Fix Missing ClusterPipeline Lock (#2190)

* ClusterPipeline needs to initialize self._lock, otherwise a class instance will fail when calling get_redis_connection on the node

* fix bad lint picked up from master

* added change to CHANGES file

* force ci build again

* force ci build again

CHANGES

@@ -9,6 +9,7 @@
     * Remove verbose logging when initializing ClusterPubSub, ClusterPipeline or RedisCluster
     * Fix broken connection writer lock-up for asyncio (#2065)
     * Fix auth bug when provided with no username (#2086)
+    * Fix missing ClusterPipeline._lock (#2189)
 
 * 4.1.3 (Feb 8, 2022)
     * Fix flushdb and flushall (#1926)

redis/cluster.py

@@ -754,6 +754,7 @@ def pipeline(self, transaction=None, shard_hint=None):
             cluster_error_retry_attempts=self.cluster_error_retry_attempts,
             read_from_replicas=self.read_from_replicas,
             reinitialize_steps=self.reinitialize_steps,
+            lock=self._lock,
         )
 
     def lock(
@@ -1754,6 +1755,7 @@ def __init__(
         read_from_replicas=False,
         cluster_error_retry_attempts=5,
         reinitialize_steps=10,
+        lock=None,
         **kwargs,
     ):
         """ """
@@ -1776,6 +1778,9 @@ def __init__(
             kwargs.get("encoding_errors", "strict"),
             kwargs.get("decode_responses", False),
         )
+        if lock is None:
+            lock = threading.Lock()
+        self._lock = lock
 
     def __repr__(self):
         """ """

redis/commands/json/__init__.py

@@ -113,6 +113,7 @@ def pipeline(self, transaction=True, shard_hint=None):
                 cluster_error_retry_attempts=self.client.cluster_error_retry_attempts,
                 read_from_replicas=self.client.read_from_replicas,
                 reinitialize_steps=self.client.reinitialize_steps,
+                lock=self.client._lock,
             )
 
         else:

redis/commands/timeseries/__init__.py

@@ -77,6 +77,7 @@ def pipeline(self, transaction=True, shard_hint=None):
                 cluster_error_retry_attempts=self.client.cluster_error_retry_attempts,
                 read_from_replicas=self.client.read_from_replicas,
                 reinitialize_steps=self.client.reinitialize_steps,
+                lock=self.client._lock,
             )
 
         else: