Fix empty primary may have dirty slots data due to bad migration (#1285)

If we become an empty primary for some reason, we still need to
check if we need to delete dirty slots, because we may have dirty
slots data left over from a bad migration. Like the target node forcibly
executes CLUSTER SETSLOT NODE to take over the slot without
performing key migration.

Signed-off-by: Binbin <[email protected]>

src/cluster_legacy.c

@@ -2451,6 +2451,7 @@ void clusterUpdateSlotsConfigWith(clusterNode *sender, uint64_t senderConfigEpoc
      * need to delete all the keys in the slots we lost ownership. */
     uint16_t dirty_slots[CLUSTER_SLOTS];
     int dirty_slots_count = 0;
+    int delete_dirty_slots = 0;
 
     /* We should detect if sender is new primary of our shard.
      * We will know it if all our slots were migrated to sender, and sender
@@ -2677,6 +2678,12 @@ void clusterUpdateSlotsConfigWith(clusterNode *sender, uint64_t senderConfigEpoc
             serverLog(LL_NOTICE,
                       "My last slot was migrated to node %.40s (%s) in shard %.40s. I am now an empty primary.",
                       sender->name, sender->human_nodename, sender->shard_id);
+            /* We may still have dirty slots when we became a empty primary due to
+             * a bad migration.
+             *
+             * In order to maintain a consistent state between keys and slots
+             * we need to remove all the keys from the slots we lost. */
+            delete_dirty_slots = 1;
         }
     } else if (dirty_slots_count) {
         /* If we are here, we received an update message which removed
@@ -2686,6 +2693,10 @@ void clusterUpdateSlotsConfigWith(clusterNode *sender, uint64_t senderConfigEpoc
          *
          * In order to maintain a consistent state between keys and slots
          * we need to remove all the keys from the slots we lost. */
+        delete_dirty_slots = 1;
+    }
+
+    if (delete_dirty_slots) {
         for (int j = 0; j < dirty_slots_count; j++) {
             serverLog(LL_NOTICE, "Deleting keys in dirty slot %d on node %.40s (%s) in shard %.40s", dirty_slots[j],
                       myself->name, myself->human_nodename, myself->shard_id);
@@ -6069,7 +6080,7 @@ void removeChannelsInSlot(unsigned int slot) {
 /* Remove all the keys in the specified hash slot.
  * The number of removed items is returned. */
 unsigned int delKeysInSlot(unsigned int hashslot) {
-    if (!kvstoreDictSize(server.db->keys, hashslot)) return 0;
+    if (!countKeysInSlot(hashslot)) return 0;
 
     unsigned int j = 0;
 

tests/unit/cluster/replica-migration.tcl

@@ -400,3 +400,23 @@ start_cluster 4 4 {tags {external:skip cluster} overrides {cluster-node-timeout
 start_cluster 4 4 {tags {external:skip cluster} overrides {cluster-node-timeout 1000 cluster-migration-barrier 999}} {
     test_cluster_setslot "setslot"
 } my_slot_allocation cluster_allocate_replicas ;# start_cluster
+
+start_cluster 3 0 {tags {external:skip cluster} overrides {cluster-node-timeout 1000 cluster-migration-barrier 999}} {
+    test "Empty primary will check and delete the dirty slots" {
+        R 2 config set cluster-allow-replica-migration no
+
+        # Write a key to slot 0.
+        R 2 incr key_977613
+
+        # Move slot 0 from primary 2 to primary 0.
+        R 0 cluster bumpepoch
+        R 0 cluster setslot 0 node [R 0 cluster myid]
+
+        # Wait for R 2 to report that it is an empty primary (cluster-allow-replica-migration no)
+        wait_for_log_messages -2 {"*I am now an empty primary*"} 0 1000 50
+
+        # Make sure primary 0 will delete the dirty slots.
+        verify_log_message -2 "*Deleting keys in dirty slot 0*" 0
+        assert_equal [R 2 dbsize] 0
+    }
+} my_slot_allocation cluster_allocate_replicas ;# start_cluster