gsadmin/rustfs
1
0
Fork 0
mirror of https://github.com/rustfs/rustfs.git synced 2026-09-10 06:05:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

test(scanner): stabilize W13 release gate evidence (#7613)

Browse Source 
* feat(nightly): publish packages as assets of the rolling 'nightly' release (sync from release) (#7593)

feat(nightly): publish packages as assets of the rolling 'nightly' release (#7592)

Replace the assets-branch scheme with a proper GitHub Release on
rustfs/auto-testing: a single 'nightly' release whose deb/rpm assets
are replaced in place on every build. This is the standard channel —
visible on the repo's Releases page, stable download URLs, no git
history growth (release assets live outside the repository).

- New scripts/release/publish_nightly_assets.sh: resolves-or-creates
  the 'nightly' release via the REST API, deletes same-name assets,
  uploads rustfs-nightly-latest.{deb,rpm}, then PATCHes the release
  body with the build provenance (ref@sha, run link, sizes, SHA256).
  Plain curl + python3, no gh CLI (the build fleet has none — #7586).
- The workflow step shrinks to invoking the script; full flow
  exercised end-to-end against the real release with probe files
  (create / upload / overwrite / download round-trip / body update).

* test(scanner): stabilize W13 release gate evidence

Treat only real raw-entry windows as replayed scanner enumeration in the restart diagnostic, so final completion rounds without raw entries are not fail-closed as raw replays.

Allow the EC8:4 multi-pool heal evidence case to select an outage object key that routes to an online pool while an entire target pool is down, preserving strict behavior for single-pool cases.

Co-Authored-By: heihutu <heihutu@gmail.com>

Co-Authored-By: zhi22915 <qiuzgang@gmail.com>

---------

Co-authored-by: hector <42570491+majinghe@users.noreply.github.com>
Co-authored-by: zhi22915 <qiuzgang@gmail.com>
This commit is contained in:
houseme
2026-09-10 10:08:55 +08:00
committed by GitHub
parent 1643cb29f1
commit 54bd3c8e24
2 changed files with 56 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/e2e_test/src/heal_erasure_disk_rebuild_test.rs
+52 -21
View File
@@ -22,7 +22,11 @@ mod tests {
init_logging, rustfs_binary_path,
};
use crate::storage_api::RUSTFS_META_BUCKET;
use aws_sdk_s3::primitives::ByteStream;
use aws_sdk_s3::{
error::{ProvideErrorMetadata, SdkError},
operation::put_object::PutObjectError,
primitives::ByteStream,
};
use http::Method;
use sha2::{Digest, Sha256};
use std::collections::HashSet;
@@ -586,6 +590,10 @@ mod tests {
&& operations["activeBySource"]["admin"].as_u64() == Some(1)
}
fn is_service_unavailable_put(error: &SdkError<PutObjectError>) -> bool {
error.as_service_error().and_then(ProvideErrorMetadata::code) == Some("ServiceUnavailable")
}
async fn replacement_recovery_status(
cluster: &RustFSTestClusterEnvironment,
) -> Result<serde_json::Value, Box<dyn Error + Send + Sync>> {
@@ -1342,18 +1350,41 @@ mod tests {
"replacement target must retain only its preformatted topology identity"
);
let outage_key = "cluster/written-while-node-down.bin";
let outage_payload_seed = 0xf1;
timeout(
Duration::from_secs(30),
clients[2]
.put_object()
.bucket(bucket)
.key(outage_key)
.body(ByteStream::from(deterministic_object_body(object_size_bytes, outage_payload_seed)))
.send(),
)
.await??;
let max_outage_write_attempts = if outage_target_manifest_required {
1
} else {
topology.total_drives().max(1)
};
let mut outage_key = None;
for attempt in 0..max_outage_write_attempts {
let candidate_key = if max_outage_write_attempts == 1 {
"cluster/written-while-node-down.bin".to_string()
} else {
format!("cluster/written-while-node-down-{attempt:04}.bin")
};
let put_result = timeout(
Duration::from_secs(30),
clients[2]
.put_object()
.bucket(bucket)
.key(&candidate_key)
.body(ByteStream::from(deterministic_object_body(object_size_bytes, outage_payload_seed)))
.send(),
)
.await;
match put_result {
Ok(Ok(_)) => {
outage_key = Some(candidate_key);
break;
}
Ok(Err(error)) if !outage_target_manifest_required && is_service_unavailable_put(&error) => {}
Ok(Err(error)) => return Err(error.into()),
Err(error) => return Err(error.into()),
}
}
let outage_key = outage_key
.ok_or_else(|| format!("no online pool accepted an outage object after {max_outage_write_attempts} candidates"))?;
let mut outage_peer_erasure_indices = HashSet::new();
for (node_index, node) in cluster.nodes.iter().enumerate() {
@@ -1361,7 +1392,7 @@ mod tests {
continue;
}
for (drive_index, drive) in node.data_dirs.iter().enumerate() {
let census = census_object_version_on_disk(Path::new(drive), bucket, outage_key, None)?;
let census = census_object_version_on_disk(Path::new(drive), bucket, &outage_key, None)?;
if !census.has_xl_meta {
continue;
}
@@ -1457,7 +1488,7 @@ mod tests {
"non-admin Heal is disabled, so the replacement target must remain empty before the explicit root heal"
);
assert!(
!census_object_version_on_disk(&replaced_disk, bucket, outage_key, None)?.has_xl_meta,
!census_object_version_on_disk(&replaced_disk, bucket, &outage_key, None)?.has_xl_meta,
"the object written during the outage must be absent before the explicit root heal"
);
assert_eq!(
@@ -1744,10 +1775,10 @@ mod tests {
loop {
let baseline_recovered = metadata_count(&replaced_disk, bucket, &expected_manifests) == expected_manifests.len();
let outage_recovered =
!outage_target_manifest_required || object_metadata_exists_on_disk(&replaced_disk, bucket, outage_key);
!outage_target_manifest_required || object_metadata_exists_on_disk(&replaced_disk, bucket, &outage_key);
if baseline_recovered && outage_recovered {
let matching = matching_manifest_count(&replaced_disk, bucket, &expected_manifests)?;
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, outage_key, None)?;
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, &outage_key, None)?;
let pool_metadata_matches = match &expected_pool_metadata {
Some(expected) => {
census_object_version_on_disk(&replaced_disk, RUSTFS_META_BUCKET, POOL_METADATA_OBJECT, None)?
@@ -1764,7 +1795,7 @@ mod tests {
}
if Instant::now() >= heal_deadline {
let matching = matching_manifest_count(&replaced_disk, bucket, &expected_manifests)?;
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, outage_key, None)?;
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, &outage_key, None)?;
let pool_metadata =
census_object_version_on_disk(&replaced_disk, RUSTFS_META_BUCKET, POOL_METADATA_OBJECT, None)?;
let final_status = signed_admin_post(&status_url, None, &cluster.access_key, &cluster.secret_key)
@@ -1802,7 +1833,7 @@ mod tests {
expected.key
);
}
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, outage_key, None)?;
let outage_census = census_object_version_on_disk(&replaced_disk, bucket, &outage_key, None)?;
if outage_target_manifest_required {
assert!(
outage_census.is_complete(),
@@ -1825,7 +1856,7 @@ mod tests {
.iter()
.map(|(key, _)| key.clone())
.collect::<HashSet<_>>();
assert!(expected_keys.insert(outage_key.to_string()));
assert!(expected_keys.insert(outage_key.clone()));
let node_listings = assert_all_nodes_list_exact_keys(&clients, bucket, &expected_keys).await?;
let target_client = cluster.create_s3_client(1)?;
@@ -1849,7 +1880,7 @@ mod tests {
}));
}
}
let response = target_client.get_object().bucket(bucket).key(outage_key).send().await?;
let response = target_client.get_object().bucket(bucket).key(&outage_key).send().await?;
let actual = response.body.collect().await?.into_bytes();
let expected_outage_body = deterministic_object_body(object_size_bytes, outage_payload_seed);
assert_eq!(actual.as_ref(), expected_outage_body.as_slice(), "object body changed for {outage_key}");
@@ -1860,7 +1891,7 @@ mod tests {
"expected_sha256": sha256_hex(&expected_outage_body),
"actual_sha256": sha256_hex(&actual),
"expected_physical": null,
"physical": census_object_version_on_disk(&replaced_disk, bucket, outage_key, None)?,
"physical": census_object_version_on_disk(&replaced_disk, bucket, &outage_key, None)?,
}));
}
scripts/diagnose_scanner_enumeration_restart.py
+4
View File
@@ -74,6 +74,10 @@ def converged(report, objects):
def replays_raw_window(previous, current):
if (previous["raw_entries"] == 0 or current["raw_entries"] == 0
or previous["raw_first_entry"] is None or current["raw_first_entry"] is None
or previous["raw_last_entry"] is None or current["raw_last_entry"] is None):
return False
return (previous["raw_first_entry"] == current["raw_first_entry"]
and previous["raw_last_entry"] == current["raw_last_entry"]
and previous["objects_retained"] == current["objects_before"]