mariadb/mysql-test/suite/galera/r/galera_bf_kill_debug.result
sjaakola 7bffe468b2 MDEV-21910 Deadlock between BF abort and manual KILL command
When high priority replication slave applier encounters lock conflict in innodb,
it will force the conflicting lock holder transaction (victim) to rollback.
This is a must in multi-master sychronous replication model to avoid cluster lock-up.
This high priority victim abort (aka "brute force" (BF) abort), is started
from innodb lock manager while holding the victim's transaction's (trx) mutex.
Depending on the execution state of the victim transaction, it may happen that the
BF abort will call for THD::awake() to wake up the victim transaction for the rollback.
Now, if BF abort requires THD::awake() to be called, then the applier thread executed
locking protocol of: victim trx mutex -> victim THD::LOCK_thd_data

If, at the same time another DBMS super user issues KILL command to abort the same victim,
it will execute locking protocol of: victim THD::LOCK_thd_data  -> victim trx mutex.
These two locking protocol acquire mutexes in opposite order, hence unresolvable mutex locking
deadlock may occur.

The fix in this commit adds THD::wsrep_aborter flag to synchronize who can kill the victim
This flag is set both when BF is called for from innodb and by KILL command.
Either path of victim killing will bail out if victim's wsrep_killed is already
set to avoid mutex conflicts with the other aborter execution. THD::wsrep_aborter
records the aborter THD's ID. This is needed to preserve the right to kill
the victim from different locations for the same aborter thread.
It is also good error logging, to see who is reponsible for the abort.

A new test case was added in galera.galera_bf_kill_debug.test for scenario where
wsrep applier thread and manual KILL command try to kill same idle victim
2020-07-22 08:20:10 +03:00

54 lines
1.6 KiB
Text

connection node_2;
connection node_1;
connection node_2;
CREATE TABLE t1(a int not null primary key auto_increment,b int) engine=InnoDB;
insert into t1 values (NULL,1);
connect node_2a, 127.0.0.1, root, , test, $NODE_MYPORT_2;
connection node_2a;
truncate t1;
insert into t1 values (1,0);
begin;
update t1 set b=2 where a=1;
connection node_2;
set session wsrep_sync_wait=0;
connect node_2b, 127.0.0.1, root, , test, $NODE_MYPORT_2;
connection node_2b;
SET GLOBAL debug_dbug = "d,sync.before_wsrep_thd_abort";
connection node_1;
select * from t1;
a b
1 0
update t1 set b= 1 where a=1;
connection node_2b;
SET SESSION DEBUG_SYNC = "now WAIT_FOR sync.before_wsrep_thd_abort_reached";
connection node_2;
SET DEBUG_SYNC= 'before_awake_no_mutex SIGNAL awake_reached WAIT_FOR continue_kill';
connection node_2b;
SET DEBUG_SYNC='now WAIT_FOR awake_reached';
SET GLOBAL debug_dbug = "";
SET DEBUG_SYNC = "now SIGNAL signal.before_wsrep_thd_abort";
SET DEBUG_SYNC = "now SIGNAL continue_kill";
connection node_2;
connection node_2a;
select * from t1;
connection node_2;
SET DEBUG_SYNC = "RESET";
drop table t1;
disconnect node_2a;
connect node_2a, 127.0.0.1, root, , test, $NODE_MYPORT_2;
connection node_2a;
CREATE TABLE t1 (i int primary key);
SET DEBUG_SYNC = "before_wsrep_ordered_commit SIGNAL bwoc_reached WAIT_FOR bwoc_continue";
INSERT INTO t1 VALUES (1);
connection node_2;
SET DEBUG_SYNC = "now WAIT_FOR bwoc_reached";
SET DEBUG_SYNC = "now SIGNAL bwoc_continue";
SET DEBUG_SYNC='RESET';
connection node_2a;
connection node_2;
select * from t1;
i
1
disconnect node_2a;
connection node_1;
drop table t1;