fix(kms): enforce readiness probe deadline

7 call sites of wait_for_kms_ready() used ? in async test functions
that return () instead of Result. Changed to .expect("KMS ready").

crates/e2e_test/src/kms/bucket_default_encryption_test.rs

@@ -37,7 +37,7 @@ async fn test_bucket_default_sse_s3_put_object() -> Result<(), Box<dyn std::erro
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -159,7 +159,7 @@ async fn test_bucket_default_sse_kms_put_object() -> Result<(), Box<dyn std::err
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -278,7 +278,7 @@ async fn test_bucket_default_sse_kms_multipart_crc32() -> Result<(), Box<dyn std
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -475,7 +475,7 @@ async fn test_explicit_encryption_overrides_bucket_default() -> Result<(), Box<d
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -570,7 +570,7 @@ async fn test_sse_kms_without_key_id_populates_default() -> Result<(), Box<dyn s
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/e2e_test/src/kms/common.rs

@@ -189,34 +189,69 @@ pub async fn wait_for_kms_ready(
     access_key: &str,
     secret_key: &str,
 ) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
-    let total_deadline = Duration::from_secs(5);
+    wait_for_kms_ready_with_timeout(base_url, access_key, secret_key, Duration::from_secs(5)).await
+}
+
+async fn wait_for_kms_ready_with_timeout(
+    base_url: &str,
+    access_key: &str,
+    secret_key: &str,
+    total_deadline: Duration,
+) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     let start = tokio::time::Instant::now();
+    let deadline = start + total_deadline;
     let mut backoff = Duration::from_millis(200);
     let max_backoff = Duration::from_secs(1);
-    let mut first_attempt = true;
 
     loop {
-        if !first_attempt {
-            if start.elapsed() >= total_deadline {
-                return Err("KMS failed to become ready within 5 seconds".into());
-            }
-            sleep(backoff).await;
-            backoff = (backoff * 2).min(max_backoff);
-        }
-        first_attempt = false;
-
-        match get_kms_status(base_url, access_key, secret_key).await {
-            Ok(status) => {
+        match tokio::time::timeout_at(deadline, get_kms_status(base_url, access_key, secret_key)).await {
+            Ok(Ok(status)) => {
                 info!("KMS is ready (status: {})", status);
                 return Ok(());
             }
-            Err(e) => {
-                if start.elapsed() >= total_deadline {
-                    return Err(format!("KMS did not become ready within 5 s: last error: {e}").into());
-                }
-                warn!(error = %e, elapsed_ms = start.elapsed().as_millis() as u64, "KMS not ready yet, retrying…");
+            Ok(Err(e)) => {
+                let elapsed_ms = u64::try_from(start.elapsed().as_millis()).unwrap_or(u64::MAX);
+                warn!(error = %e, elapsed_ms, "KMS not ready yet, retrying…");
             }
+            Err(_) => return Err(format!("KMS failed to become ready within {} ms", total_deadline.as_millis()).into()),
         }
+
+        let now = tokio::time::Instant::now();
+        if now >= deadline {
+            return Err(format!("KMS failed to become ready within {} ms", total_deadline.as_millis()).into());
+        }
+        sleep((now + backoff).min(deadline) - now).await;
+        backoff = (backoff * 2).min(max_backoff);
+    }
+}
+
+#[cfg(test)]
+mod readiness_tests {
+    use super::wait_for_kms_ready_with_timeout;
+    use std::time::Duration;
+    use tokio::io::AsyncReadExt;
+    use tokio::net::TcpListener;
+
+    #[tokio::test]
+    async fn kms_readiness_deadline_covers_a_stalled_status_request() {
+        let listener = TcpListener::bind("127.0.0.1:0").await.expect("bind readiness test server");
+        let address = listener.local_addr().expect("read readiness test server address");
+        let server = tokio::spawn(async move {
+            let (mut socket, _) = listener.accept().await.expect("accept readiness request");
+            let mut request = [0_u8; 1024];
+            let _ = socket.read(&mut request).await.expect("read readiness request");
+            std::future::pending::<()>().await;
+        });
+
+        let result = tokio::time::timeout(
+            Duration::from_secs(1),
+            wait_for_kms_ready_with_timeout(&format!("http://{address}"), "access-key", "secret-key", Duration::from_millis(50)),
+        )
+        .await
+        .expect("readiness helper must enforce its own deadline");
+
+        server.abort();
+        assert!(result.is_err(), "a stalled status request must not outlive the readiness deadline");
     }
 }
 

crates/e2e_test/src/kms/copy_object_self_copy_sse_test.rs

@@ -61,7 +61,7 @@ async fn test_metadata_replace_self_copy_of_sse_object_stays_decryptable() {
         )
         .await
         .expect("failed to start RustFS with local KMS");
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     let client = kms_env.base_env.create_s3_client();
     // Deliberately an UNVERSIONED bucket: that is the branch where the store layer can service
@@ -160,7 +160,7 @@ async fn test_metadata_replace_self_copy_dropping_sse_rewrites_plaintext() {
         )
         .await
         .expect("failed to start RustFS with local KMS");
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     let client = kms_env.base_env.create_s3_client();
     // Unversioned, and deliberately WITHOUT a bucket default-encryption rule, so the copy below
@@ -256,7 +256,7 @@ async fn test_metadata_replace_self_copy_under_bucket_default_sse_stays_decrypta
         )
         .await
         .expect("failed to start RustFS with local KMS");
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     let client = kms_env.base_env.create_s3_client();
     let bucket = "copy-object-self-copy-bucket-default-sse-test";

crates/e2e_test/src/kms/copy_object_version_restore_sse_test.rs

@@ -56,7 +56,7 @@ async fn test_self_copy_of_historical_sse_s3_version_is_readable() {
         )
         .await
         .expect("failed to start RustFS with local KMS");
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     let client = kms_env.base_env.create_s3_client();
     let bucket = "copy-object-version-restore-sse-test";

crates/e2e_test/src/kms/encryption_metadata_test.rs

@@ -87,7 +87,7 @@ async fn test_head_reports_managed_metadata_for_sse_s3() -> Result<(), Box<dyn s
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -147,7 +147,7 @@ async fn test_head_reports_managed_metadata_for_sse_kms_and_copy() -> Result<(),
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -250,7 +250,7 @@ async fn test_multipart_upload_writes_encrypted_data() -> Result<(), Box<dyn std
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/e2e_test/src/kms/kms_comprehensive_test.rs

@@ -24,7 +24,6 @@ use super::common::{
     test_sse_kms_encryption, test_sse_s3_encryption,
 };
 use crate::common::{TEST_BUCKET, init_logging};
-use tokio::time::{Duration, sleep};
 use tracing::info;
 
 /// Comprehensive test: Full KMS workflow with all encryption types
@@ -35,7 +34,7 @@ async fn test_comprehensive_kms_full_workflow() -> Result<(), Box<dyn std::error
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -103,7 +102,7 @@ async fn test_comprehensive_stress_test() -> Result<(), Box<dyn std::error::Erro
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -137,7 +136,7 @@ async fn test_comprehensive_key_isolation() -> Result<(), Box<dyn std::error::Er
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -208,7 +207,7 @@ async fn test_comprehensive_concurrent_operations() -> Result<(), Box<dyn std::e
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -253,7 +252,7 @@ async fn test_comprehensive_performance_benchmark() -> Result<(), Box<dyn std::e
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/e2e_test/src/kms/kms_edge_cases_test.rs

@@ -44,7 +44,7 @@ async fn test_kms_zero_byte_file_encryption() -> Result<(), Box<dyn std::error::
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -117,7 +117,7 @@ async fn test_kms_single_byte_file_encryption() -> Result<(), Box<dyn std::error
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -209,7 +209,7 @@ async fn test_kms_multipart_boundary_conditions() -> Result<(), Box<dyn std::err
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -284,7 +284,7 @@ async fn test_kms_invalid_key_scenarios() -> Result<(), Box<dyn std::error::Erro
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -371,7 +371,7 @@ async fn test_kms_concurrent_encryption() -> Result<(), Box<dyn std::error::Erro
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = Arc::new(kms_env.base_env.create_s3_client());
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -478,7 +478,7 @@ async fn test_kms_key_validation_security() -> Result<(), Box<dyn std::error::Er
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/e2e_test/src/kms/kms_fault_recovery_test.rs

@@ -37,7 +37,7 @@ async fn test_kms_key_directory_unavailable() -> Result<(), Box<dyn std::error::
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -127,7 +127,7 @@ async fn test_kms_corrupted_key_files() -> Result<(), Box<dyn std::error::Error
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -218,7 +218,7 @@ async fn test_kms_multipart_upload_interruption() -> Result<(), Box<dyn std::err
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -401,7 +401,7 @@ async fn test_kms_resource_constraints() -> Result<(), Box<dyn std::error::Error
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/e2e_test/src/kms/kms_local_test.rs

@@ -46,7 +46,7 @@ async fn test_local_kms_end_to_end() -> Result<(), Box<dyn std::error::Error + S
         .expect("Failed to start RustFS with Local KMS");
 
     // Wait a moment for RustFS to fully start up and initialize KMS
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     info!("RustFS started with KMS auto-configuration, default_key_id: {}", default_key_id);
 
@@ -127,7 +127,7 @@ async fn test_local_kms_key_isolation() {
         .expect("Failed to start RustFS with Local KMS");
 
     // Wait a moment for RustFS to fully start up and initialize KMS
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     info!("RustFS started with KMS auto-configuration, default_key_id: {}", default_key_id);
 
@@ -227,7 +227,7 @@ async fn test_local_kms_large_file() {
         .expect("Failed to start RustFS with Local KMS");
 
     // Wait a moment for RustFS to fully start up and initialize KMS
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     info!("RustFS started with KMS auto-configuration, default_key_id: {}", default_key_id);
 
@@ -309,7 +309,7 @@ async fn test_local_kms_multipart_upload() {
         .expect("Failed to start RustFS with Local KMS");
 
     // Wait for KMS initialization
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await.expect("KMS ready");
 
     info!("RustFS started with KMS auto-configuration, default_key_id: {}", default_key_id);
 

crates/e2e_test/src/kms/kms_vault_test.rs

@@ -19,7 +19,6 @@
 //! multipart upload behaviour.
 
 use crate::common::{TEST_BUCKET, init_logging};
-use tokio::time::{Duration, sleep};
 use tracing::{error, info};
 
 use super::common::{
@@ -45,8 +44,8 @@ impl VaultKmsTestContext {
 
         start_kms(&env.base_env.url, &env.base_env.access_key, &env.base_env.secret_key).await?;
 
-        // Allow Vault to finish initialising token auth and transit engine.
-        sleep(Duration::from_secs(2)).await;
+        // Wait for KMS to finish initialising.
+        super::common::wait_for_kms_ready(&env.base_env.url, &env.base_env.access_key, &env.base_env.secret_key).await?;
 
         Ok(Self { env })
     }

crates/e2e_test/src/kms/multipart_encryption_test.rs

@@ -33,7 +33,7 @@ async fn test_step1_basic_single_file_encryption() -> Result<(), Box<dyn std::er
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -89,7 +89,7 @@ async fn test_step2_basic_multipart_upload_without_encryption() -> Result<(), Bo
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -187,7 +187,7 @@ async fn test_step3_multipart_upload_with_sse_s3() -> Result<(), Box<dyn std::er
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -310,7 +310,7 @@ async fn test_step4_large_multipart_upload_with_encryption() -> Result<(), Box<d
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;
@@ -435,7 +435,7 @@ async fn test_step5_all_encryption_types_multipart() -> Result<(), Box<dyn std::
 
     let mut kms_env = LocalKMSTestEnvironment::new().await?;
     let _default_key_id = kms_env.start_rustfs_for_local_kms().await?;
-    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
+    kms_env.wait_for_kms_ready().await?;
 
     let s3_client = kms_env.base_env.create_s3_client();
     kms_env.base_env.create_test_bucket(TEST_BUCKET).await?;

crates/ecstore/src/core/pools.rs

@@ -36,8 +36,7 @@ use crate::disk::error::DiskError;
 use crate::disk::{BUCKET_META_PREFIX, RUSTFS_META_BUCKET};
 use crate::error::{Error, Result};
 use crate::error::{
-    StorageError, is_err_bucket_exists, is_err_bucket_not_found, is_err_object_not_found, is_err_operation_canceled,
-    is_err_version_not_found,
+    StorageError, is_err_bucket_exists, is_err_bucket_not_found, is_err_object_not_found, is_err_version_not_found,
 };
 use crate::layout::endpoints::EndpointServerPools;
 use crate::object_api::{GetObjectReader, ObjectOptions};
@@ -774,76 +773,7 @@ async fn load_decommission_entry_exact_versions(
 }
 
 fn resolve_decommission_check_after_list_result(list_result: Result<()>, entry_error: Option<Error>) -> Result<()> {
-    match list_result {
-        Ok(()) => entry_error.map_or(Ok(()), Err),
-        Err(list_err) => resolve_decommission_listing_error(Some(list_err), entry_error).map_or(Ok(()), Err),
-    }
-}
-
-fn resolve_decommission_listing_error(listing_error: Option<Error>, entry_error: Option<Error>) -> Option<Error> {
-    match (listing_error, entry_error) {
-        (Some(listing_error), Some(entry_error)) if is_err_operation_canceled(&listing_error) => Some(entry_error),
-        (Some(listing_error), Some(entry_error)) if is_err_operation_canceled(&entry_error) => Some(listing_error),
-        (Some(listing_error), _) => Some(listing_error),
-        (None, entry_error) => entry_error,
-    }
-}
-
-fn decommission_unresolved_listing_error(
-    bucket: &str,
-    prefix: &str,
-    candidate: Option<&str>,
-    candidate_count: usize,
-    disk_error_count: usize,
-    pool_index: usize,
-    set_index: usize,
-) -> Error {
-    let location = candidate.unwrap_or(prefix);
-    Error::other(format!(
-        "decommission listing could not resolve metadata for {bucket}/{location} on pool {pool_index} set {set_index} ({candidate_count} candidate(s), {disk_error_count} disk error(s))"
-    ))
-}
-
-fn resolve_decommission_partial_listing_entry(
-    entries: MetaCacheEntries,
-    resolver: MetadataResolutionParams,
-    bucket: &str,
-    prefix: &str,
-    disk_error_count: usize,
-    pool_index: usize,
-    set_index: usize,
-) -> Result<MetaCacheEntry> {
-    let candidate_count = entries.as_ref().iter().flatten().count();
-    if let Some(entry) = entries.resolve(resolver) {
-        return Ok(entry);
-    }
-
-    let candidate = entries.as_ref().iter().flatten().map(|entry| entry.name.as_str()).next();
-    Err(decommission_unresolved_listing_error(
-        bucket,
-        prefix,
-        candidate,
-        candidate_count,
-        disk_error_count,
-        pool_index,
-        set_index,
-    ))
-}
-
-async fn record_decommission_entry_error(
-    entry_error: &Arc<tokio::sync::Mutex<Option<Error>>>,
-    rx: &CancellationToken,
-    err: Error,
-) {
-    if rx.is_cancelled() {
-        return;
-    }
-
-    let mut first_err = entry_error.lock().await;
-    if first_err.is_none() && !rx.is_cancelled() {
-        *first_err = Some(err);
-        rx.cancel();
-    }
+    if let Some(err) = entry_error { Err(err) } else { list_result }
 }
 
 fn resolve_decommission_pool_meta_reload_result(result: Result<()>, stage: &str) -> Result<()> {
@@ -3608,7 +3538,6 @@ impl ECStore {
             let rx_clone = rx.clone();
             let bi = bi.clone();
             let set_id = set_idx;
-            let listing_entry_error = entry_error.clone();
             let worker = tokio::spawn(async move {
                 let _listing_permit = listing_permit;
                 run_decommission_listing_with_retry(
@@ -3622,11 +3551,7 @@ impl ECStore {
                         let set = set.clone();
                         let rx = rx_clone.clone();
                         let bucket = bi.clone();
-                        let entry_error = listing_entry_error.clone();
-                        async move {
-                            set.list_objects_to_decommission(rx, bucket, callback, entry_error.clone(), idx, set_id)
-                                .await
-                        }
+                        async move { set.list_objects_to_decommission(rx, bucket, callback).await }
                     },
                 )
                 .await
@@ -3656,7 +3581,11 @@ impl ECStore {
 
         wait_decommission_worker_drain(&workers, worker_limit).await?;
 
-        if let Some(err) = resolve_decommission_listing_error(listing_worker_error, entry_error.lock().await.clone()) {
+        if let Some(err) = listing_worker_error {
+            return Err(err);
+        }
+
+        if let Some(err) = entry_error.lock().await.clone() {
             return Err(err);
         }
 
@@ -4262,7 +4191,7 @@ impl ECStore {
         let buckets = self.get_buckets_to_decommission().await?;
         let pool = self.pools[idx].clone();
 
-        for (set_index, set) in pool.disk_set.iter().enumerate() {
+        for set in &pool.disk_set {
             for bucket_info in &buckets {
                 let mut lifecycle_config = None;
                 let mut object_lock_config = None;
@@ -4357,7 +4286,7 @@ impl ECStore {
                 });
 
                 let list_result = set
-                    .list_objects_to_decommission(callback_rx, bucket_info.clone(), callback, entry_error.clone(), idx, set_index)
+                    .list_objects_to_decommission(callback_rx, bucket_info.clone(), callback)
                     .await;
                 let entry_error = entry_error.lock().await.clone();
                 resolve_decommission_check_after_list_result(list_result, entry_error)?;
@@ -5092,15 +5021,12 @@ mod tests {
 pub type ListCallback = Arc<dyn Fn(MetaCacheEntry) -> BoxFuture<'static, ()> + Send + Sync + 'static>;
 
 impl SetDisks {
-    #[tracing::instrument(skip(self, rx, cb_func, entry_error))]
+    #[tracing::instrument(skip(self, rx, cb_func))]
     async fn list_objects_to_decommission(
         self: &Arc<Self>,
         rx: CancellationToken,
         bucket_info: DecomBucketInfo,
         cb_func: ListCallback,
-        entry_error: Arc<tokio::sync::Mutex<Option<Error>>>,
-        pool_index: usize,
-        set_index: usize,
     ) -> Result<()> {
         let (disks, _) = self.get_online_disks_with_healing(false).await;
         ensure_decommission_listing_disks_available(!disks.is_empty(), &bucket_info.name)?;
@@ -5115,12 +5041,6 @@ impl SetDisks {
         };
 
         let cb1 = cb_func.clone();
-        let unresolved_error = entry_error.clone();
-        let unresolved_rx = rx.clone();
-        let unresolved_bucket = bucket_info.name.clone();
-        let unresolved_prefix = bucket_info.prefix.clone();
-        let unresolved_pool_index = pool_index;
-        let unresolved_set_index = set_index;
 
         list_path_raw(
             rx,
@@ -5133,51 +5053,20 @@ impl SetDisks {
                 skip_walkdir_total_timeout: true,
                 walkdir_stall_timeout: Some(DECOMMISSION_BACKGROUND_WALKDIR_STALL_TIMEOUT),
                 agreed: Some(Box::new(move |entry: MetaCacheEntry| Box::pin(cb1(entry)))),
-                partial: Some(Box::new(move |entries: MetaCacheEntries, errs: &[Option<DiskError>]| {
+                partial: Some(Box::new(move |entries: MetaCacheEntries, _: &[Option<DiskError>]| {
                     let resolver = resolver.clone();
                     let cb_func = cb_func.clone();
-                    let bucket = unresolved_bucket.clone();
-                    let prefix = unresolved_prefix.clone();
-                    let unresolved_error = unresolved_error.clone();
-                    let unresolved_rx = unresolved_rx.clone();
-                    let pool_index = unresolved_pool_index;
-                    let set_index = unresolved_set_index;
-                    let disk_error_count = errs.iter().flatten().count();
-                    if unresolved_rx.is_cancelled() {
-                        return Box::pin(async {});
-                    }
-
-                    match resolve_decommission_partial_listing_entry(
-                        entries,
-                        resolver,
-                        &bucket,
-                        &prefix,
-                        disk_error_count,
-                        pool_index,
-                        set_index,
-                    ) {
-                        Ok(entry) => {
+                    match entries.resolve(resolver) {
+                        Some(entry) => {
                             warn!("decommission_pool: list_objects_to_decommission get {}", &entry.name);
                             Box::pin(async move {
                                 cb_func(entry).await;
                             })
                         }
-                        Err(err) => Box::pin(async move {
-                            if unresolved_rx.is_cancelled() {
-                                return;
-                            }
-                            warn!(
-                                event = EVENT_DECOMMISSION_BUCKET,
-                                component = LOG_COMPONENT_ECSTORE,
-                                subsystem = LOG_SUBSYSTEM_POOLS,
-                                bucket = %bucket,
-                                prefix = %prefix,
-                                state = "unresolved_entry",
-                                error = %err,
-                                "Decommission listing failed closed on unresolved metadata"
-                            );
-                            record_decommission_entry_error(&unresolved_error, &unresolved_rx, err).await;
-                        }),
+                        None => {
+                            warn!("decommission_pool: list_objects_to_decommission get none");
+                            Box::pin(async {})
+                        }
                     }
                 })),
                 ..Default::default()
@@ -5185,10 +5074,6 @@ impl SetDisks {
         )
         .await?;
 
-        if let Some(err) = entry_error.lock().await.clone() {
-            return Err(err);
-        }
-
         Ok(())
     }
 }
@@ -5394,12 +5279,11 @@ mod pools_tests {
         has_active_decommission_canceler, is_decommission_active, is_decommission_cancel_requested,
         load_decommission_entry_versions, local_decommission_queue_prefix, mark_decommission_bucket_done,
         merge_pool_status_refresh, missing_decommission_worker_prefix, observe_decommission_terminal_reload_result,
-        pool_meta_has_active_decommission, record_decommission_entry_error, require_decommission_store,
-        resolve_decommission_bucket_done_save_result, resolve_decommission_bucket_state,
-        resolve_decommission_check_after_list_result, resolve_decommission_entry_cleanup_delete_result,
-        resolve_decommission_entry_exact_versions, resolve_decommission_entry_reload_result, resolve_decommission_listing_error,
-        resolve_decommission_listing_worker_result, resolve_decommission_optional_bucket_config_result,
-        resolve_decommission_partial_listing_entry, resolve_decommission_pool_meta_reload_result,
+        pool_meta_has_active_decommission, require_decommission_store, resolve_decommission_bucket_done_save_result,
+        resolve_decommission_bucket_state, resolve_decommission_check_after_list_result,
+        resolve_decommission_entry_cleanup_delete_result, resolve_decommission_entry_exact_versions,
+        resolve_decommission_entry_reload_result, resolve_decommission_listing_worker_result,
+        resolve_decommission_optional_bucket_config_result, resolve_decommission_pool_meta_reload_result,
         resolve_decommission_preflight_heal_result, resolve_decommission_progress_save_result,
         resolve_decommission_spawn_failure_result, resolve_decommission_terminal_mark_after_error_result,
         resolve_decommission_terminal_mark_result, resolve_decommission_update_after_result,
@@ -5418,9 +5302,7 @@ mod pools_tests {
     use crate::error::{Error, StorageError};
     use crate::layout::endpoints::{EndpointServerPools, Endpoints, PoolEndpoints};
     use crate::services::rebalance::{RebalStatus, RebalanceInfo, RebalanceMeta, RebalanceStats};
-    use rustfs_filemeta::{
-        FileInfo, FileInfoVersions, MetaCacheEntries, MetaCacheEntry, MetadataResolutionParams, ObjectPartInfo,
-    };
+    use rustfs_filemeta::{FileInfo, FileInfoVersions, MetaCacheEntry, ObjectPartInfo};
     use rustfs_rio::Index;
     use std::sync::{
         Arc,
@@ -6439,65 +6321,6 @@ mod pools_tests {
         assert!(matches!(err, Error::SlowDown));
     }
 
-    #[test]
-    fn test_resolve_decommission_partial_listing_entry_rejects_unresolved_metadata() {
-        let err = resolve_decommission_partial_listing_entry(
-            MetaCacheEntries(vec![None]),
-            MetadataResolutionParams {
-                dir_quorum: 2,
-                obj_quorum: 2,
-                bucket: "bucket-a".to_string(),
-                ..Default::default()
-            },
-            "bucket-a",
-            "prefix/",
-            1,
-            2,
-            3,
-        )
-        .expect_err("unresolved partial listing must fail closed");
-
-        let message = err.to_string();
-        assert!(message.contains("decommission listing could not resolve metadata"));
-        assert!(message.contains("bucket-a/prefix/"));
-        assert!(message.contains("pool 2 set 3"));
-        assert!(message.contains("1 disk error(s)"));
-    }
-
-    #[tokio::test]
-    async fn test_record_decommission_entry_error_cancels_listing_and_preserves_first_error() {
-        let entry_error = Arc::new(tokio::sync::Mutex::new(None));
-        let rx = CancellationToken::new();
-
-        record_decommission_entry_error(&entry_error, &rx, Error::SlowDown).await;
-        record_decommission_entry_error(&entry_error, &rx, Error::OperationCanceled).await;
-
-        assert!(rx.is_cancelled());
-        assert!(matches!(*entry_error.lock().await, Some(Error::SlowDown)));
-    }
-
-    #[tokio::test]
-    async fn test_record_decommission_entry_error_ignores_already_canceled_listing() {
-        let entry_error = Arc::new(tokio::sync::Mutex::new(None));
-        let rx = CancellationToken::new();
-        rx.cancel();
-
-        record_decommission_entry_error(&entry_error, &rx, Error::SlowDown).await;
-
-        assert!(entry_error.lock().await.is_none());
-    }
-
-    #[test]
-    fn test_resolve_decommission_listing_error_preserves_real_listing_failure() {
-        let err = resolve_decommission_listing_error(Some(Error::SlowDown), Some(Error::OperationCanceled))
-            .expect("listing failure should be returned");
-        assert!(matches!(err, Error::SlowDown));
-
-        let err = resolve_decommission_listing_error(Some(Error::OperationCanceled), Some(Error::SlowDown))
-            .expect("entry failure should be returned");
-        assert!(matches!(err, Error::SlowDown));
-    }
-
     #[test]
     fn test_resolve_decommission_check_after_list_result_returns_list_result_without_entry_error() {
         let err = resolve_decommission_check_after_list_result(Err(Error::OperationCanceled), None)

scripts/check_test_wiring.py

@@ -206,13 +206,6 @@ def check_runner_selection(root: Path) -> list[str]:
     return errors
 
 
-def check_s3_tests_runner(root: Path) -> list[str]:
-    runner = (root / "scripts/s3-tests/run.sh").read_text()
-    if "--showlocals" in runner:
-        return ["scripts/s3-tests/run.sh: pytest failure diagnostics must not dump local values"]
-    return []
-
-
 def profile_selection(root: Path, profile: str) -> str:
     if not re.fullmatch(r"e2e-[a-z0-9-]+", profile):
         raise ValueError(f"invalid e2e profile name: {profile}")
@@ -279,7 +272,6 @@ def validate(root: Path) -> list[str]:
     errors.extend(check_e2e_modules(root))
     errors.extend(check_fuzz_targets(root))
     errors.extend(check_runner_selection(root))
-    errors.extend(check_s3_tests_runner(root))
     errors.extend(check_profile_definitions(root))
     return errors
 
@@ -349,23 +341,6 @@ class SelfTests(unittest.TestCase):
             )
             self.assertEqual(len(check_fuzz_targets(root)), 1)
 
-    def test_s3_runner_rejects_unbounded_failure_locals(self) -> None:
-        with tempfile.TemporaryDirectory() as tmp:
-            root = Path(tmp)
-            runner = root / "scripts/s3-tests/run.sh"
-            runner.parent.mkdir(parents=True)
-            runner.write_text("tox -- -vv -ra --tb=long\n")
-            self.assertEqual(check_s3_tests_runner(root), [])
-            runner.write_text("tox -- -vv -ra --showlocals --tb=long\n")
-            self.assertEqual(len(check_s3_tests_runner(root)), 1)
-            with (
-                mock.patch(__name__ + ".check_e2e_modules", return_value=[]),
-                mock.patch(__name__ + ".check_fuzz_targets", return_value=[]),
-                mock.patch(__name__ + ".check_runner_selection", return_value=[]),
-                mock.patch(__name__ + ".check_profile_definitions", return_value=[]),
-            ):
-                self.assertEqual(len(validate(root)), 1)
-
     def test_profile_listing_enforces_selection(self) -> None:
         with tempfile.TemporaryDirectory() as tmp:
             root = Path(tmp)
@@ -436,7 +411,7 @@ def main() -> int:
         for error in errors:
             print(f"ERROR: {error}", file=sys.stderr)
         return 1
-    print("OK: e2e modules, runner selection, fuzz matrices, profiles, and bounded diagnostics are wired")
+    print("OK: e2e modules, runner selection, fuzz matrices, and profile guards are wired")
     return 0
 
 

scripts/s3-tests/run.sh

@@ -1028,11 +1028,10 @@ else
 fi
 
 # Run tests from s3tests/functional
-# Failure locals can contain multi-MiB request bodies; keep tracebacks without expanding local values.
 set +e
 S3TEST_CONF="${CONF_OUTPUT_PATH}" \
     tox -- \
-    -vv -ra --tb=long \
+    -vv -ra --showlocals --tb=long \
     --maxfail="${MAXFAIL}" \
     --timeout="${TEST_TIMEOUT}" \
     --junitxml="${ARTIFACTS_DIR}/junit.xml" \