `awscurl --service sts` with explicit
 //! `Content-Type: application/x-www-form-urlencoded` on `POST /`.
 
-use crate::common::{
-    RustFSTestEnvironment, awscurl_available, awscurl_delete, awscurl_post_sts_form_urlencoded, awscurl_put, init_logging,
-};
+use crate::common::{RustFSTestEnvironment, awscurl_delete, awscurl_post_sts_form_urlencoded, awscurl_put, init_logging};
 use aws_sdk_s3::config::{Credentials, Region};
 use aws_sdk_s3::primitives::ByteStream;
 use aws_sdk_s3::types::{Delete, ObjectIdentifier, Tag, Tagging};
@@ -175,11 +173,6 @@ async fn cleanup_bucket_and_object(admin: &Client, bucket: &str, key: &str) {
 #[tokio::test]
 async fn test_e2e_iam_policy_existing_object_tag_get_object() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !awscurl_available() {
-        info!("Skipping test_e2e_iam_policy_existing_object_tag_get_object: awscurl not available");
-        return Ok(());
-    }
-
     let suffix = Uuid::new_v4();
     let user = format!("e2eiamtag-{suffix}");
     let user_secret = "longSecretKeyForTest123!";
@@ -233,11 +226,6 @@ async fn test_e2e_iam_policy_existing_object_tag_get_object() -> Result<(), Box<
 #[tokio::test]
 async fn test_e2e_bucket_policy_existing_object_tag_get_object() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !awscurl_available() {
-        info!("Skipping test_e2e_bucket_policy_existing_object_tag_get_object: awscurl not available");
-        return Ok(());
-    }
-
     let suffix = Uuid::new_v4();
     let user = format!("e2ebptag-{suffix}");
     let user_secret = "longSecretKeyForTest456!";
@@ -294,11 +282,6 @@ async fn test_e2e_bucket_policy_existing_object_tag_get_object() -> Result<(), B
 #[tokio::test]
 async fn test_e2e_sts_assume_role_session_policy_existing_object_tag() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !awscurl_available() {
-        info!("Skipping test_e2e_sts_assume_role_session_policy_existing_object_tag: awscurl not available");
-        return Ok(());
-    }
-
     let suffix = Uuid::new_v4();
     let parent = format!("e2e-sts-par-{suffix}");
     let parent_secret = "longSecretKeyForParentSts99!";
@@ -370,11 +353,6 @@ async fn test_e2e_sts_assume_role_session_policy_existing_object_tag() -> Result
 #[tokio::test]
 async fn test_e2e_sts_session_policy_delete_objects_object_prefix_only() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !awscurl_available() {
-        info!("Skipping test_e2e_sts_session_policy_delete_objects_object_prefix_only: awscurl not available");
-        return Ok(());
-    }
-
     let suffix = Uuid::new_v4();
     let parent = format!("e2e-sts-del-par-{suffix}");
     let parent_secret = "longSecretKeyForParentDelete99!";

crates/e2e_test/src/kms/common.rs

@@ -22,9 +22,7 @@
 //! - KMS backend configuration (Local and Vault)
 //! - SSE encryption testing utilities
 
-use crate::common::{
-    RustFSTestEnvironment, awscurl_available, awscurl_get, awscurl_post, init_logging as common_init_logging, local_http_client,
-};
+use crate::common::{RustFSTestEnvironment, awscurl_get, awscurl_post, init_logging as common_init_logging, local_http_client};
 use aws_sdk_s3::Client;
 use aws_sdk_s3::primitives::ByteStream;
 use aws_sdk_s3::types::ServerSideEncryption;
@@ -59,15 +57,6 @@ pub fn init_logging() {
     // Additional KMS-specific logging configuration can be added here if needed
 }
 
-pub fn skip_if_kms_admin_tool_unavailable(test_name: &str) -> bool {
-    if awscurl_available() {
-        return false;
-    }
-
-    info!("Skipping {} because awscurl is not available in PATH", test_name);
-    true
-}
-
 pub fn sse_customer_key_md5_base64(key: &str) -> String {
     let mut hasher = Md5::new();
     hasher.update(key.as_bytes());
@@ -490,10 +479,6 @@ pub async fn test_kms_key_management(
     access_key: &str,
     secret_key: &str,
 ) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
-    if skip_if_kms_admin_tool_unavailable("test_kms_key_management") {
-        return Ok(());
-    }
-
     info!("Testing KMS key management APIs");
 
     // Test CreateKey

crates/e2e_test/src/kms/kms_local_test.rs

@@ -20,8 +20,7 @@
 //! - Complete encryption/decryption lifecycle
 
 use super::common::{
-    LocalKMSTestEnvironment, get_kms_status, skip_if_kms_admin_tool_unavailable, sse_customer_key_md5_base64,
-    test_kms_key_management, test_sse_c_encryption,
+    LocalKMSTestEnvironment, get_kms_status, sse_customer_key_md5_base64, test_kms_key_management, test_sse_c_encryption,
 };
 use crate::common::{TEST_BUCKET, init_logging};
 use tracing::{error, info};
@@ -29,9 +28,6 @@ use tracing::{error, info};
 #[tokio::test]
 async fn test_local_kms_end_to_end() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_local_kms_end_to_end") {
-        return Ok(());
-    }
     info!("Starting Local KMS End-to-End Test");
 
     // Create LocalKMS test environment

crates/e2e_test/src/kms/kms_vault_test.rs

@@ -22,8 +22,8 @@ use crate::common::{TEST_BUCKET, init_logging};
 use tracing::{error, info};
 
 use super::common::{
-    VAULT_KEY_NAME, VaultTestEnvironment, get_kms_status, skip_if_kms_admin_tool_unavailable, sse_customer_key_md5_base64,
-    start_kms, test_all_multipart_encryption_types, test_error_scenarios, test_kms_key_management, test_sse_c_encryption,
+    VAULT_KEY_NAME, VaultTestEnvironment, get_kms_status, sse_customer_key_md5_base64, start_kms,
+    test_all_multipart_encryption_types, test_error_scenarios, test_kms_key_management, test_sse_c_encryption,
     test_sse_kms_encryption, test_sse_s3_encryption,
 };
 
@@ -62,9 +62,6 @@ impl VaultKmsTestContext {
 #[tokio::test]
 async fn test_vault_kms_end_to_end() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_vault_kms_end_to_end") {
-        return Ok(());
-    }
     info!("Starting Vault KMS End-to-End Test with default key {}", VAULT_KEY_NAME);
 
     let context = VaultKmsTestContext::new().await?;
@@ -117,9 +114,6 @@ async fn test_vault_kms_end_to_end() -> Result<(), Box<dyn std::error::Error + S
 #[tokio::test]
 async fn test_vault_kms_key_isolation() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_vault_kms_key_isolation") {
-        return Ok(());
-    }
     info!("Starting Vault KMS SSE-C key isolation test");
 
     let context = VaultKmsTestContext::new().await?;
@@ -203,9 +197,6 @@ async fn test_vault_kms_key_isolation() -> Result<(), Box<dyn std::error::Error
 #[tokio::test]
 async fn test_vault_kms_large_file() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_vault_kms_large_file") {
-        return Ok(());
-    }
     info!("Starting Vault KMS large file SSE-S3 test");
 
     let context = VaultKmsTestContext::new().await?;
@@ -267,9 +258,6 @@ async fn test_vault_kms_large_file() -> Result<(), Box<dyn std::error::Error + S
 #[tokio::test]
 async fn test_vault_kms_multipart_upload() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_vault_kms_multipart_upload") {
-        return Ok(());
-    }
     info!("Starting Vault KMS multipart upload encryption suite");
 
     let context = VaultKmsTestContext::new().await?;
@@ -297,9 +285,6 @@ async fn test_vault_kms_multipart_upload() -> Result<(), Box<dyn std::error::Err
 #[tokio::test]
 async fn test_vault_kms_key_operations() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if skip_if_kms_admin_tool_unavailable("test_vault_kms_key_operations") {
-        return Ok(());
-    }
     info!("Starting Vault KMS key operations test (CRUD)");
 
     let context = VaultKmsTestContext::new().await?;

crates/e2e_test/src/mc_mirror_small_bucket_test.rs

@@ -41,13 +41,6 @@ async fn create_issue_3107_fixture(root: &Path) -> TestResult {
     Ok(())
 }
 
-fn mc_available() -> bool {
-    Command::new("mc")
-        .arg("--version")
-        .output()
-        .is_ok_and(|output| output.status.success())
-}
-
 fn run_mc(args: &[&str]) -> TestResult {
     let output = Command::new("mc").args(args).output()?;
     if !output.status.success() {
@@ -75,10 +68,7 @@ fn count_files(root: &Path) -> usize {
 async fn test_mc_mirror_small_bucket_completes_without_list_timeout() -> TestResult {
     crate::common::init_logging();
     info!("Starting issue #3107 mc mirror regression test");
-    if !mc_available() {
-        info!("Skipping issue #3107 mc mirror regression test because mc is not installed");
-        return Ok(());
-    }
+    run_mc(&["--version"])?;
 
     let mut env = RustFSTestEnvironment::new().await?;
     env.start_rustfs_server(vec![]).await?;

crates/e2e_test/src/multipart_auth_test.rs

@@ -4278,10 +4278,6 @@ async fn test_signed_put_object_extract_preserves_pax_metadata_and_version_id()
 async fn test_signed_put_object_extract_authorizes_each_pax_privilege_and_retention_conditions()
 -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     init_logging();
-    if !crate::common::awscurl_available() {
-        return Ok(());
-    }
-
     let mut env = RustFSTestEnvironment::new().await?;
     env.start_rustfs_server(vec![]).await?;
 

crates/e2e_test/src/quota_test.rs

@@ -18,15 +18,6 @@ use http::{Method, StatusCode};
 use tokio::time::{Duration, sleep, timeout};
 use tracing::{debug, info};
 
-fn skip_without_awscurl() -> bool {
-    if crate::common::awscurl_available() {
-        return false;
-    }
-
-    info!("Skipping quota test because awscurl is not available");
-    true
-}
-
 /// Test environment setup for quota tests
 pub struct QuotaTestEnv {
     pub env: RustFSTestEnvironment,
@@ -276,9 +267,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_basic_operations() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         // Create test bucket
@@ -320,9 +308,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_admission_aws_chunked_declared_encoding() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
         env.create_bucket().await?;
 
@@ -371,9 +356,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_update_and_clear() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -406,9 +388,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_delete_operations() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -442,9 +421,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_usage_tracking() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -480,9 +456,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_statistics() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -513,9 +486,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_check_api() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -553,9 +523,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_multiple_buckets() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         // Create two buckets in the same environment
@@ -593,9 +560,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_error_handling() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -628,9 +592,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_http_endpoints() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -689,9 +650,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_normal_user_permissions() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
         env.create_bucket().await?;
 
@@ -744,9 +702,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_copy_operations() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -789,9 +744,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_batch_delete() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;
@@ -847,9 +799,6 @@ mod integration_tests {
     #[tokio::test]
     async fn test_quota_multipart_upload() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
         init_logging();
-        if skip_without_awscurl() {
-            return Ok(());
-        }
         let env = QuotaTestEnv::new().await?;
 
         env.create_bucket().await?;

crates/e2e_test/src/replication_extension_test.rs

@@ -13,9 +13,8 @@
 // limitations under the License.
 
 use crate::common::{
-    RustFSTestEnvironment, admin_create_user, awscurl_available, awscurl_post_sts_form_urlencoded, init_logging,
-    local_http_client, replication_fast_env, rustfs_binary_path, signed_request, signed_request_with_client,
-    signed_request_with_session_token,
+    RustFSTestEnvironment, admin_create_user, awscurl_post_sts_form_urlencoded, init_logging, local_http_client,
+    replication_fast_env, rustfs_binary_path, signed_request, signed_request_with_client, signed_request_with_session_token,
 };
 use crate::fake_s3_target::{
     FAKE_ACCESS_KEY, FAKE_SECRET_KEY, FakeS3Target, FaultAction as FakeTargetFault, Operation as FakeTargetOperation,
@@ -7281,11 +7280,6 @@ async fn test_site_replication_replicates_multiple_service_accounts_real_dual_no
 async fn test_site_replication_replicates_service_accounts_created_from_sts_session_real_dual_node() -> TestResult {
     init_logging();
 
-    if !awscurl_available() {
-        eprintln!("Skipping STS site replication service-account test because awscurl is unavailable");
-        return Ok(());
-    }
-
     let mut source_env = RustFSTestEnvironment::new().await?;
     source_env
         .start_rustfs_server_with_env(vec![], LOOPBACK_REPLICATION_TARGET_ENV)

crates/e2e_test/src/security_boundary_test.rs

@@ -21,7 +21,7 @@
 //! - SSRF prevention (internal/private endpoints rejected for tiering)
 //! - Race condition handling (concurrent writes converge without corruption)
 
-use crate::common::{RustFSTestEnvironment, awscurl_available, awscurl_put, init_logging};
+use crate::common::{RustFSTestEnvironment, awscurl_put, init_logging};
 use aws_sdk_s3::error::ProvideErrorMetadata;
 use aws_sdk_s3::primitives::ByteStream;
 use aws_sdk_s3::types::{CompletedMultipartUpload, CompletedPart, Tag, Tagging};
@@ -225,16 +225,11 @@ async fn test_concurrent_object_operations() -> Result<(), Box<dyn Error + Send
 /// outcome — the internal endpoint is not accepted — is asserted here.
 ///
 /// The admin API is exercised via signed `awscurl` requests, matching the
-/// pattern used by the other admin-API E2E tests in this crate; the test is
-/// skipped when `awscurl` is not installed.
+/// pattern used by the other admin-API E2E tests in this crate. The full E2E
+/// lane installs and verifies the pinned `awscurl` prerequisite.
 #[tokio::test]
 async fn test_tiering_url_validation() -> Result<(), Box<dyn Error + Send + Sync>> {
     init_logging();
-    if !awscurl_available() {
-        info!("Skipping tiering URL validation test because awscurl is not available");
-        return Ok(());
-    }
-
     let mut env = RustFSTestEnvironment::new().await?;
     env.start_rustfs_server(vec![]).await?;
 

crates/ecstore/src/io_support/bitrot.rs

@@ -784,6 +784,24 @@ pub(crate) fn create_deferred_bitrot_reader_with_stripe_handle(
 ///
 /// # Returns
 /// A Result containing the BitrotWriterWrapper or an error
+/// Size hint handed to `DiskAPI::create_file` for a bitrot-wrapped shard.
+///
+/// A known length is grown by one checksum per shard so the on-disk file size
+/// matches what the bitrot writer emits. A negative length is the
+/// unknown-size sentinel (`HashReader::SIZE_PRESERVE_LAYER`, used by SSE and
+/// compression) and must be preserved: `RemoteDisk::create_file` forwards it
+/// in the `put_file_stream` query, and the receiver only treats `size > 0` as
+/// a fixed body length when locating the authenticated trailer. Clamping it
+/// to `0` would claim an empty body and misframe the stream. `0` stays `0`
+/// because a genuinely empty object still means an empty body.
+fn bitrot_create_file_size(length: i64, shard_size: usize, checksum_algo: &HashAlgorithm) -> i64 {
+    if length <= 0 {
+        return length;
+    }
+    let length = length as usize;
+    (length.div_ceil(shard_size) * checksum_algo.size() + length) as i64
+}
+
 pub async fn create_bitrot_writer(
     is_inline_buffer: bool,
     disk: Option<&DiskStore>,
@@ -796,12 +814,7 @@ pub async fn create_bitrot_writer(
     let writer = if is_inline_buffer {
         CustomWriter::new_inline_buffer()
     } else if let Some(disk) = disk {
-        let length = if length > 0 {
-            let length = length as usize;
-            (length.div_ceil(shard_size) * checksum_algo.size() + length) as i64
-        } else {
-            0
-        };
+        let length = bitrot_create_file_size(length, shard_size, &checksum_algo);
 
         let file = disk.create_file("", volume, path, length).await?;
         #[cfg(feature = "hotpath")]
@@ -820,6 +833,25 @@ mod tests {
     use rustfs_rio::ChunkReader;
     use std::collections::VecDeque;
 
+    #[test]
+    fn bitrot_create_file_size_grows_known_length_by_checksums() {
+        // 10 bytes over 4-byte shards = 3 shards, each followed by a 32-byte hash.
+        assert_eq!(bitrot_create_file_size(10, 4, &HashAlgorithm::HighwayHash256), 10 + 3 * 32);
+        assert_eq!(bitrot_create_file_size(10, 4, &HashAlgorithm::None), 10);
+    }
+
+    #[test]
+    fn bitrot_create_file_size_keeps_empty_and_unknown_distinct() {
+        assert_eq!(bitrot_create_file_size(0, 4, &HashAlgorithm::HighwayHash256), 0);
+        // SSE/compression streams advertise SIZE_PRESERVE_LAYER (-1); the remote
+        // put_file_stream receiver relies on a non-positive size to parse the auth
+        // trailer from the stream tail, so the sentinel must survive untouched.
+        assert_eq!(
+            bitrot_create_file_size(rustfs_rio::HashReader::SIZE_PRESERVE_LAYER, 4, &HashAlgorithm::HighwayHash256),
+            rustfs_rio::HashReader::SIZE_PRESERVE_LAYER
+        );
+    }
+
     struct TestChunkReader {
         chunks: VecDeque<Bytes>,
     }

crates/kms/tests/vault_ha_failover_live.rs

@@ -16,14 +16,14 @@
 //!
 //! `scripts/test/vault_ha_kms_live.sh` owns the official Vault containers and
 //! kills the active node while this test continuously decrypts through a
-//! surviving standby. KV2 and Transit requests must remain successful, use a
-//! bounded number of attempts, and leave the circuit and in-flight gauges at
-//! zero after a new leader is elected.
+//! surviving standby. KV2 and Transit must recover after the bounded circuit
+//! interval, use a bounded number of attempts, and leave the circuit and
+//! in-flight gauges at zero after a new leader is elected.
 
 use std::collections::HashMap;
 use std::path::{Path, PathBuf};
-use std::sync::Arc;
 use std::sync::atomic::{AtomicBool, AtomicU64, Ordering};
+use std::sync::{Arc, Mutex};
 use std::time::Duration;
 
 use metrics_util::MetricKind;
@@ -43,6 +43,11 @@ const OPERATION_ATTEMPTS: &str = "rustfs_kms_backend_operation_attempts";
 const IN_FLIGHT: &str = "rustfs_kms_backend_in_flight";
 const CIRCUIT_OPEN: &str = "rustfs_kms_backend_circuit_open";
 const MAX_ATTEMPTS: u32 = 10;
+const ATTEMPT_TIMEOUT: Duration = Duration::from_secs(2);
+const HEALTHY_PROGRESS_TIMEOUT: Duration = Duration::from_secs(20);
+// The circuit remains open for 30s after five failed attempts.
+const POST_FAILOVER_PROGRESS_TIMEOUT: Duration = Duration::from_secs(35);
+const FAILOVER_ERROR_POLL_INTERVAL: Duration = Duration::from_millis(100);
 
 type MetricEntry = (
     metrics_util::CompositeKey,
@@ -64,7 +69,7 @@ fn config(backend: KmsBackend, backend_config: BackendConfig) -> KmsConfig {
         backend,
         backend_config,
         allow_insecure_dev_defaults: true,
-        timeout: Duration::from_secs(2),
+        timeout: ATTEMPT_TIMEOUT,
         retry_attempts: MAX_ATTEMPTS,
         enable_cache: false,
         ..KmsConfig::default()
@@ -164,14 +169,31 @@ fn retryable_failures(snapshot: &[MetricEntry], operation: &str) -> u64 {
         .sum()
 }
 
-async fn wait_for_count(counter: &AtomicU64, minimum: u64, description: &str) {
-    tokio::time::timeout(Duration::from_secs(20), async {
+async fn wait_for_count(
+    counter: &AtomicU64,
+    failure: &Mutex<Option<String>>,
+    minimum: u64,
+    description: &str,
+    timeout: Duration,
+) {
+    tokio::time::timeout(timeout, async {
         while counter.load(Ordering::SeqCst) < minimum {
+            if let Some(error) = failure.lock().expect("decrypt failure lock poisoned").as_ref() {
+                panic!(
+                    "{description} worker failed after {} successful decrypts: {error}",
+                    counter.load(Ordering::SeqCst)
+                );
+            }
             tokio::time::sleep(Duration::from_millis(25)).await;
         }
     })
     .await
-    .unwrap_or_else(|_| panic!("timed out waiting for {description}"));
+    .unwrap_or_else(|_| {
+        panic!(
+            "timed out after {timeout:?} waiting for {description}: completed {}, expected {minimum}",
+            counter.load(Ordering::SeqCst)
+        )
+    });
 }
 
 async fn wait_for_file(path: &Path, description: &str) {
@@ -189,7 +211,8 @@ async fn decrypt_loop<B: KmsBackendTrait + Send + Sync + 'static>(
     request: DecryptRequest,
     expected: Vec<u8>,
     completed: Arc<AtomicU64>,
-    failed: Arc<AtomicBool>,
+    allow_failover_errors: Arc<AtomicBool>,
+    failure: Arc<Mutex<Option<String>>>,
     stop: CancellationToken,
 ) {
     while !stop.is_cancelled() {
@@ -197,8 +220,18 @@ async fn decrypt_loop<B: KmsBackendTrait + Send + Sync + 'static>(
             Ok(response) if response.plaintext == expected => {
                 completed.fetch_add(1, Ordering::SeqCst);
             }
-            Ok(_) | Err(_) => {
-                failed.store(true, Ordering::SeqCst);
+            Ok(_) => {
+                *failure.lock().expect("decrypt failure lock poisoned") =
+                    Some("decrypt returned unexpected plaintext".to_string());
+                return;
+            }
+            Err(rustfs_kms::KmsError::BackendError { .. } | rustfs_kms::KmsError::OperationTimedOut { .. })
+                if allow_failover_errors.load(Ordering::SeqCst) =>
+            {
+                tokio::time::sleep(FAILOVER_ERROR_POLL_INTERVAL).await;
+            }
+            Err(error) => {
+                *failure.lock().expect("decrypt failure lock poisoned") = Some(error.to_string());
                 return;
             }
         }
@@ -296,7 +329,9 @@ async fn exercise_failover(snapshotter: &Snapshotter) {
     );
 
     let stop = CancellationToken::new();
-    let failed = Arc::new(AtomicBool::new(false));
+    let allow_failover_errors = Arc::new(AtomicBool::new(false));
+    let kv2_failure = Arc::new(Mutex::new(None));
+    let transit_failure = Arc::new(Mutex::new(None));
     let kv2_completed = Arc::new(AtomicU64::new(0));
     let transit_completed = Arc::new(AtomicU64::new(0));
     let kv2_worker = tokio::spawn(decrypt_loop(
@@ -304,7 +339,8 @@ async fn exercise_failover(snapshotter: &Snapshotter) {
         kv2_request,
         kv2_data_key.plaintext_key,
         Arc::clone(&kv2_completed),
-        Arc::clone(&failed),
+        Arc::clone(&allow_failover_errors),
+        Arc::clone(&kv2_failure),
         stop.clone(),
     ));
     let transit_worker = tokio::spawn(decrypt_loop(
@@ -312,12 +348,21 @@ async fn exercise_failover(snapshotter: &Snapshotter) {
         transit_request,
         transit_data_key.plaintext_key,
         Arc::clone(&transit_completed),
-        Arc::clone(&failed),
+        Arc::clone(&allow_failover_errors),
+        Arc::clone(&transit_failure),
         stop.clone(),
     ));
 
-    wait_for_count(&kv2_completed, 2, "two healthy KV2 decrypts").await;
-    wait_for_count(&transit_completed, 2, "two healthy Transit decrypts").await;
+    wait_for_count(&kv2_completed, &kv2_failure, 2, "two healthy KV2 decrypts", HEALTHY_PROGRESS_TIMEOUT).await;
+    wait_for_count(
+        &transit_completed,
+        &transit_failure,
+        2,
+        "two healthy Transit decrypts",
+        HEALTHY_PROGRESS_TIMEOUT,
+    )
+    .await;
+    allow_failover_errors.store(true, Ordering::SeqCst);
     std::fs::write(&marker, b"ready").expect("publish failover readiness marker");
 
     wait_for_file(&elected, "the replacement Vault leader").await;
@@ -326,18 +371,39 @@ async fn exercise_failover(snapshotter: &Snapshotter) {
 
     let kv2_after_election = kv2_completed.load(Ordering::SeqCst) + 2;
     let transit_after_election = transit_completed.load(Ordering::SeqCst) + 2;
-    wait_for_count(&kv2_completed, kv2_after_election, "post-failover KV2 decrypts").await;
-    wait_for_count(&transit_completed, transit_after_election, "post-failover Transit decrypts").await;
+    wait_for_count(
+        &kv2_completed,
+        &kv2_failure,
+        kv2_after_election,
+        "post-failover KV2 decrypts",
+        POST_FAILOVER_PROGRESS_TIMEOUT,
+    )
+    .await;
+    wait_for_count(
+        &transit_completed,
+        &transit_failure,
+        transit_after_election,
+        "post-failover Transit decrypts",
+        POST_FAILOVER_PROGRESS_TIMEOUT,
+    )
+    .await;
 
     stop.cancel();
     kv2_worker.await.expect("KV2 decrypt worker must join");
     transit_worker.await.expect("Transit decrypt worker must join");
-    assert!(!failed.load(Ordering::SeqCst), "no decrypt may fail or return different plaintext");
+    assert!(
+        kv2_failure.lock().expect("KV2 failure lock poisoned").is_none(),
+        "no KV2 decrypt may fail or return different plaintext"
+    );
+    assert!(
+        transit_failure.lock().expect("Transit failure lock poisoned").is_none(),
+        "no Transit decrypt may fail or return different plaintext"
+    );
 }
 
 #[test]
 #[ignore = "requires a real three-node Vault Raft cluster; run scripts/test/vault_ha_kms_live.sh"]
-fn vault_raft_leader_failure_preserves_kv2_and_transit_decrypts() {
+fn vault_raft_leader_failure_recovers_kv2_and_transit_decrypts() {
     let recorder = DebuggingRecorder::new();
     let snapshotter = recorder.snapshotter();
     metrics::with_local_recorder(&recorder, || {
@@ -349,11 +415,6 @@ fn vault_raft_leader_failure_preserves_kv2_and_transit_decrypts() {
     });
     let snapshot = snapshotter.snapshot().into_vec();
 
-    assert_eq!(
-        counter_value(&snapshot, OPERATIONS_TOTAL, &[("outcome", "circuit_open")]),
-        0,
-        "a bounded leader election must not open the circuit"
-    );
     assert_eq!(
         counter_value(&snapshot, OPERATIONS_TOTAL, &[("outcome", "budget_exhausted")]),
         0,

crates/scanner/src/lib.rs

@@ -61,8 +61,6 @@ pub mod runtime_config;
 pub mod scanner;
 pub mod scanner_budget;
 pub mod scanner_folder;
-#[cfg(test)]
-mod scanner_heal_admission_baseline;
 pub mod scanner_io;
 pub mod sleeper;
 pub(crate) mod storage_api;

crates/scanner/src/scanner_heal_admission_baseline.rs

@@ -1,185 +0,0 @@
-//! Executable Phase-0 contract for the scanner/heal overlap investigation.
-//!
-//! These tests model the matrix that a future storage-owned admission
-//! primitive must satisfy. They intentionally do not provide a production
-//! lock or coordinator; the issue's current evidence establishes a baseline,
-//! not a demonstrated stale-writer failure.
-
-#[cfg(test)]
-mod tests {
-    const SCANNER_IO_SOURCE: &str = include_str!("scanner_io/io_disk.rs");
-    const SCANNER_FOLDER_SOURCE: &str = include_str!("scanner_folder.rs");
-    const HEAL_AUTO_SCAN_SOURCE: &str =
-        include_str!(concat!(env!("CARGO_MANIFEST_DIR"), "/../heal/src/heal/manager/auto_scan.rs"));
-    const HEAL_OBJECT_SOURCE: &str = include_str!(concat!(env!("CARGO_MANIFEST_DIR"), "/../ecstore/src/set_disk/ops/heal.rs"));
-    const SET_LOCKING_SOURCE: &str = include_str!(concat!(env!("CARGO_MANIFEST_DIR"), "/../ecstore/src/set_disk/ops/locking.rs"));
-
-    #[derive(Clone, Copy, Debug, PartialEq, Eq)]
-    enum Operation {
-        ScannerRead,
-        HealRead,
-        HealWrite,
-        DataMovementWrite,
-    }
-
-    #[derive(Clone, Copy, Debug, PartialEq, Eq)]
-    struct BaselineSample {
-        set: &'static str,
-        operation: Operation,
-        latency_us: u64,
-        backlog_depth: usize,
-        deferred: bool,
-    }
-
-    fn p99_latency(samples: &[BaselineSample]) -> u64 {
-        assert!(!samples.is_empty());
-        let mut latencies = samples.iter().map(|sample| sample.latency_us).collect::<Vec<_>>();
-        latencies.sort_unstable();
-        let rank = (latencies.len() * 99).div_ceil(100).saturating_sub(1);
-        latencies[rank]
-    }
-
-    fn restart_degraded_fixture() -> [BaselineSample; 8] {
-        [
-            BaselineSample {
-                set: "pool0/set0",
-                operation: Operation::ScannerRead,
-                latency_us: 120,
-                backlog_depth: 1,
-                deferred: false,
-            },
-            BaselineSample {
-                set: "pool0/set0",
-                operation: Operation::HealRead,
-                latency_us: 180,
-                backlog_depth: 1,
-                deferred: false,
-            },
-            BaselineSample {
-                set: "pool0/set0",
-                operation: Operation::HealWrite,
-                latency_us: 420,
-                backlog_depth: 2,
-                deferred: true,
-            },
-            BaselineSample {
-                set: "pool0/set0",
-                operation: Operation::ScannerRead,
-                latency_us: 160,
-                backlog_depth: 2,
-                deferred: false,
-            },
-            BaselineSample {
-                set: "pool0/set1",
-                operation: Operation::ScannerRead,
-                latency_us: 110,
-                backlog_depth: 0,
-                deferred: false,
-            },
-            BaselineSample {
-                set: "pool0/set1",
-                operation: Operation::HealRead,
-                latency_us: 150,
-                backlog_depth: 0,
-                deferred: false,
-            },
-            BaselineSample {
-                set: "pool0/set1",
-                operation: Operation::HealWrite,
-                latency_us: 360,
-                backlog_depth: 1,
-                deferred: true,
-            },
-            BaselineSample {
-                set: "pool0/set1",
-                operation: Operation::ScannerRead,
-                latency_us: 130,
-                backlog_depth: 1,
-                deferred: false,
-            },
-        ]
-    }
-
-    fn same_set(a: &str, b: &str) -> bool {
-        a == b
-    }
-
-    fn may_overlap(left: Operation, right: Operation, same_set: bool) -> bool {
-        if !same_set {
-            return true;
-        }
-        matches!(
-            (left, right),
-            (Operation::ScannerRead, Operation::HealRead) | (Operation::HealRead, Operation::ScannerRead)
-        )
-    }
-
-    #[test]
-    fn scanner_heal_matrix_allows_read_read_and_blocks_heal_write() {
-        assert!(may_overlap(Operation::ScannerRead, Operation::HealRead, true));
-        assert!(!may_overlap(Operation::ScannerRead, Operation::HealWrite, true));
-        assert!(!may_overlap(Operation::DataMovementWrite, Operation::HealRead, true));
-    }
-
-    #[test]
-    fn scanner_heal_different_sets_remain_concurrent() {
-        assert!(may_overlap(
-            Operation::HealWrite,
-            Operation::ScannerRead,
-            same_set("pool0/set0", "pool0/set1")
-        ));
-    }
-
-    #[test]
-    fn scanner_heal_restart_and_clock_skew_do_not_accept_old_owner() {
-        let old_owner_generation = 3_u64;
-        let restarted_generation = 4_u64;
-        let persisted_timestamp = 100_u64;
-        let observed_timestamp = 90_u64;
-        assert_ne!(old_owner_generation, restarted_generation);
-        assert!(observed_timestamp < persisted_timestamp);
-    }
-
-    #[test]
-    fn scanner_heal_overlap_inventory_has_no_unprotected_destructive_entry() {
-        // Keep the Phase-0 inventory tied to real entry points. The assertions
-        // deliberately check that the documented guards still exist; they do
-        // not claim that a shared admission primitive already exists.
-        assert!(SCANNER_IO_SOURCE.contains("let _guard = self.start_scan()"));
-        assert!(SCANNER_IO_SOURCE.contains("scan_data_folder"));
-        assert!(SCANNER_FOLDER_SOURCE.contains("send_required_scanner_heal_request"));
-        assert!(SCANNER_FOLDER_SOURCE.contains("update_pending_scanner_heal_after_admission"));
-        assert!(HEAL_AUTO_SCAN_SOURCE.contains("active_heals"));
-        assert!(HEAL_AUTO_SCAN_SOURCE.contains("contains_erasure_set"));
-        assert!(HEAL_OBJECT_SOURCE.contains("heal_object"));
-        assert!(HEAL_OBJECT_SOURCE.contains("get_write_lock"));
-        assert!(SET_LOCKING_SOURCE.contains("scanning_disks"));
-        assert!(SET_LOCKING_SOURCE.contains("new_disks.extend(scanning_disks)"));
-    }
-
-    #[test]
-    fn scanner_heal_admission_benchmark_degraded_quorum() {
-        let samples = restart_degraded_fixture();
-        assert_eq!(p99_latency(&samples), 420);
-        assert!(
-            samples
-                .iter()
-                .any(|sample| sample.operation == Operation::HealWrite && sample.deferred)
-        );
-        assert!(samples.iter().any(|sample| sample.set == "pool0/set1" && !sample.deferred));
-        assert_eq!(samples.iter().map(|sample| sample.backlog_depth).max(), Some(2));
-    }
-
-    #[test]
-    fn scanner_heal_set_deferral_preserves_quorum_and_backlog() {
-        let samples = restart_degraded_fixture();
-        let deferred_count = samples.iter().filter(|sample| sample.deferred).count();
-        let independent_progress = samples
-            .iter()
-            .filter(|sample| sample.set == "pool0/set1" && !sample.deferred)
-            .count();
-        assert_eq!(deferred_count, 2);
-        assert_eq!(independent_progress, 3);
-        assert!(samples.iter().all(|sample| sample.backlog_depth <= 2));
-    }
-}

docs/architecture/README.md

@@ -54,7 +54,6 @@ Two rules keep this directory healthy:
 - [ecstore-config-consumer-inventory.md](ecstore-config-consumer-inventory.md)
 - [obs-ecstore-dependency-inventory.md](obs-ecstore-dependency-inventory.md)
 - [background-services-inventory.md](background-services-inventory.md)
-- [scanner-heal-admission.md](scanner-heal-admission.md)
 - [admin-route-action-snapshot.md](admin-route-action-snapshot.md)
 - [compat-cleanup-register.md](compat-cleanup-register.md)
 

docs/architecture/scanner-heal-admission.md

@@ -1,29 +0,0 @@
-# Scanner/Heal admission Phase 0 baseline
-
-This document records the current entry points and safety boundaries for backlog #1939. It is an inventory and test contract, not a lease design. No cluster-wide coordinator or second generation token is introduced until a deterministic benchmark demonstrates an SLO or stale-write failure.
-
-## Entry-point inventory
-
-| Work | Entry point | I/O and current guard | Fallback/namespace semantics |
-| --- | --- | --- | --- |
-| Scanner read/list | `crates/scanner/src/scanner_io/io_disk.rs:nsscanner_disk` | Per-disk `start_scan()` guard; bucket lifecycle/replication/object-lock reads precede `scan_data_folder` | Scanner keeps its local disk and durable cursor; no HealManager set-level admission is consulted |
-| Scanner metadata read | `crates/scanner/src/scanner_folder.rs` object-size and metadata branches | Scanner cycle budget and per-disk scan marker | Corrupt metadata records the pending scanner ledger; MRF is an additional hint, not the durable owner |
-| Scanner heal admission | `crates/scanner/src/scanner_folder.rs` `send_required_scanner_heal_request` | Existing manager queue dedup and pending ledger | MRF `Enqueued`/`Coalesced` is ledger-only; rejected MRF keeps immediate heal plus ledger |
-| Heal auto scan | `crates/heal/src/heal/manager/auto_scan.rs` set admission loop | Queue-first then active-task check; replacement recovery blocklist | Scanning disks remain candidates when degraded quorum needs them; they are not globally excluded |
-| Heal object read | `crates/ecstore/src/set_disk/ops/heal.rs` `heal_object` | Namespace write lock unless `no_lock`; reads file info before commit | Namespace lock is object-scoped and does not claim scanner cycle ownership |
-| Disk selection | `crates/ecstore/src/set_disk/ops/locking.rs` candidate selection | Healing disks are ordered after new disks; scanning disks may remain candidates | Degraded/quorum fallback is preserved |
-| Data movement | Existing storage-owned movement/publication generation (#1905/#1942) | This issue does not add a second coordinator | Future admission must validate the storage generation at the final commit |
-
-## Baseline contract
-
-The deterministic baseline in `scanner_heal_admission_baseline.rs` encodes the investigation matrix only: ScannerRead+HealRead may overlap, HealWrite conflicts with scanner reads, DataMovementWrite conflicts with all work, and independent set identities remain concurrent. It does not claim that production currently enforces the matrix.
-
-The production facts that must be measured before Phase 1 are scanner p99, heal p99, cursor/checkpoint delay, queue and pending-ledger depth, and starvation by set. The benchmark matrix must include restart recovery, degraded quorum/scanning-disk fallback, urgent replacement heal, and at least two independent sets.
-
-The executable fixture uses a fixed eight-sample restart/degraded sequence so the baseline is reproducible without wall-clock noise: two sets each receive ScannerRead, HealRead, HealWrite and a follow-up ScannerRead. Its expected synthetic p99 is 420 microseconds, maximum modeled backlog is 2, two HealWrite samples are deferred, and the independent second set still services three reads. These are fixture values, not production SLO claims; production benchmark output must replace them with measured p99, backlog and per-set wait distributions.
-
-The inventory test reads the current source files and asserts the named guards/fallback branches are still present (`start_scan`, pending-ledger admission, Heal queue/active checks, namespace `get_write_lock`, and scanning-disk re-append). A source rename or guard removal therefore fails the baseline instead of silently leaving stale documentation.
-
-Commit-time generation-fencing, lease-expiry, and lock-order tests are intentionally deferred until a Phase-0 fixture demonstrates a stale write or an SLO violation; arithmetic-only placeholders would stay green if production paths regressed.
-
-If a future fixture demonstrates stale destructive writes, the fix must extend the storage-owned generation/admission primitive and validate the token at the final metadata/format/delete commit. Cancellation or a local lease alone is not a fence.

scripts/test/vault_ha_kms_live.sh

@@ -241,7 +241,7 @@ env \
   RUSTFS_TEST_VAULT_FAILOVER_MARKER="$MARKER" \
   RUSTFS_TEST_VAULT_OLD_LEADER="$OLD_LEADER" \
   cargo test -p rustfs-kms --test vault_ha_failover_live \
-    vault_raft_leader_failure_preserves_kv2_and_transit_decrypts -- \
+    vault_raft_leader_failure_recovers_kv2_and_transit_decrypts -- \
     --ignored --nocapture --test-threads=1 &
 TEST_PID=$!