fix(ilm): preserve duplicate metadata and reject unapplied purges

Merge lock and tag revisions for duplicate versions before all-pool
lifecycle evaluation using the existing metadata reconciliation rules.

Distinguish ordinary lifecycle refusal from object absence and clear
rejected batch replication state so a late policy withdrawal cannot
schedule a purge. Preserve restored-copy cleanup and real absence.

Cover single and batch protections, legitimate metadata removals, and
actual replication queue and HTTP target behavior.

Signed-off-by: Feng Ruohang <rh@vonng.com>

cmd/erasure-server-pool-consistency.go

@@ -19,6 +19,7 @@ package cmd
 
 import (
 	"context"
+	"errors"
 	"maps"
 	"sort"
 	"strings"
@@ -769,6 +770,11 @@ func (z *erasureServerPools) deleteObjectsReconciled(ctx context.Context, bucket
 			deleteOpts.Versioned = opts.PrefixEnabledFn(obj.ObjectName)
 		}
 		_, err := z.deleteObjectReconciled(ctx, bucket, obj.ObjectName, deleteOpts)
+		if errors.Is(err, errLifecycleNotApplicable) {
+			// A later lifecycle rejection did not commit a local purge. Do
+			// not let the worker replicate the request's pending decision.
+			deleted[i].ReplicationState = ReplicationState{}
+		}
 		if err != nil && !isErrObjectNotFound(err) && !isErrVersionNotFound(err) {
 			errs[i] = err
 		} else {