fix(readiness): gate on lock quorum health (#3100)

* fix(readiness): gate on lock quorum health

Include distributed lock quorum in runtime readiness decisions and expose the signal in health responses.

Map namespace lock quorum failures to ServiceUnavailable instead of InternalError so clients can safely retry while keeping quorum enforcement unchanged.

* fix(ecstore): restore namespace lock error decoding

Add the missing from_u32 arm for NamespaceLockQuorumUnavailable and cover the numeric roundtrip in error tests.
This commit is contained in:
houseme
2026-05-28 01:12:52 +08:00
committed by GitHub
parent 4648de9e62
commit 5dfc1b5f07
9 changed files with 502 additions and 113 deletions
+7 -13
View File
@@ -39,12 +39,12 @@ impl SetDisks {
let write_lock_guard = if !opts.no_lock {
let ns_lock = self.new_ns_lock(bucket, object).await?;
Some(ns_lock.get_write_lock(get_lock_acquire_timeout()).await.map_err(|e| {
StorageError::other(format!(
"Failed to acquire write lock: {}",
self.format_lock_error_from_error(bucket, object, "write", &e)
))
})?)
Some(
ns_lock
.get_write_lock(get_lock_acquire_timeout())
.await
.map_err(|e| self.map_namespace_lock_error(bucket, object, "write", e))?,
)
} else {
None
};
@@ -713,13 +713,7 @@ impl SetDisks {
.await?
.get_write_lock(get_lock_acquire_timeout())
.await
.map_err(|e| {
let message = format!(
"Failed to acquire write lock: {}",
self.format_lock_error_from_error(bucket, object, "write", &e)
);
DiskError::other(message)
})?;
.map_err(|e| DiskError::other(self.map_namespace_lock_error(bucket, object, "write", e).to_string()))?;
self.heal_object_dir_locked(bucket, object, dry_run, remove).await
}
+24
View File
@@ -50,6 +50,30 @@ impl SetDisks {
}
}
pub(super) fn map_namespace_lock_error(
&self,
bucket: &str,
object: &str,
mode: &'static str,
err: rustfs_lock::error::LockError,
) -> StorageError {
match err {
rustfs_lock::error::LockError::QuorumNotReached { required, achieved } => {
StorageError::NamespaceLockQuorumUnavailable {
mode,
bucket: bucket.to_string(),
object: object.to_string(),
required,
achieved,
}
}
other => StorageError::other(format!(
"Failed to acquire {mode} lock: {}",
self.format_lock_error_from_error(bucket, object, mode, &other)
)),
}
}
pub(super) async fn get_disks_internal(&self) -> Vec<Option<DiskStore>> {
let rl = self.disks.read().await;