Redisのレプリケーション設定

概要

スレーブのコンフィグにマスターの情報を記述して、Redisを再起動するだけ。
なおスレーブは読み込み専用になり、書き込みはできなくなる。

127.0.0.1:6379> set key1 val1
(error) READONLY You can't write against a read only slave.

障害時やメンテナンスでフェイルオーバーさせる場合は、手動でマスター/スレーブの情報を変更(slaveof no oneでスレーブ解除、slaveof new_master_ip new_master_portで新マスター指定)するか、sentinelを使う必要があるっぽい。

スレーブ側のconfig修正

• slaveof にマスターサーバの情報(IPとPort)を記述する
  

slaveof 192.168.22.11 6379

redis再起動

# systemctl restart redis

確認

マスター側

• レプリケーションの設定確認
  

# redis-cli info replication

# Replication
role:master
connected_slaves:1
slave0:ip=192.168.22.12,port=6379,state=online,offset=56,lag=1
master_replid:9778dff7ac400670b6c48c9f03ebdcb6173ee3dd
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:56
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1
repl_backlog_histlen:56

• ログ確認
  

9716:M 05 Apr 00:19:48.743 * Slave 192.168.22.12:6379 asks for synchronization
9716:M 05 Apr 00:19:48.743 * Full resync requested by slave 192.168.22.12:6379
9716:M 05 Apr 00:19:48.743 * Starting BGSAVE for SYNC with target: disk
9716:M 05 Apr 00:19:48.743 * Background saving started by pid 9763
9763:C 05 Apr 00:19:48.745 * DB saved on disk
9763:C 05 Apr 00:19:48.745 * RDB: 0 MB of memory used by copy-on-write
9716:M 05 Apr 00:19:48.751 * Background saving terminated with success
9716:M 05 Apr 00:19:48.751 * Synchronization with slave 192.168.22.12:6379 succeeded

スレーブ側

• レプリケーションの設定確認
  

# redis-cli info replication

# Replication
role:slave
master_host:192.168.22.11
master_port:6379
master_link_status:up
master_last_io_seconds_ago:10
master_sync_in_progress:0
slave_repl_offset:28
slave_priority:100
slave_read_only:1
connected_slaves:0
master_replid:9778dff7ac400670b6c48c9f03ebdcb6173ee3dd
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:28
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1
repl_backlog_histlen:28

• ログ確認
  

8468:S 04 Apr 15:12:34.078 * Connecting to MASTER 192.168.22.11:6379
8468:S 04 Apr 15:12:34.079 * MASTER <-> SLAVE sync started
8468:S 04 Apr 15:12:34.079 * Non blocking connect for SYNC fired the event.
8468:S 04 Apr 15:12:34.080 * Master replied to PING, replication can continue...
8468:S 04 Apr 15:12:34.081 * Partial resynchronization not possible (no cached master)
8468:S 04 Apr 15:12:34.082 * Full resync from master: 9778dff7ac400670b6c48c9f03ebdcb6173ee3dd:0
8468:S 04 Apr 15:12:34.144 * MASTER <-> SLAVE sync: receiving 175 bytes from master
8468:S 04 Apr 15:12:34.145 * MASTER <-> SLAVE sync: Flushing old data
8468:S 04 Apr 15:12:34.145 * MASTER <-> SLAVE sync: Loading DB in memory
8468:S 04 Apr 15:12:34.145 * MASTER <-> SLAVE sync: Finished with success