I tried to reproduce with lastest PXC version 5.6.22-72.0-56 ercona XtraDB Cluster (GPL), Release rel72.0, Revision 978, WSREP version 25.8, wsrep_25.8.r4150. After deleting record on master i.e. writer node I got below error on other nodes of the cluster 2015-06-09 05:49:25 21598 [ERROR] Slave SQL: Could not execute Delete_rows event on table test.a; Can't find record in 'a', Error_code: 1032; handler error HA_ERR_END_OF_FILE; the event's master log FIRST, end_log_pos 149, Error_code: 1032 2015-06-09 05:49:25 21598 [Warning] WSREP: RBR event 3 Delete_rows apply warning: 137, 48483717 2015-06-09 05:49:25 21598 [Warning] WSREP: Failed to apply app buffer: seqno: 48483717, status: 1 at galera/src/trx_handle.cpp:apply():351 Retrying 2th time 2015-06-09 05:49:25 21598 [ERROR] Slave SQL: Could not execute Delete_rows event on table test.a; Can't find record in 'a', Error_code: 1032; handler error HA_ERR_END_OF_FILE; the event's master log FIRST, end_log_pos 149, Error_code: 1032 2015-06-09 05:49:25 21598 [Warning] WSREP: RBR event 3 Delete_rows apply warning: 137, 48483717 2015-06-09 05:49:25 21598 [Warning] WSREP: Failed to apply app buffer: seqno: 48483717, status: 1 at galera/src/trx_handle.cpp:apply():351 Retrying 3th time 2015-06-09 05:49:25 21598 [ERROR] Slave SQL: Could not execute Delete_rows event on table test.a; Can't find record in 'a', Error_code: 1032; handler error HA_ERR_END_OF_FILE; the event's master log FIRST, end_log_pos 149, Error_code: 1032 2015-06-09 05:49:25 21598 [Warning] WSREP: RBR event 3 Delete_rows apply warning: 137, 48483717 2015-06-09 05:49:25 21598 [Warning] WSREP: Failed to apply app buffer: seqno: 48483717, status: 1 at galera/src/trx_handle.cpp:apply():351 Retrying 4th time 2015-06-09 05:49:25 21598 [ERROR] Slave SQL: Could not execute Delete_rows event on table test.a; Can't find record in 'a', Error_code: 1032; handler error HA_ERR_END_OF_FILE; the event's master log FIRST, end_log_pos 149, Error_code: 1032 2015-06-09 05:49:25 21598 [Warning] WSREP: RBR event 3 Delete_rows apply warning: 137, 48483717 2015-06-09 05:49:25 21598 [ERROR] WSREP: Failed to apply trx: source: 6697faad-0e8b-11e5-b4d9-3f7722481adf version: 3 local: 0 state: APPLYING flags: 1 conn_id: 13 trx_id: 141946374 seqnos (l: 5, g: 48483717, s: 48483716, d: 48483716, ts: 5046077371079801) 2015-06-09 05:49:25 21598 [ERROR] WSREP: Failed to apply trx 48483717 4 times 2015-06-09 05:49:25 21598 [ERROR] WSREP: Node consistency compromized, aborting... 2015-06-09 05:49:25 21598 [Note] WSREP: Closing send monitor... 2015-06-09 05:49:25 21598 [Note] WSREP: Closed send monitor. 2015-06-09 05:49:25 21598 [Note] WSREP: gcomm: terminating thread 2015-06-09 05:49:25 21598 [Note] WSREP: gcomm: joining thread 2015-06-09 05:49:25 21598 [Note] WSREP: gcomm: closing backend 2015-06-09 05:49:25 21598 [Note] WSREP: view(view_id(NON_PRIM,6697faad,3) memb { be72198a,0 } joined { } left { } partitioned { 6697faad,102 9a8c7599,0 }) 2015-06-09 05:49:25 21598 [Note] WSREP: view((empty)) 2015-06-09 05:49:25 21598 [Note] WSREP: New COMPONENT: primary = no, bootstrap = no, my_idx = 0, memb_num = 1 2015-06-09 05:49:25 21598 [Note] WSREP: gcomm: closed 2015-06-09 05:49:25 21598 [Note] WSREP: Flow-control interval: [16, 16] 2015-06-09 05:49:25 21598 [Note] WSREP: Received NON-PRIMARY. 2015-06-09 05:49:25 21598 [Note] WSREP: Shifting SYNCED -> OPEN (TO: 48483717) 2015-06-09 05:49:25 21598 [Note] WSREP: Received self-leave message. 2015-06-09 05:49:25 21598 [Note] WSREP: Flow-control interval: [0, 0] 2015-06-09 05:49:25 21598 [Note] WSREP: Received SELF-LEAVE. Closing connection. 2015-06-09 05:49:25 21598 [Note] WSREP: Shifting OPEN -> CLOSED (TO: 48483717) 2015-06-09 05:49:25 21598 [Note] WSREP: RECV thread exiting 0: Success 2015-06-09 05:49:25 21598 [Note] WSREP: recv_thread() joined. 2015-06-09 05:49:25 21598 [Note] WSREP: Closing replication queue. 2015-06-09 05:49:25 21598 [Note] WSREP: Closing slave action queue. 2015-06-09 05:49:25 21598 [Note] WSREP: /usr/sbin/mysqld: Terminated. Aborted 150609 05:49:25 mysqld_safe Number of processes running now: 0 150609 05:49:25 mysqld_safe WSREP: not restarting wsrep node automatically 150609 05:49:25 mysqld_safe mysqld from pid file /var/run/mysqld/mysqld.pid ended while on master node i.e. writer node: 2015-06-09 05:49:25 21452 [Note] WSREP: declaring 9a8c7599 at tcp://10.0.3.10:4567 stable 2015-06-09 05:49:25 21452 [Note] WSREP: forgetting be72198a (tcp://10.0.3.169:4567) 2015-06-09 05:49:25 21452 [Note] WSREP: Node 6697faad state prim 2015-06-09 05:49:25 21452 [Warning] WSREP: 6697faad sending install message failed: Resource temporarily unavailable 2015-06-09 05:49:25 21452 [Note] WSREP: view(view_id(NON_PRIM,6697faad,4) memb { 6697faad,102 } joined { } left { } partitioned { 9a8c7599,0 be72198a,0 }) 2015-06-09 05:49:25 21452 [Note] WSREP: forgetting 9a8c7599 (tcp://10.0.3.10:4567) 2015-06-09 05:49:25 21452 [Note] WSREP: New COMPONENT: primary = no, bootstrap = no, my_idx = 0, memb_num = 1 2015-06-09 05:49:25 21452 [Note] WSREP: view(view_id(NON_PRIM,6697faad,5) memb { 6697faad,102 } joined { } left { } partitioned { 9a8c7599,0 be72198a,0 }) 2015-06-09 05:49:25 21452 [Note] WSREP: Flow-control interval: [16, 16] 2015-06-09 05:49:25 21452 [Note] WSREP: Received NON-PRIMARY. 2015-06-09 05:49:25 21452 [Note] WSREP: Shifting SYNCED -> OPEN (TO: 48483717) 2015-06-09 05:49:25 21452 [Note] WSREP: New COMPONENT: primary = no, bootstrap = no, my_idx = 0, memb_num = 1 2015-06-09 05:49:25 21452 [Note] WSREP: Flow-control interval: [16, 16] 2015-06-09 05:49:25 21452 [Note] WSREP: Received NON-PRIMARY. 2015-06-09 05:49:25 21452 [Note] WSREP: New cluster view: global state: 51476e75-5313-11e4-a0c8-c21460d01a80:48483717, view# -1: non-Primary, number of nodes: 1, my index: 0, protocol version 3 2015-06-09 05:49:25 21452 [Note] WSREP: Setting wsrep_ready to 0 2015-06-09 05:49:25 21452 [Note] WSREP: wsrep_notify_cmd is not defined, skipping notification. 2015-06-09 05:49:25 21452 [Note] WSREP: New cluster view: global state: 51476e75-5313-11e4-a0c8-c21460d01a80:48483717, view# -1: non-Primary, number of nodes: 1, my index: 0, protocol version 3 2015-06-09 05:49:25 21452 [Note] WSREP: Setting wsrep_ready to 0 2015-06-09 05:49:25 21452 [Note] WSREP: wsrep_notify_cmd is not defined, skipping notification. I think it seems normal as after deleting record on writer node remaining nodes of cluster loosing quorum and writer node can't operate any more alone. As far as data inconsistency concerned all needed message and GRA files are logged on the slave nodes (non-writer nodes). You have more messages on writer node because of debug messages enabled.