Fixes for redis/redis-py
Verified fixes and incident notes extracted from GitHub Issues for redis/redis-py.
Repository
redis / redis-py
Updated: 2026-02-13
- slowlog_get() raises TypeError when decode_responses=True
- Question: why is `blocking` missing as a lock() parameter?
- error in set when ex is a float is confusing
- 4.0.0b1 breaks script_flush on pre-6.2 server
- AttributeError: 'Redis' object has no attribute 'sentinel_masters'
- Another race condition in health checks and pubsub
- exception in listen
- Question: resolving hostname when retry on connection timeout
- AttributeError in UnixDomainSocketConnection starting v4.1.0
- redis.asyncio trying to reuse a closed connection writer
- asyncio RedisCluster is not working for Amazon Elasticache
- byte vector is incorrectly decoded as utf-8 string in ft result class
- WRONGPASS response doesn't raise AuthenticationError exception
- Invalid Default Argument For LMPOP
- AsyncIO Cluster Mode Lock doesn't work
- Missing "blocking" argument when initializing blocking in asynchronous mode
- KeyError in asyncio cluster get_node_from_slot
- 'UnixDomainSocketConnection' object has no attribute '_command_packer'
- [Bug] Aioredis cancellation of blocking operations stucks next operations
- Redis pubsub task cancellation raises ConnectionError
- TimeoutError during ClusterPipeline makes the client unrecoverable
- `SentinelManagedConnection` delays new master detection and Risks data loss
- Retry Mechanism Fails When Redis Container is Paused
- max_connections is not being honoured in RedisCluster mode
- Async lock can enter a deadlock if it is cancelled while being released
- query object is not outputing the `TIMEOUT` argument in case of 0 ( 0 means unlimited for RediSearch TIMEOUT parameter )
- Instantiating BlockingConnectionPool.from_url with timeout in query args fails
- `redis.cluster.RedisCluster` with RESP3 and hiredis causes segmentation fault
- Sub-unsub-resub causes redispy to forget the channel
- redis._compat: global name 'InterruptedError' is not defined
- connection pool in multi-thread is not thread safe
- KeyError on repr in any ConnectionPool without host/port/db arguments
- Misleading docstring of class `Redis`
- LMOVE, BLMOVE return incorrect response
- v4.3.2 contains backwards incompatible change to `redis.lock.Lock`
- v4.3.3 continues to contain backwards incompatible change to ``redis.lock.Lock``
- RediSearch: search command doesn't support asyncio Pipeline
- 4.6.0 - ConnectionError max number of clients reached
- sentinel_get_master_addr_by_name() returns boolean instead of (host, port) pair
- Adding required `message` to `LockError` broke backwards compatibility
- Redundant async-timeout dependency in Redis 5.0.2
- Release a lock from a non-owned thread should raise proper error
- Async performance issue since version 5.3.0