replication-manager icon indicating copy to clipboard operation
replication-manager copied to clipboard

GTID enabled,does'nt support percona and and oracle mysql

Open tangweichun opened this issue 6 years ago • 5 comments

error log like below 2018/07/05 15:31:33 ERROR Failed to autojoin indirect slave server 172.17.5.101:3307, stopping slave as a precaution Change master statement CHANGE MASTER TO master_host='172.17.5.12', master_port=3307, master_user='repl', master_password='mysql', master_connect_retry=5, master_heartbeat_period=3, MASTER_USE_GTID=SLAVE_POS failed, reason: Error 1064: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'MASTER_USE_GTID=SLAVE_POS' at line 1 2018/07/05 15:31:33 INFO Rejoining slave via GTID 2018/07/05 15:31:33 INFO Found slave to rejoin 172.17.5.101:3307 slave was previously in state SlaveErr replication io thread {No %!s(bool=true)}, pointing currently to 172.17.5.12:3307 2018/07/05 15:31:33 INFO Crash info on current master &{172.17.5.201:3307 mysql-bin.000011 947336839 mysql-bin.000011 952931766 %!s(bool=false) %!s(*gtid.List=) 172.17.5.12:3307}

tangweichun avatar Jul 05 '18 07:07 tangweichun

5.7.21-log MySQL Community Server (GPL)

my configuration: cat /etc/replication-manager/config.toml [Default] title = "Cluster1" log-file = "/var/log/replication-manager.log" db-servers-hosts = "172.17.5.12:3307,172.17.5.101:3307,172.17.5.201:3307" db-servers-credential = "root:xxx" replication-credential = "repl:xxx" failover-mode = "manual"

##########

HTTP

##########

http-server = true http-bind-address = "0.0.0.0"

tangweichun avatar Jul 05 '18 08:07 tangweichun

Hi i think i fixed this recently What is the release you are using ?

svaroqui avatar Jul 05 '18 08:07 svaroqui

Ok found it tx, i 'll try to push a fix

svaroqui avatar Jul 05 '18 08:07 svaroqui

Can you confirm me that you shutdown the slave than shutdown the master , the master failover to an other slave and than restart the stopped slave when this scenario was trigger ?

svaroqui avatar Jul 05 '18 08:07 svaroqui

that's great! release:2.0.1-1-gaea9c25e

scenario:

  1. use sysbench make data on master(172.17.5.201),make a big lag between the master and slave;
  2. shutdown the master(172.17.5.201)
  3. use replication manager to do a failover,then error occurs

tangweichun avatar Jul 05 '18 08:07 tangweichun