master: repair the lookup index after a vacuum that raced a disconnect (#10226)

* master: re-create a vacuum-committed volume the index has lost

SetVolumeAvailable dereferenced vid2location[vid] with no nil check. A
disconnect during a long vacuum can drop a single-replica volume from the
lookup index while it stays on the node; the commit then panics the master
instead of re-registering it. Re-create the entry, and seed its size tracking
so assigns are counted right away rather than after the next heartbeat.

* master: re-register a vacuumed volume the index lost on mark-writable

The maintenance worker re-enables a volume by marking it writable after the
vacuum commit, but VolumeMarkReadonly only updated nodes the lookup index
already knew. If a disconnect race dropped the volume from the index during the
vacuum, that path was a no-op and the volume stayed "not found" until the next
full heartbeat healed it. Re-register it from the node that still holds it.

weed/server/master_grpc_server_volume.go

@@ -15,6 +15,7 @@ import (
 	"github.com/seaweedfs/raft"
 
 	"github.com/seaweedfs/seaweedfs/weed/glog"
+	"github.com/seaweedfs/seaweedfs/weed/pb"
 	"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
 	"github.com/seaweedfs/seaweedfs/weed/security"
 	"github.com/seaweedfs/seaweedfs/weed/storage/needle"
@@ -323,18 +324,33 @@ func (ms *MasterServer) VolumeMarkReadonly(ctx context.Context, req *master_pb.V
 
 	replicaPlacement, _ := super_block.NewReplicaPlacementFromByte(byte(req.ReplicaPlacement))
 	vl := ms.Topo.GetVolumeLayout(req.Collection, replicaPlacement, needle.LoadTTLFromUint32(req.Ttl), types.ToDiskType(req.DiskType))
-	dataNodes := ms.Topo.Lookup(req.Collection, needle.VolumeId(req.VolumeId))
+	vid := needle.VolumeId(req.VolumeId)
+	dataNodes := ms.Topo.Lookup(req.Collection, vid)
 
+	found := false
 	for _, dn := range dataNodes {
 		if dn.Ip == req.Ip && dn.Port == int(req.Port) {
+			found = true
 			if req.IsReadonly {
-				vid := needle.VolumeId(req.VolumeId)
 				vl.SetVolumeReadOnly(dn, vid)
 				if pending := vl.GetPendingSize(vid); pending > 0 {
 					glog.V(0).Infof("volume %d marked readonly with %d pending bytes", vid, pending)
 				}
 			} else {
-				vl.SetVolumeWritable(dn, needle.VolumeId(req.VolumeId))
+				vl.SetVolumeWritable(dn, vid)
+			}
+		}
+	}
+
+	// A vacuum worker marks the volume writable after a successful commit. If the
+	// index lost it meanwhile (e.g. a disconnect race during the vacuum), the loop
+	// found nothing to update; re-register it from the node that still holds it so
+	// LookupVolume stops returning "not found".
+	if !found && !req.IsReadonly {
+		if dn := ms.Topo.LookupDataNodeByAddress(pb.NewServerAddress(req.Ip, int(req.Port), 0)); dn != nil {
+			if vi, err := dn.GetVolumesById(vid); err == nil {
+				ms.Topo.RegisterVolumeLayout(vi, dn)
+				glog.V(0).Infof("volume %d re-registered from %s:%d, was missing from the lookup index at mark-writable", vid, req.Ip, req.Port)
 			}
 		}
 	}