Replica選擇切換機制

• 先剔除不健康的Replica

  • Replica與Master失去連線時間，或是資料落後Master太久的都會優先背剔除

• 依照Replica-priority進行排序

  • 依照Redis.conf中設定的組態，越小越優先

• 資料同步複製處理之同步差

  • 如果Replica-priority相同，則會選擇與Master資料同步差異最小的優先選擇

• RunID

  • 如果Replica-priority相同且資料同步差也相同，則選擇RunID最小的．
  • RunID越小代表Redis Server越先啟動的．

• 在開始測試Redis Server異常自動切換Master前，先確認目前Sentinel叢集狀態．

# 連線到sentinel叢集
# 目前監控的redis server master是mymaster中的 IP: 127.0.0.1 Port: 6380
# Sentinel 26379 狀態
./redis-cli -h 127.0.0.1 -p 26379
127.0.0.1:26379> info sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
sentinel_simulate_failure_flags:0
master0:name=mymaster,status=ok,address=127.0.0.1:6380,slaves=2,sentinels=3

# Sentinel 26380 狀態
./redis-cli -h 127.0.0.1 -p 26380
127.0.0.1:26380> info sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
sentinel_simulate_failure_flags:0
master0:name=mymaster,status=ok,address=127.0.0.1:6380,slaves=2,sentinels=3

# Sentinel 26381 狀態
./redis-cli -h 127.0.0.1 -p 26381
127.0.0.1:26381> info sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
sentinel_simulate_failure_flags:0
master0:name=mymaster,status=ok,address=127.0.0.1:6380,slaves=2,sentinels=3

• 連線到目前Redis Server Master確認目前狀態．

# 連線至目前Redis Server (Master) 6380
./redis-cli -h 127.0.0.1 -p 6380
127.0.0.1:6380> info replication
# Replication
role:master
connected_slaves:2
slave0:ip=127.0.0.1,port=6381,state=online,offset=3204281,lag=0
slave1:ip=127.0.0.1,port=6379,state=online,offset=3204281,lag=0
master_failover_state:no-failover
master_replid:ea3bc11d99568d87b0496fb83c6f2ccc58425baa
master_replid2:a30878092cdd31e404af76eaa7a2d024c8ae843e
master_repl_offset:3204281
second_repl_offset:19299
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2155706
repl_backlog_histlen:1048576

• 找出目前Redis Server Ｍaster並讓其異常終止．

# 
ps -ef|grep redis-server

# kill
kill -9 409413

• 終止目前Redis Server Master Process

• Redis Server Replica 6379 log顯示內容

• Redis Server Replica 6381 log顯示內容

• Sentinel 26379 log 顯示偵測到Redis Server Master異常並進行Replica選舉，切換Replica為Master．

• Sentinel 已將Redis Server Replica 6381切換為Master．

# 原Redis Server Master 6380 已偵測到異常
+sdown master mymaster 127.0.0.1 6380

# 投票選取Sentinel 此次進行切換的Leader
+vote-for-leader 6337212980ee71b6db0222c50ad525c94c6c48a8 1

# 選取Sentinel 26380為此次進行切換的Leader
+config-update-from sentinel 6337212980ee71b6db0222c50ad525c94c6c48a8 127.0.0.1 26380 @ mymaster 127.0.0.1 6380

# 切換Redis Server Master 6381為新的Ｍaster取代原有的
switch-master mymaster 127.0.0.1 6380 127.0.0.1 6381

• Redis Server Master 6381 顯示已成為Master．

• Redis Server Replica 6379 與新的Redis Server Ｍaster 6381 建立同步．

• 將Redis Server 6380 復原重新加入叢集．

Connecting to MASTER 127.0.0.1:6381
Full resync from master: c2f5749e8daef12058d98addb48e954afbc92698:1306638

• Sentinel 會將 Redis Server 6380角色切換為Replica．

• 確認結果是否符合異常自動切換的高可用情境．

# 確認Redis Server 6380角色為Replica
./redis-cli -h 127.0.0.1 -p 6380
127.0.0.1:6380> info replication
# Replication
role:slave
master_host:127.0.0.1
master_port:6381
master_link_status:up
master_last_io_seconds_ago:0
master_sync_in_progress:0
slave_repl_offset:1377655
slave_priority:100
slave_read_only:1
replica_announced:1
connected_slaves:0
master_failover_state:no-failover
master_replid:c2f5749e8daef12058d98addb48e954afbc92698
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:1377655
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1306639
repl_backlog_histlen:71017

# 確認Sentinel與Redis Server都已恢復
./redis-cli -h 127.0.0.1 -p 26379
127.0.0.1:26379> info sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
sentinel_simulate_failure_flags:0
master0:name=mymaster,status=ok,address=127.0.0.1:6381,slaves=2,sentinels=3