Prepare all nodes in the cluster by installing build dependencies and extracting the Redis source distribution:

yum install -y gcc make tcl
curl -fsSL https://download.redis.io/releases/redis-7.0.12.tar.gz -o /usr/src/redis.tar.gz
tar -xzf /usr/src/redis.tar.gz -C /usr/src/
cd /usr/src/redis-7.0.12
make && make install PREFIX=/opt/redis

Create the runtime directories and install the service script:

mkdir -p /opt/redis/{conf,data,logs}
cp utils/redis_init_script /etc/init.d/redis
sed -i 's/REDISPORT=6379/REDISPORT=6380/' /etc/init.d/redis
chmod +x /etc/init.d/redis
ln -s /opt/redis/conf /etc/redis

Generate the base configuraton file on each server:

cat > /etc/redis/6380.conf << 'EOF'
daemonize yes
port 6380
bind 0.0.0.0
dir /opt/redis/data
logfile /opt/redis/logs/redis-server.log
protected-mode no
supervised auto
EOF

Initialize the Redis service across all nodes:

/etc/init.d/redis start
ss -tlnp | grep 6380

Configure the replica instances to synchronize with the primary. On secondary nodes (192.168.1.11 and 192.168.1.12), append the replication directive:

echo "replicaof 192.168.1.10 6380" >> /etc/redis/6380.conf
/etc/init.d/redis restart

Verify the replication topology by writing test data on the primary node:

redis-cli -h 192.168.1.10 -p 6380 SET cluster:validation "master-write-test"
redis-cli -h 192.168.1.10 -p 6380 GET cluster:validation

Confirm data propagation on replicas:

redis-cli -h 192.168.1.11 -p 6380 GET cluster:validation
redis-cli -h 192.168.1.12 -p 6380 GET cluster:validation

Deploy Redis Sentinel for automatic failover monitoring. Establish the configuration on all nodes:

cat > /etc/redis/sentinel.conf << 'EOF'
port 26380
daemonize yes
logfile "/opt/redis/logs/sentinel.log"
dir /tmp
sentinel monitor primary-node 192.168.1.10 6380 2
sentinel down-after-milliseconds primary-node 5000
sentinel parallel-syncs primary-node 1
sentinel failover-timeout primary-node 120000
EOF

Initialize the Setninel processes:

redis-sentinel /etc/redis/sentinel.conf
redis-cli -p 26380 INFO Sentinel

The output should indicate the primary node with two replicas attached.

To validate automatic failover, terminate the Redis process on the current primary (192.168.1.10):

redis-cli -h 192.168.1.10 -p 6380 DEBUG SEGFAULT

Monitor the election process via Sentinel:

redis-cli -p 26380 INFO Sentinel

Look for the address field changing to 192.168.1.11 or 192.168.1.12 in the output.

Once promotion comlpetes, verify write operations on the new primary:

redis-cli -h 192.168.1.11 -p 6380 SET failover:check "promoted-node"
redis-cli -h 192.168.1.12 -p 6380 GET failover:check

Confirm the old primary reconnects as a replica when restarted.