redis/redis-py · error · RedisClusterException

No targets were found to execute {cmd.args} command on

Error message

No targets were found to execute {cmd.args} command on

What it means

The pipeline-execution analogue of error 65: raised inside the pipeline execute path when _determine_nodes() returns an empty list for a queued command. A pipeline command must resolve to at least one target node; zero means the routing policy found no candidate (e.g. replica policy with no replicas, or empty node group) and the pipeline cannot place the command.

Source

Thrown at redis/asyncio/cluster.py:2939

                                response_policy=ResponsePolicy.DEFAULT_KEYED,
                            )
                    else:
                        if command_flag in client._command_flags_mapping:
                            command_policies = CommandPolicies(
                                request_policy=client._command_flags_mapping[
                                    command_flag
                                ]
                            )
                        else:
                            command_policies = CommandPolicies()

                target_nodes = await client._determine_nodes(
                    *cmd.args,
                    request_policy=command_policies.request_policy,
                    node_flag=passed_targets,
                )
                if not target_nodes:
                    raise RedisClusterException(
                        f"No targets were found to execute {cmd.args} command on"
                    )
            cmd.command_policies = command_policies
            if len(target_nodes) > 1:
                raise RedisClusterException(f"Too many targets for command {cmd.args}")
            node = target_nodes[0]
            if node.name not in nodes:
                nodes[node.name] = (node, [])
            nodes[node.name][1].append(cmd)

        # Start timing for observability
        start_time = time.monotonic()

        errors = await asyncio.gather(
            *(
                asyncio.create_task(node[0].execute_pipeline(node[1]))
                for node in nodes.values()
            )

View on GitHub (pinned to da03cdc7e8)

Solutions

1. Pass target_nodes explicitly on the pipeline command so routing bypasses policy resolution.
2. If using read_from_replicas, ensure replicas exist (CLUSTER NODES) or set read_from_replicas=False.
3. Refresh topology (await rc.refresh_nodes()) and retry the pipeline after failover completes.

Example fix

// before
pipe = rc.pipeline()
pipe.get('k')
await pipe.execute()  # read_from_replicas=True, no replicas

// after
pipe = rc.pipeline()
pipe.get('k', target_nodes=RedisCluster.PRIMARIES)
await pipe.execute()

async def pipelined_read(rc, key):
    if rc.read_from_replicas and not rc.get_replicas():
        pipe = rc.pipeline()
        pipe.get(key, target_nodes=RedisCluster.PRIMARIES)
        return (await pipe.execute())[0]
    pipe = rc.pipeline()
    pipe.get(key)
    return (await pipe.execute())[0]

from redis.asyncio.cluster import RedisCluster

def pipeline_has_targets(rc: RedisCluster) -> bool:
    return bool(rc.get_primaries())

from redis.cluster import RedisClusterException

pipe = rc.pipeline()
pipe.get('k')
try:
    await pipe.execute()
except RedisClusterException as e:
    if 'No targets' in str(e):
        await rc.refresh_table_nodes()
        pipe = rc.pipeline()
        pipe.get('k', target_nodes=RedisCluster.PRIMARIES)
        await pipe.execute()
    else:
        raise

When it happens

Trigger: Queuing a replica-routed read in a pipeline when the cluster has no replicas; a pipeline command whose policy targets a node group that is empty in the current topology; transient empty node set during failover when the pipeline flushes.

Common situations: read_from_replicas=True with no replicas present; long-lived pipeline that buffers commands while topology changes underneath; pipeline issued right after client construction before topology settled.

Related errors

• No targets were found to execute {args} command on
• Cannot execute FT.CURSOR commands without FT.AGGREGATE
• No way to dispatch this command to Redis Cluster. Missing ke
• shard_hint is deprecated in cluster mode
• Too many targets for command {cmd.args}

AI-assisted analysis of redis/redis-py@da03cdc7e8 (2026-08-04). Data as JSON: /data/errors/860ade2484c2cfc4.json. Report an issue: GitHub.