Changed dynamic_startup_nodes default value to 'True' to prevent brea…

…king backward compatibility (#2251)

README.md

@@ -1268,15 +1268,15 @@ Using scripting within pipelines in cluster mode is **not supported**.
      When set to true, read commands will be assigned between the
      primary and its replications in a Round-Robin manner.
      
- dynamic_startup_nodes: (default=False)
+ dynamic_startup_nodes: (default=True)
 
      Set the RedisCluster's startup nodes to all of the discovered nodes.
      If true, the cluster's discovered nodes will be used to determine the
      cluster nodes-slots mapping in the next topology refresh.
      It will remove the initial passed startup nodes if their endpoints aren't
      listed in the CLUSTER SLOTS output.
      If you use dynamic DNS endpoints for startup nodes but CLUSTER SLOTS lists
-     specific IP addresses, keep it at false.
+     specific IP addresses, it is best to set it to false.
      
  cluster_error_retry_attempts: (default=3)
 

redis/cluster.py

@@ -482,7 +482,7 @@ def __init__(
         require_full_coverage=False,
         reinitialize_steps=10,
         read_from_replicas=False,
-        dynamic_startup_nodes=False,
+        dynamic_startup_nodes=True,
         url=None,
         **kwargs,
     ):
@@ -512,12 +512,12 @@ def __init__(
              primary and its replications in a Round-Robin manner.
          :dynamic_startup_nodes: 'bool'
              Set the RedisCluster's startup nodes to all of the discovered nodes.
-             If true, the cluster's discovered nodes will be used to determine the
-             cluster nodes-slots mapping in the next topology refresh.
+             If true (default value), the cluster's discovered nodes will be used to
+             determine the cluster nodes-slots mapping in the next topology refresh.
              It will remove the initial passed startup nodes if their endpoints aren't
              listed in the CLUSTER SLOTS output.
              If you use dynamic DNS endpoints for startup nodes but CLUSTER SLOTS lists
-             specific IP addresses, keep it at false.
+             specific IP addresses, it is best to set it to false.
         :cluster_error_retry_attempts: 'int'
              Retry command execution attempts when encountering ClusterDownError
              or ConnectionError
@@ -1293,7 +1293,7 @@ def __init__(
         from_url=False,
         require_full_coverage=False,
         lock=None,
-        dynamic_startup_nodes=False,
+        dynamic_startup_nodes=True,
         **kwargs,
     ):
         self.nodes_cache = {}