mysql - Nodes unexpectedly closing in Mariadb Galera Cluster -
i've setup 3 node galera cluster application, there stored procedure in application creates table temporarily(not temporary table), table created dynamically , executed. crud operation done on table muptiple times , after usage completed dropped @ end of stored procedure. problem everytime run particular stored procedure 2 out of 3 node fails , shuts down automatically. here log node fails :-
05 seqnos (l: 38, g: 47, s: 46, d: 46, ts: 186965868554819) 150423 8:31:35 [error] wsrep: failed apply trx 47 4 times 150423 8:31:35 [error] wsrep: node consistency compromized, aborting... 150423 8:31:35 [note] wsrep: closing send monitor... 150423 8:31:35 [note] wsrep: closed send monitor. 150423 8:31:35 [note] wsrep: gcomm: terminating thread 150423 8:31:35 [note] wsrep: gcomm: joining thread 150423 8:31:35 [note] wsrep: gcomm: closing backend 150423 8:31:35 [note] wsrep: view(view_id(non_prim,4d2adf77-e972-11e4-be7e-6bcb3d5c882f,3) memb { 849bbb26-e981-11e4-9f9f-f37ebca03acf,0 } joined { } left { } partitioned { 4d2adf77-e972-11e4-be7e-6bcb3d5c882f,0 634f32cb-e972-11e4-a226-bf661daf79d7,0 }) 150423 8:31:35 [note] wsrep: view((empty)) 150423 8:31:35 [note] wsrep: new component: primary = no, bootstrap = no, my_idx = 0, memb_num = 1 150423 8:31:35 [note] wsrep: gcomm: closed 150423 8:31:35 [note] wsrep: flow-control interval: [16, 16] 150423 8:31:35 [note] wsrep: received non-primary. 150423 8:31:35 [note] wsrep: shifting synced -> open (to: 47) 150423 8:31:35 [note] wsrep: received self-leave message. 150423 8:31:35 [note] wsrep: flow-control interval: [0, 0] 150423 8:31:35 [note] wsrep: received self-leave. closing connection. 150423 8:31:35 [note] wsrep: shifting open -> closed (to: 47) 150423 8:31:35 [note] wsrep: recv thread exiting 0: success 150423 8:31:35 [note] wsrep: recv_thread() joined. 150423 8:31:35 [note] wsrep: closing replication queue. 150423 8:31:35 [note] wsrep: closing slave action queue. 150423 8:31:35 [note] wsrep: /usr/sbin/mysqld: terminated. 150423 08:31:35 mysqld_safe number of processes running now: 0 150423 08:31:35 mysqld_safe wsrep: not restarting wsrep node automatically 150423 08:31:35 mysqld_safe mysqld pid file /mnt/mysql/db3-prod.pid ended
is there particular can solve without modifying internal structure of stored procedure.
the output of
show variables 'default%'; +---------------------------+--------+ | variable_name | value | +---------------------------+--------+ | default_master_connection | | | default_regex_flags | | | default_storage_engine | innodb | | default_week_format | 0 | +---------------------------+--------+