0

I have Master-Slave(3 nodes) setup using Redis Sentinel but when I try to do failover, I am seeing the below error

127.0.0.1:26379> sentinel failover mymaster 
(error) NOGOODSLAVE No suitable replica to promote

Below is the Redis server configurations for master and slave nodes

#redis.conf of master node
bind 127.0.0.1 192.26.x.1
protected-mode no
daemonize yes
logfile /opt/softwares/redis-6.0.16/log/redis-server.log

#redis.conf of slave node 1
bind 127.0.0.1 192.26.x.2
protected-mode no
daemonize yes
logfile /opt/softwares/redis-6.0.16/log/redis-server.log
replicaof 192.26.x.1 6379

#redis.conf of slave node 2
bind 127.0.0.1 192.26.x.3
protected-mode no
daemonize yes
logfile /opt/softwares/redis-6.0.16/log/redis-server.log
replicaof 192.26.x.1 6379

Below is my sentinel nodes config

# Same for all sentinel nodes
bind 127.0.0.1 192.26.x.1
protected-mode no
port 26379
daemonize yes
pidfile "/var/run/redis-sentinel.pid"
logfile "/var/log/redis-sentinel.log"
sentinel monitor mymaster 192.26.x.1 6379 2
sentinel down-after-milliseconds mymaster 5000
sentinel failover-timeout mymaster 60000

When I queried the slave's status, I can see master-related config are not defined.

127.0.0.1:26379> SENTINEL masters
   31) "master-link-down-time"
   32) "0"
   33) "master-link-status"
   34) "err"
   35) "master-host"
   36) "?"
   37) "master-port"
   38) "0"

I am starting the Redis-servers in all nodes then Redis-sentinel. Not sure the ordering of starting nodes matters.

Please let me know whether I am missing some configurations or doing something wrong. The Redis version that I am using is 6.0.16.

Thanks in advance.

CrazyCoder
  • 2,465
  • 8
  • 36
  • 57

1 Answers1

0

Changing the order of IP addresses in a bind in all configuration files solved the issue i.e.

bind <public_ip> <localhost>
CrazyCoder
  • 2,465
  • 8
  • 36
  • 57