Redis REPL-DISKLESS-SYNC-DELAY Parameter

레디스 설정 파일인 redis.conf 에 있는 repl-diskless-sync-delay 파라미터에 대한 설명입니다.  

설명

Diskless 방식으로 복제 서버와 동기화를 하도록 설정했을때, 마스터 서버는 거의 비슷한 시간에 여러 복제에서 요청이 들어올 것을 대비해서 첫 번째 동기화 요청이 들어온 후 설정한 시간(초)를 기다린다.   그 시간안에 다른 복제 서버에서 요청이 들어오면 마스터 서버는 요청이 접수된 복제 서버 여러 대에 동시에 소켓 통신으로 전송한다.
두 번째 요청이 들어오면 다시 5초를 기다린다.
기본값은 5(초)이다.   첫 번재 복제 서버 입장에서는 요청 후 5초 후에 동기화 데이터를 받기 시작할 것이다. 즉시 받도록 하려면 0으로 설정하면 된다. 마스터 서버 당 복제 서버의 개수와 네크워크 상태를 고려해서 정하면 될 것이다.
마스터 서버 로그 3537:M 18:11:17.464 * Replica 192.168.56.102:6003 asks for synchronization 3537:M 18:11:17.464 * Delay next BGSAVE for diskless SYNC 3537:M 18:11:17.464 * Partial resynchronization not accepted: Replication ID mismatch (Replica asked for 'cc95d64e1e28d0a7e69c86af13bad772dbb4ca84', my replication IDs are '7b5402f50d07bf5e7b36420bc5baa7f3b02b51e3' and '0000000000000000000000000000000000000000') 3537:M 18:11:18.987 * Replica 192.168.56.102:6004 asks for synchronization 3537:M 18:11:18.987 * Partial resynchronization not accepted: Replication ID mismatch (Replica asked for '9b59afcf26fed76cbf73afd6ffc9bf8a8ec28b1b', my replication IDs are '7b5402f50d07bf5e7b36420bc5baa7f3b02b51e3' and '0000000000000000000000000000000000000000') 3537:M 18:11:18.987 * Delay next BGSAVE for diskless SYNC 3537:M 18:11:23.510 * Starting BGSAVE for SYNC with target: replicas sockets 3537:M 18:11:23.512 * Background RDB transfer started by pid 4981 4981:C 18:11:25.852 * RDB: 0 MB of memory used by copy-on-write 3537:M 18:11:25.892 * Background RDB transfer terminated with success 3537:M 18:11:25.892 # Slave 192.168.56.102:6003 correctly received the streamed RDB file. 3537:M 18:11:25.892 # Slave 192.168.56.102:6004 correctly received the streamed RDB file. 3537:M 18:11:25.892 * Streamed RDB transfer with replica 192.168.56.102:6003 succeeded (socket). Waiting for REPLCONF ACK from slave to enable streaming 3537:M 18:11:25.892 * Streamed RDB transfer with replica 192.168.56.102:6004 succeeded (socket). Waiting for REPLCONF ACK from slave to enable streaming 3537:M 18:11:30.020 * Synchronization with replica 192.168.56.102:6004 succeeded 3537:M 18:11:30.422 * Synchronization with replica 192.168.56.102:6003 succeeded 복제 서버 로그: 6004 4622:S 18:11:18.734 * Before turning into a replica, using my master parameters to synthesize a cached master: I may be able to synchronize with the new master with just a partial transfer. 4622:S 18:11:18.735 * REPLICAOF 192.168.56.102:6001 enabled (user request from 'id=3 addr=127.0.0.1:47723 fd=8 name= age=54 idle=0 flags=N db=0 sub=0 psub=0 multi=-1 qbuf=50 qbuf-free=32718 obl=0 oll=0 omem=0 events=r cmd=replicaof') 4622:S 18:11:18.985 * Connecting to MASTER 192.168.56.102:6001 4622:S 18:11:18.986 * MASTER <-> REPLICA sync started 4622:S 18:11:18.986 * Non blocking connect for SYNC fired the event. 4622:S 18:11:18.986 * Master replied to PING, replication can continue... 4622:S 18:11:18.987 * Trying a partial resynchronization (request 9b59afcf26fed76cbf73afd6ffc9bf8a8ec28b1b:1). 4622:S 18:11:23.513 * Full resync from master: 7b5402f50d07bf5e7b36420bc5baa7f3b02b51e3:616 4622:S 18:11:23.513 * Discarding previously cached master state. 4622:S 18:11:23.526 * MASTER <-> REPLICA sync: receiving streamed RDB from master 4622:S 18:11:25.963 * MASTER <-> REPLICA sync: Flushing old data 4622:S 18:11:25.963 * MASTER <-> REPLICA sync: Loading DB in memory 4622:S 18:11:29.719 * MASTER <-> REPLICA sync: Finished with success 복제에 대한 자세한 내용은 여기를 보세요.

사용 방법