My environment: I have three Ubuntu servers. One server is used as a load balancer using Nginx. The other two servers contain the exact same project (are identical apart from redis where one box is the master and other is the slave).
The Programs/Applications I'm using Python, Gunicorn, Django, Celery, Redis, Sentinel
What My project does: I have a URL which takes in a GET request, does some logic regarding the request and saves into redis.
def save(key, value):
conn = redis_cluster_connect()
print conn
conn.set(key, value)
conn.set('bar','foo')
This was working fine when my connection was:
def redis_connect():
print 'redis connected'
return redis.Redis(host="xxx.xx.xxx.x", port=6380, db=1) # CHANGE TO THE REDIS PORT THAT IS ACTIVE
But when I use the cluster connection:
def redis_cluster_connect():
startup_nodes = [{"host": "xxx.xx.xxx.x", "port": "6380"}]
rc = StrictRedisCluster(startup_nodes=startup_nodes, decode_responses=True)
return rc
The Error I get:
[2016-09-15 13:28:59,682: INFO/MainProcess] Received task: testapp.tasks.mobilise_stops[cc64c425-bd37-4896-b6ab-4319de5fb743]
[2016-09-15 13:28:59,684: WARNING/Worker-1] Oh no! Task failed: RedisClusterException("ERROR sending 'cluster slots' command to redis server: {'host': 'xxx.xx.xxx.x', 'port': '6380'}",)
[2016-09-15 13:28:59,685: ERROR/MainProcess] Task testapp.tasks.mobilise_stops[cc64c425-bd37-4896-b6ab-4319de5fb743] raised unexpected: RedisClusterException("ERROR sending 'cluster slots' command to redis server: {'host': 'xxx.xx.xxx.x', 'port': '6380'}",)
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/celery/app/trace.py", line 240, in trace_task
R = retval = fun(*args, **kwargs)
File "/usr/local/lib/python2.7/dist-packages/celery/app/trace.py", line 438, in __protected_call__
return self.run(*args, **kwargs)
File "/var/www/gateway/venv/gateway/testapp/tasks.py", line 50, in mobilise_stops
save(mobilise_stops.request.id, 'Success')
File "/var/www/gateway/venv/gateway/testapp/tasks.py", line 28, in save
conn = redis_cluster_connect()
File "/var/www/gateway/venv/gateway/testapp/tasks.py", line 19, in redis_cluster_connect
rc = StrictRedisCluster(startup_nodes=startup_nodes, decode_responses=True)
File "/usr/local/lib/python2.7/dist-packages/rediscluster/client.py", line 157, in __init__
**kwargs
File "/usr/local/lib/python2.7/dist-packages/rediscluster/connection.py", line 83, in __init__
self.nodes.initialize()
File "/usr/local/lib/python2.7/dist-packages/rediscluster/nodemanager.py", line 148, in initialize
raise RedisClusterException("ERROR sending 'cluster slots' command to redis server: {0}".format(node))
RedisClusterException: ERROR sending 'cluster slots' command to redis server: {'host': 'xxx.xx.xxx.x', 'port': '6380'}
When I run redis-cli and set a variable on the master (on server 1) I can retrieve it on the slave (on server 2)
server 1:
xxx.xx.xxx.x:6380> set test 100
OK
server 2:
xxx.xx.xxx.x:6381> get test
"100"
When I try doing the cluster slots command in the redis client I get the same error as the above.
xxx.xx.xxx.x:6380> cluster slots
(error) ERR unknown command 'cluster'
One thing to consider is my redis config file doesn't have a tcp socket, when I tried using tcp redis would not work. It gave me an error when I tried running redis. This was changed to the default Unixsocket that comes in the default redis.conf file
From:
tcp-backlog 511
To:
unixsocket /var/run/redis/redis.sock
unixsocketperm 700
