Redis Version：4.0.10、4.0.11
Deploy Node：3 sentinel、1 master、2 slave
启动顺序：先启动master，然后slave，最后启动sentinel
sentinel.conf #按照redis官网配置
port 30001
daemonize yes
loglevel notice
dir "/home/work/redis_setinel_test/tmp"
logfile "/home/work/redis_setinel_test/30001.log"
sentinel monitor redis-config-1 127.0.0.1 6666 2
sentinel down-after-milliseconds redis-config-1 5000
sentinel failover-timeout redis-config-1 60000
redis6666.conf
port 6666
daemonize yes
pidfile "/home/work/redis_setinel_test/6666.pid"
tcp-backlog 511
timeout 0
tcp-keepalive 0
loglevel notice
logfile "/home/work/redis_setinel_test/6666.log"
databases 16
save 3600 1
stop-writes-on-bgsave-error yes
rdbcompression yes
rdbchecksum yes
dbfilename "dump.rdb"
dir "/home/work/redis_setinel_test/db"
slave-serve-stale-data yes
slave-read-only yes
repl-diskless-sync no
repl-diskless-sync-delay 5
repl-timeout 60
repl-disable-tcp-nodelay no
repl-backlog-size 10mb
repl-backlog-ttl 3600
slave-priority 100
# requirepass foobared
rename-command CONFIG _CONFIG
rename-command FLUSHALL ""
maxclients 20000
maxmemory 14gb
maxmemory-policy allkeys-lru
appendonly no
appendfilename "appendonly.aof"
appendfsync everysec
no-appendfsync-on-rewrite yes
auto-aof-rewrite-percentage 80
auto-aof-rewrite-min-size 128mb
aof-load-truncated yes
aof-rewrite-incremental-fsync yes
lua-time-limit 5000
#REDIS CLUSTER
cluster-enabled no
cluster-config-file "nodes.conf"
cluster-node-timeout 15000
cluster-migration-barrier 1
cluster-require-full-coverage no
slowlog-log-slower-than 10000
slowlog-max-len 128
latency-monitor-threshold 0
hash-max-ziplist-entries 512
hash-max-ziplist-value 64
list-max-ziplist-entries 512
list-max-ziplist-value 64
set-max-intset-entries 512
zset-max-ziplist-entries 128
zset-max-ziplist-value 64
hll-sparse-max-bytes 3000
activerehashing yes
client-output-buffer-limit normal 0 0 0
client-output-buffer-limit slave 1gb 256mb 600
client-output-buffer-limit pubsub 32mb 8mb 60
hz 10
sentinel日志报错：
31307:X 31 Oct 20:23:40.307 # +sdown master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:23:40.399 # +new-epoch 1
31307:X 31 Oct 20:23:40.399 # +vote-for-leader 36b0a741a9288f4ff5e26441c748a82e0b476e5e 1
31307:X 31 Oct 20:23:41.397 # +odown master redis-config-1 127.0.0.1 6666 #quorum 3/2
31307:X 31 Oct 20:23:41.397 # Next failover delay: I will not start a failover before Wed Oct 31 20:25:41 2018
31307:X 31 Oct 20:25:40.546 # +new-epoch 2
31307:X 31 Oct 20:25:40.546 # +vote-for-leader 36b0a741a9288f4ff5e26441c748a82e0b476e5e 2
31307:X 31 Oct 20:25:40.607 # Next failover delay: I will not start a failover before Wed Oct 31 20:27:40 2018
31307:X 31 Oct 20:27:40.592 # +new-epoch 3
31307:X 31 Oct 20:27:40.592 # +try-failover master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:27:40.592 # +vote-for-leader 4e0d6414c879007e3b55e8517bd9c319a9ae6da5 3
31307:X 31 Oct 20:27:40.593 # 36b0a741a9288f4ff5e26441c748a82e0b476e5e voted for 4e0d6414c879007e3b55e8517bd9c319a9ae6d
a5 3
31307:X 31 Oct 20:27:40.593 # d7cd4ee54fcc63cf3b0419daa065c836b4d56634 voted for 4e0d6414c879007e3b55e8517bd9c319a9ae6d
a5 3
31307:X 31 Oct 20:27:40.654 # +elected-leader master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:27:40.654 # +failover-state-select-slave master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:27:40.721 # +selected-slave slave 127.0.0.1:6667 127.0.0.1 6667 @ redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:27:40.721 * +failover-state-send-slaveof-noone slave 127.0.0.1:6667 127.0.0.1 6667 @ redis-config-1 1
27.0.0.1 6666
31307:X 31 Oct 20:27:40.804 * +failover-state-wait-promotion slave 127.0.0.1:6667 127.0.0.1 6667 @ redis-config-1 127.0
.0.1 6666
31307:X 31 Oct 20:28:40.867 # -failover-abort-slave-timeout master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:28:40.957 # Next failover delay: I will not start a failover before Wed Oct 31 20:29:40 2018
31307:X 31 Oct 20:28:51.194 * +reboot master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:28:51.277 # -sdown master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:28:51.277 # -odown master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:30:02.674 # +sdown master redis-config-1 127.0.0.1 6666
31307:X 31 Oct 20:30:02.691 # +new-epoch 4
即master已经处于odown，但failover一直没发生切换，sentinel节点无法向redis节点发出slaveof no one命令。
尝试解决：
1. 更新redis版本，由4.0.10 --> 4.0.11
2. 在线下机器部署测试。
但都无法解决。
====================== 解决了 ======================
咳咳咳～ 去掉redis.conf 中rename-command 配置后，重启redis节点主从切换正常。
最重要的是 rename-command CONFIG _CONFIG 这个。
总结：前期测试sentinel/cluster 尽量不要改动默认配置，除了基本配置信息。
后期可以根据实际需要修改配置。
