fix(s3): round-trip null-version delete-marker identity (#6765)

* fix(s3): round-trip null-version delete-marker identity through listing and delete responses

On a versioning-suspended bucket, a null delete marker's identity was lost on the way back to the client at three points (issue #6745): ListObjectVersions advertised the marker's VersionId as the literal nil UUID instead of null; deleting by that id succeeded but the DeleteObjects/DeleteObject response reported the identity as null with no way to correlate it to the request; and the response lacked DeleteMarker/DeleteMarkerVersionId because the marker-ness comparison mixed the client-facing identity (Some(nil)) with the storage identity (None), so the removal also mis-recorded accounting and fired DeleteMarkerCreated semantics on later paths.

- Listing (bucket_usecase, s3_api/bucket, build_list_versions_next_marker) now maps the synthesized nil UUID to the literal null everywhere it reaches the wire, and VersionMarker::parse folds a nil-UUID marker from older listings into VersionMarker::Null so pagination resumes correctly.
- delete_objects normalizes both sides of the marker-ness comparison via delete_file_info_version_id (matching the adjacent explicit_delete_marker admission check) and reports DeleteMarkerVersionId as null for an explicit null-marker removal.
- resolve_delete_version_state reports delete_marker for an explicit-version delete whose target is a delete marker even when the bucket is versioning-suspended, fixing x-amz-delete-marker on the single-object path.
- The DeleteObjects response entry echoes the version identity the request addressed for marker removals, marker-removal accounting no longer records a marker creation, and notification events fire DeleteMarkerCreated only for actual marker creation.

Fixes #6745

* fix(s3): keep null-marker removal write shape undeleted and report marker semantics response-side

The first cut marked the storage delete request deleted for a null-marker removal, which FileMeta::delete_version interprets as the suspended-bucket delete-mints-a-marker write and re-creates the marker just removed. Carry marker-ness to responses via explicit_delete_removed_marker (single path) and a response-only branch flag (batch path) instead, keeping every storage write shape byte-identical to the pre-fix behavior. Adds an embedded end-to-end regression test covering the full issue #6745 round trip.
This commit is contained in:
Zhengchao An
2026-08-28 07:55:05 +08:00
committed by GitHub
parent db57fabcbd
commit 3c89c71f66
8 changed files with 449 additions and 23 deletions
+131
View File
@@ -21,6 +21,7 @@
use aws_sdk_s3::config::{Credentials, Region};
use aws_sdk_s3::primitives::ByteStream;
use aws_sdk_s3::types::{BucketVersioningStatus, Delete, ObjectIdentifier, VersioningConfiguration};
use aws_sdk_s3::{Client, Config};
use rustfs::embedded::{RustFSServerBuilder, find_available_port};
@@ -118,3 +119,133 @@ async fn test_embedded_server_basic_s3_operations_body() {
server.shutdown().await;
}
// Regression test for issue #6745: on a versioning-suspended bucket, a null
// delete marker's version identity must round-trip as the literal `null`
// through ListObjectVersions, DeleteObject, and DeleteObjects, and removing
// the marker by version id must carry the delete-marker flags.
#[test]
fn test_null_version_delete_marker_round_trip() {
common::run_embedded_test(test_null_version_delete_marker_round_trip_body);
}
async fn test_null_version_delete_marker_round_trip_body() {
let port = match find_available_port() {
Ok(port) => port,
Err(err) if err.kind() == std::io::ErrorKind::PermissionDenied => return,
Err(err) => panic!("find free port: {err}"),
};
let server = RustFSServerBuilder::new()
.address(format!("127.0.0.1:{port}"))
.access_key("testaccesskey")
.secret_key("testsecretkey")
.build()
.await
.expect("start embedded server");
let client = s3_client(&server.endpoint(), server.access_key(), server.secret_key());
let bucket = "null-marker-bucket";
client.create_bucket().bucket(bucket).send().await.expect("create bucket");
for status in [BucketVersioningStatus::Enabled, BucketVersioningStatus::Suspended] {
client
.put_bucket_versioning()
.bucket(bucket)
.versioning_configuration(VersioningConfiguration::builder().status(status).build())
.send()
.await
.expect("set bucket versioning state");
}
// A versionless DELETE on the suspended bucket mints a null delete marker
// and must report it as version `null`.
for key in ["doc/f1.txt", "doc/f2.txt"] {
client
.put_object()
.bucket(bucket)
.key(key)
.body(ByteStream::from_static(b"null-version payload"))
.send()
.await
.expect("put object");
let deleted = client
.delete_object()
.bucket(bucket)
.key(key)
.send()
.await
.expect("delete object without version id");
assert_eq!(deleted.delete_marker(), Some(true), "suspended-bucket delete should mint a marker");
assert_eq!(deleted.version_id(), Some("null"), "the minted marker is the null version");
}
// The markers must be listed under the literal `null`, never a nil UUID.
let listed = client
.list_object_versions()
.bucket(bucket)
.send()
.await
.expect("list object versions");
assert!(listed.versions().is_empty(), "the null versions were replaced by markers");
let markers = listed.delete_markers();
assert_eq!(markers.len(), 2);
for marker in markers {
assert_eq!(marker.version_id(), Some("null"), "listing must advertise the null version as `null`");
}
// Removing one marker by its listed version id over DeleteObject must
// acknowledge the marker identity on the wire.
let removed = client
.delete_object()
.bucket(bucket)
.key("doc/f1.txt")
.version_id("null")
.send()
.await
.expect("delete marker by null version id");
assert_eq!(
removed.delete_marker(),
Some(true),
"x-amz-delete-marker must be true for a marker removal"
);
assert_eq!(removed.version_id(), Some("null"));
// Removing the other via DeleteObjects must produce an entry a client can
// correlate with its request: same key, version `null`, marker flags set.
let delete = Delete::builder()
.objects(
ObjectIdentifier::builder()
.key("doc/f2.txt")
.version_id("null")
.build()
.expect("object identifier"),
)
.build()
.expect("delete payload");
let batch = client
.delete_objects()
.bucket(bucket)
.delete(delete)
.send()
.await
.expect("delete objects by null version id");
assert!(batch.errors().is_empty(), "batch delete reported errors: {:?}", batch.errors());
let entries = batch.deleted();
assert_eq!(entries.len(), 1);
assert_eq!(entries[0].key(), Some("doc/f2.txt"));
assert_eq!(entries[0].version_id(), Some("null"), "the entry must echo the requested identity");
assert_eq!(entries[0].delete_marker(), Some(true));
assert_eq!(entries[0].delete_marker_version_id(), Some("null"));
// With identity round-tripping, one pass leaves the bucket truly empty
// and deletable.
let after = client
.list_object_versions()
.bucket(bucket)
.send()
.await
.expect("list versions after cleanup");
assert!(after.versions().is_empty() && after.delete_markers().is_empty());
client.delete_bucket().bucket(bucket).send().await.expect("delete bucket");
server.shutdown().await;
}