Hi all, has anyone been able to run fleet using r...
# fleet
c
Hi all, has anyone been able to run fleet using redis v7.0.5? We are having an issue starting up the process. More details in the 🧵
Copy code
$ sudo /usr/local/bin/fleet serve --config /etc/fleetdm/fleetdm.yml
Using config file: /etc/fleetdm/fleetdm.yml
Failed to start: initialize Redis: refresh cluster: redisc: all nodes failed
ERR unknown command 'CLUSTER', with args beginning with: 'SLOTS'
I think the problem is that
CLUSTER SLOTS
has been deprecated in 7.0. I've found this open issue but without any updates for a while.
Also for additional context, this is my fleet configuration:
Copy code
mysql:
  address: mysql.server.domain:17315
  database: fleetdm
  username: username
  password: mysql_secret
redis:
  address: redis.server.domain:17316
  username: default
  password: redis_secret
  use_tls: true
  cluster_follow_redirections: true
  connect_retry_attempts: 3
server:
  address: 0.0.0.0:8080
  cert: /etc/fleetdm/server.crt
  key: /etc/fleetdm/server.key
auth:
  jwt_key: key_secret
filesystem:
  status_log_file: /var/log/fleetdm/status.log
  result_log_file: /var/log/fleetdm/result.log
  enable_log_rotation: true
logging:
  json: true
I also tried disabling tls in the redis instance and set
use_tls: false
, then just using
address
and
username
in the config, but this time the ouput doesn't really show what's is wrong. See below. So, again I suspect it might be related to the redis version
Copy code
Using config file: /etc/fleetdm/fleetdm.yml
Failed to start: initialize Redis: refresh cluster: redisc: all nodes failed
EOF
Another note: Both log files configured in the config file are not created
k
Hey @Camilo! We've tested Fleet with Redis 5.0.14 and 6.2.7. I'm reaching out to the team to see if anyone has tested with v7.0.5
Can you access your Redis server with
redis-cli
and run
INFO server
and get us the output (mostly interested in
redis_mode
, it might not be a Redis Cluster), and if it does say
cluster
, then also run
CLUSTER INFO
which will give health information about the cluster (it may be in a failed mode).
c
Hi @Kathy Satterlee you are totally right
redis_mode:standalone
Full output:
Copy code
INFO server
# Server
redis_version:7.0.11
redis_git_sha1:f9a42950
redis_git_dirty:0
redis_build_id:49adc3127143cf45
redis_mode:standalone
os:Linux XXXXXXX
arch_bits:64
monotonic_clock:POSIX clock_gettime
multiplexing_api:epoll
atomicvar_api:c11-builtin
gcc_version:12.2.1
tcp_port:17315
server_time_usec:1684269072250033
uptime_in_seconds:265
uptime_in_days:0
hz:10
configured_hz:10
lru_clock:6547472
executable:/usr/bin/redis-server
config_file:/etc/redis.conf
io_threads_active:0
I don't think there's a parameter for fleet to connect to a redis node with standalone mode? When 1st setting this up, I had a single node and did notice the need of a cluster so I set up two nodes but just learned they're not in cluster mode 🤦‍♂️ I'll try configure it in cluster mode, but is this a requirement for fleet?
k
You can definitely use a standalone Redis instance. What happens if you remove the cluster settings from your config:
Copy code
cluster_follow_redirections: true
  connect_retry_attempts: 3
c
It keeps showing the same error
Copy code
Using config file: /etc/fleetdm/fleetdm.yml
Failed to start: initialize Redis: refresh cluster: redisc: all nodes failed
ERR unknown command 'CLUSTER', with args beginning with: 'SLOTS'
Redis config:
Copy code
redis:
  address: <http://redis-1234.com:17316|redis-1234.com:17316>
  username: default
  password: redis_secret
  use_tls: true
I noticed that if I remove
use_tls: true
the error changes to just
Copy code
Using config file: /etc/fleetdm/fleetdm.yml
Failed to start: initialize Redis: refresh cluster: redisc: all nodes failed
EOF
Hi @Kathy Satterlee just wondering if you've heard from anyone about running fleet with redis v7? Do you know if there anyway to verbose/debug the connection issues to redis as per me previous messages above? Thanks!