The connection is closed by Connection closed by server.

[yuerem]

Version:
redis-py 5.0.0
Redis 6.2.6
python 3.11

Description:
I am using Amazon ElastiCache for Redis. After the execution of the maintenance window for Amazon ElastiCache for Redis, exceptions such as

redis.exceptions.RedisClusterException: Redis Cluster cannot be connected. Please provide at least one reachable node: Timeout connecting to server

and

redis.exceptions.ConnectionError: Connection closed by server.

were thrown.

I suspect that the cause was the failure of name resolution because the node's IP was changed after the execution of the maintenance window. Does anyone know a good solution to this issue?

Currently, I have globally defined the client as follows and am using it in various functions.

r = RedisCluster(
    host='xxx',
    port='yyy',
    encoding="utf-8",
    decode_responses=True,
    )

def sample1():
   r.set('foo', 'bar')

def sample2():
    r.get('foo')

[jimcoolberd]

+1

[martyanov]

Same issue. The same versions as OP, but redis-py is 5.0.1.

[martyanov]

Probaly the cause is intermittent network or DNS issue.

[martyanov]

And a bit of context. We're using redis[hiredis] and asyncio.

[jimcoolberd]

It may be a problem with network quality. Now that the network speed has been improved, it now works.

[yuerem]

The client is being used globally, so in case there is a network or DNS issue, I believe it will require a client initialization.
Currently, we are catching exceptions and initializing the client, but is there a better way to handle this?

example:

r = RedisCluster(
    host='xxx',
    port='yyy',
    encoding="utf-8",
    decode_responses=True,
    )

def sample1():
   r.set('foo', 'bar')

def sample2():
    r.get('foo')

while True:
    try:
        sample1()
        sample2()
    except ConnectionError, RedisClusterException, TimeoutError
        r = RedisCluster(
            host='xxx',
            port='yyy',
            encoding="utf-8",
            decode_responses=True,
            )