fix(storage): harden ODM and scanner publication (#7187)

* fix(storage): harden ODM and scanner publication

* fix(app): simplify absent SSE configuration matching

* test(heal): settle PUT rename tails before disk-wipe fixtures

* fix(ecstore): remove duplicate local rename implementation

Keep the canonical commit module after concurrent storage changes merged.
The control-write and rollback changes are already present there.

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

* fix(ci): satisfy new clippy lints

* style(scanner): order merged test imports

* fix(scanner): invalidate bucket work after namespace completion

* fix(scanner): fence cached snapshots by scan execution

---------

Co-authored-by: houseme <housemecn@gmail.com>
Co-authored-by: heihutu <heihutu@gmail.com>
Co-authored-by: zhi22915 <qiuzgang@gmail.com>
This commit is contained in:
Zhengchao An
2026-09-05 21:47:12 +08:00
committed by GitHub
parent 447f3c704b
commit e2a921bc16
35 changed files with 2252 additions and 274 deletions
+24 -15
View File
@@ -118,6 +118,7 @@ impl ScannerIOCache for SetDisks {
all_buckets,
scope,
digest: scan_plan_digest,
execution_digest,
leader_epoch,
tier_registry_generation,
publication_epoch,
@@ -137,20 +138,24 @@ impl ScannerIOCache for SetDisks {
.ok_or_else(|| StorageError::other("scanner cache publication is blocked by data movement"))?,
};
let mut old_cache = DataUsageCache::default();
if let Err(e) = old_cache.load(self.clone(), DATA_USAGE_CACHE_NAME).await {
warn!(
target: "rustfs::scanner::io",
event = EVENT_SCANNER_CACHE_PERSIST_STATE,
component = LOG_COMPONENT_SCANNER,
subsystem = LOG_SUBSYSTEM_IO,
pool = self.pool_index,
set = self.set_index,
cache_name = DATA_USAGE_CACHE_NAME,
state = "old_cache_load_failed",
error = %e,
"Scanner old data usage cache load failed; rebuilding from bucket caches"
);
}
let initial_revisions = match old_cache.load_with_revisions(self.clone(), DATA_USAGE_CACHE_NAME).await {
Ok(revisions) => Some(revisions),
Err(e) => {
warn!(
target: "rustfs::scanner::io",
event = EVENT_SCANNER_CACHE_PERSIST_STATE,
component = LOG_COMPONENT_SCANNER,
subsystem = LOG_SUBSYSTEM_IO,
pool = self.pool_index,
set = self.set_index,
cache_name = DATA_USAGE_CACHE_NAME,
state = "old_cache_load_failed",
error = %e,
"Scanner old data usage cache load failed; rebuilding from bucket caches"
);
None
}
};
let scoped_scan = prepare_scoped_set_scan(
&old_cache,
&buckets,
@@ -195,6 +200,7 @@ impl ScannerIOCache for SetDisks {
};
cache.info.last_update = Some(now);
cache.info.snapshot_complete = true;
cache.info.scan_execution_digest = Some(execution_digest);
cache.info.lkg_snapshot_complete = false;
cache.info.lkg_next_cycle = None;
cache.info.lkg_last_update = None;
@@ -208,6 +214,7 @@ impl ScannerIOCache for SetDisks {
self,
&updates,
cache,
initial_revisions.as_ref(),
cache_cycle_floor.as_ref(),
expected_publication_epoch,
)
@@ -637,7 +644,7 @@ impl ScannerIOCache for SetDisks {
let cache_name = path_join_buf(&[&bucket.name, DATA_USAGE_CACHE_NAME]);
let bucket_scan_plan_digest =
scanner_bucket_cache_digest(scan_plan_digest, dirty_usage_buckets_clone.get(&bucket.name).copied());
scanner_bucket_cache_digest(execution_digest, dirty_usage_buckets_clone.get(&bucket.name).copied());
if let Some(server_epoch) = remote_server_epoch {
let request_sequence = remote_session_sequence;
@@ -1360,6 +1367,7 @@ impl ScannerIOCache for SetDisks {
cache.info.next_cycle = want_cycle;
cache.info.last_update.get_or_insert_with(SystemTime::now);
cache.info.snapshot_complete = true;
cache.info.scan_execution_digest = Some(execution_digest);
cache.info.lkg_snapshot_complete = false;
cache.info.lkg_next_cycle = None;
cache.info.lkg_last_update = None;
@@ -1371,6 +1379,7 @@ impl ScannerIOCache for SetDisks {
self.clone(),
&updates,
cache_snapshot,
initial_revisions.as_ref(),
cache_cycle_floor.as_ref(),
expected_publication_epoch,
)