1
0
mirror of https://github.com/MariaDB/server.git synced 2025-07-30 16:24:05 +03:00

MDEV-9573 'Stop slave' hangs on replication slave

The reason for this is that stop slave takes LOCK_active_mi over the
whole operation while some slave operations will also need LOCK_active_mi
which causes deadlocks.

Fixed by introducing object counting for Master_info and not taking
LOCK_active_mi over stop slave or even stop_all_slaves()

Another benefit of this approach is that it allows:
- Multiple threads can run SHOW SLAVE STATUS at the same time
- START/STOP/RESET/SLAVE STATUS on a slave will not block other slaves
- Simpler interface for handling get_master_info()
- Added some missing unlock of 'log_lock' in error condtions
- Moved rpl_parallel_inactivate_pool(&global_rpl_thread_pool) to end
  of stop_slave() to not have to use LOCK_active_mi inside
  terminate_slave_threads()
- Changed argument for remove_master_info() to Master_info, as we always
  have this available
- Fixed core dump when doing FLUSH TABLES WITH READ LOCK and parallel
  replication. Problem was that waiting for pause_for_ftwrl was not done
  when deleting rpt->current_owner after a force_abort.
This commit is contained in:
Monty
2017-01-29 22:10:56 +02:00
committed by Sergei Golubchik
parent d5c54f3990
commit e65f667bb6
13 changed files with 539 additions and 345 deletions

View File

@ -174,24 +174,20 @@ bool reload_acl_and_cache(THD *thd, unsigned long long options,
slave is not likely to have the same connection names.
*/
tmp_write_to_binlog= 0;
mysql_mutex_lock(&LOCK_active_mi);
if (master_info_index)
if (!(mi= (get_master_info(&connection_name,
Sql_condition::WARN_LEVEL_ERROR))))
{
if (!(mi= (master_info_index->
get_master_info(&connection_name,
Sql_condition::WARN_LEVEL_ERROR))))
{
result= 1;
}
else
{
mysql_mutex_lock(&mi->data_lock);
if (rotate_relay_log(mi))
*write_to_binlog= -1;
mysql_mutex_unlock(&mi->data_lock);
}
result= 1;
}
else
{
mysql_mutex_lock(&mi->data_lock);
if (rotate_relay_log(mi))
*write_to_binlog= -1;
mysql_mutex_unlock(&mi->data_lock);
mi->release();
}
mysql_mutex_unlock(&LOCK_active_mi);
#endif
}
#ifdef HAVE_QUERY_CACHE
@ -349,27 +345,33 @@ bool reload_acl_and_cache(THD *thd, unsigned long long options,
LEX_MASTER_INFO* lex_mi= &thd->lex->mi;
Master_info *mi;
tmp_write_to_binlog= 0;
mysql_mutex_lock(&LOCK_active_mi);
if (master_info_index)
if (!(mi= get_master_info(&lex_mi->connection_name,
Sql_condition::WARN_LEVEL_ERROR)))
{
if (!(mi= (master_info_index->
get_master_info(&lex_mi->connection_name,
Sql_condition::WARN_LEVEL_ERROR))))
{
result= 1;
}
else if (reset_slave(thd, mi))
result= 1;
}
else
{
/* The following will fail if slave is running */
if (reset_slave(thd, mi))
{
mi->release();
/* NOTE: my_error() has been already called by reset_slave(). */
result= 1;
}
else if (mi->connection_name.length && thd->lex->reset_slave_info.all)
{
/* If not default connection and 'all' is used */
master_info_index->remove_master_info(&mi->connection_name);
mi->release();
mysql_mutex_lock(&LOCK_active_mi);
if (master_info_index->remove_master_info(mi))
result= 1;
mysql_mutex_unlock(&LOCK_active_mi);
}
else
mi->release();
}
mysql_mutex_unlock(&LOCK_active_mi);
}
#endif
if (options & REFRESH_USER_RESOURCES)