gsadmin/rustfs
1
0
Fork 0
mirror of https://github.com/rustfs/rustfs.git synced 2026-07-29 09:38:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(replication): classify resync verification HEAD errors instead of miscounting (backlog#862) (#4312)

Browse Source 
The resync result verification HEADed the target after replicating and counted
the outcome with inverted error handling:

- for a delete marker, ANY HEAD error (timeout, 5xx, auth, malformed) was
  counted as replicated (success);
- for a versioned object against an AWS-style target, HEAD was sent with the
  RustFS UUID versionId, which AWS rejects with 400, so a well-replicated
  object was counted as failed.

Classify the error before counting:

- delete marker: only a definitive 404/NoSuchKey or 405/MethodNotAllowed
  confirms the marker propagated (`is_retryable_delete_replication_head_error`
  == false); any retryable/ambiguous error now counts as failed;
- versioned object with a version-id-format rejection: re-verify via
  `head_object_fallback` (versionId-less HEAD) before deciding — present ->
  replicated, absent/error -> failed;
- all other errors: failed, as before.

Reuses the existing, unit-tested classifier helpers. Verified against the
existing resyncer suite (24 tests).

Refs backlog#799 (B13).
This commit is contained in:
Zhengchao An
2026-07-06 21:20:41 +08:00
committed by GitHub
parent 32845e03b0
commit 650a3e5734
1 changed files with 51 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/ecstore/src/bucket/replication/replication_resyncer.rs
+51 -9
View File
@@ -708,17 +708,59 @@ impl ReplicationResyncer {
roi.version_id.map(|v| v.to_string()),
)
.await;
let (size, err) = if let Err(err) = head_result {
if roi.delete_marker {
let (size, err) = match head_result {
Ok(_) => {
st.replicated_count += 1;
} else {
st.failed_count += 1;
st.replicated_size += roi.size;
(roi.size, None)
}
Err(err) if roi.delete_marker => {
// Verifying a replicated delete marker: only a
// definitive 404/NoSuchKey or 405/MethodNotAllowed
// confirms the marker propagated. Any other
// (retryable/ambiguous) HEAD error leaves the outcome
// unverified, so it must count as failed — not as a
// blanket success (backlog#862 / #799 B13).
let retryable = {
let (is_not_found, code) = err
.as_service_error()
.map(|se| (se.is_not_found(), se.code()))
.unwrap_or((false, None));
is_retryable_delete_replication_head_error(is_not_found, code)
};
if retryable {
st.failed_count += 1;
(0, Some(err))
} else {
st.replicated_count += 1;
(0, None)
}
}
Err(err) if is_version_id_format_mismatch(&err) => {
// AWS-style target rejects the RustFS UUID versionId
// (400). Re-verify without the versionId before
// concluding the object failed to replicate, instead
// of counting a well-replicated object as failed.
match head_object_fallback(&bucket_name, target_client.as_ref(), &roi.name).await {
Ok(Some(_)) => {
st.replicated_count += 1;
st.replicated_size += roi.size;
(roi.size, None)
}
Ok(None) => {
st.failed_count += 1;
(0, Some(err))
}
Err(e2) => {
st.failed_count += 1;
(0, Some(e2))
}
}
}
Err(err) => {
st.failed_count += 1;
(0, Some(err))
}
(0, Some(err))
} else {
st.replicated_count += 1;
st.replicated_size += roi.size;
(roi.size, None)
};
if err.is_some() {