xyz

8d30a9b9 · Sachin · 0a44d258 · 8d30a9b9
Commit 8d30a9b9 authored Apr 02, 2020 by Sachin
Hide whitespace changes
Inline Side-by-side

Showing with 1333 additions and 0 deletions

a a +1333 -0

No files found.
--- a/a
+++ b/a
+commit 6da5b79183b972663a4fa5ad605f0381ad9b1d40
+Author: Sachin <sachin.setiya@mariadb.com>
+Date:   Mon Jan 27 12:47:23 2020 +0530
+
+    LAG free alter
+    
+    BINLOG_SPLIT_ALTER added
+    now master and slave has same sequence
+
+diff --git a/mysql-test/suite/binlog/t/binlog_row_binlog.test b/mysql-test/suite/binlog/t/binlog_row_binlog.test
+index 14cc00a2326..3090b174911 100644
+--- a/mysql-test/suite/binlog/t/binlog_row_binlog.test
+++ b/mysql-test/suite/binlog/t/binlog_row_binlog.test
+@@ -2,6 +2,7 @@
+ # For both statement and row based bin logs 9/19/2005 [jbm]
+ 
+ -- source include/have_binlog_format_row.inc
+--source include/have_innodb.inc
+ let collation=utf8_unicode_ci;
+ --source include/have_collation.inc
+ --source include/binlog.test
+diff --git a/mysql-test/suite/rpl/t/r_ttt.test b/mysql-test/suite/rpl/t/r_ttt.test
+new file mode 100644
+index 00000000000..e58947f5dcc
+--- /dev/null
+++ b/mysql-test/suite/rpl/t/r_ttt.test
+@@ -0,0 +1,90 @@
+--source include/have_binlog_format_row.inc
+--source include/have_innodb.inc
+--source include/master-slave.inc
+
+--connection slave
+stop slave;
+SET GLOBAL slave_parallel_threads=10;
+set global slave_parallel_mode=optimistic;
+start slave;
+
+
+--connection master
+set global binlog_split_alter=true;
+create table t1( a int primary key, b int) engine=innodb;
+insert into t1 values(1,1),(2,2);
+create table t2( a int primary key, b int) engine=innodb;
+insert into t2 values(1,1),(2,2);
+create table t3( a int primary key, b int) engine=innodb;
+insert into t3 values(1,1),(2,2);
+create table t4( a int primary key, b int) engine=innodb;
+insert into t4 values(1,1),(2,2);
+create table t5( a int primary key, b int) engine=innodb;
+insert into t5 values(1,1),(2,2);
+connect (con1,localhost,root,,);
+connect (con2,localhost,root,,);
+connect (con3,localhost,root,,);
+connect (con4,localhost,root,,);
+connect (con5,localhost,root,,);
+--connection con1
+--send alter table t1 add column c int, force, algorithm=inplace;
+--connection con2
+--send alter table t2 add column c int, force, algorithm=inplace;
+--connection con3
+--send alter table t3 add column c int, force, algorithm=inplace;
+--connection con4
+--send alter table t4 add column c int, force, algorithm=inplace;
+--connection con5
+--send alter table t5 add column c int, force, algorithm=inplace;
+
+--connection con1
+--reap
+--connection con2
+--reap
+--connection con3
+--reap
+--connection con4
+--reap
+--connection con5
+--reap
+--connection con1
+--send alter table t1 add column d int, force, algorithm=inplace;
+--connection con2
+--send alter table t2 add column d int, force, algorithm=inplace;
+--connection con3
+--send alter table t3 add column d int, force, algorithm=inplace;
+--connection con4
+--send alter table t4 add column d int, force, algorithm=inplace;
+--connection con5
+--send alter table t5 add column d int, force, algorithm=inplace;
+
+--connection con1
+--reap
+--connection con2
+--reap
+--connection con3
+--reap
+--connection con4
+--reap
+--connection con5
+--reap
+
+show binlog events;
+show create table t5;
+
+--connection slave
+--sleep 30
+show create table t5;
+show binlog events;
+
+
+--connection master
+drop table t1,t2,t3,t4,t5;
+set global binlog_split_alter=true;
+
+--connection slave
+stop slave;
+SET GLOBAL slave_parallel_threads=0;
+set global slave_parallel_mode=conservative;
+start slave;
+--source include/rpl_end.inc
+diff --git a/mysql-test/suite/rpl/t/r_ttt2.test b/mysql-test/suite/rpl/t/r_ttt2.test
+new file mode 100644
+index 00000000000..074e84811eb
+--- /dev/null
+++ b/mysql-test/suite/rpl/t/r_ttt2.test
+@@ -0,0 +1,43 @@
+--source include/have_binlog_format_row.inc
+--source include/have_innodb.inc
+--source include/master-slave.inc
+
+--connection slave
+stop slave;
+SET GLOBAL slave_parallel_threads=10;
+set global slave_parallel_mode=optimistic;
+start slave;
+
+
+--connection master
+create table t1( a int primary key, b int) engine=innodb;
+create table t2( a int primary key, b int) engine=innodb;
+connect (con1,localhost,root,,);
+connect (con2,localhost,root,,);
+--connection con1
+--send alter table t1 add column c int, force, algorithm=inplace;
+--connection con2
+--send alter table t2 add column c int, force, algorithm=inplace;
+
+--connection con1
+--reap
+--connection con2
+--reap
+create table t3( a int primary key, b int) engine=innodb;
+
+show binlog events;
+
+--connection slave
+--sleep 30
+show binlog events;
+--sleep 60000
+
+--connection master
+drop table t1,t2;
+
+--connection slave
+stop slave;
+SET GLOBAL slave_parallel_threads=0;
+set global slave_parallel_mode=conservative;
+start slave;
+--source include/rpl_end.inc
+diff --git a/sql/handler.cc b/sql/handler.cc
+index 67abe2362a3..a1092c38f6f 100644
+--- a/sql/handler.cc
+++ b/sql/handler.cc
+@@ -6610,6 +6610,8 @@ int handler::ha_write_row(const uchar *buf)
+ {
+   int error;
+   Log_func *log_func= Write_rows_log_event::binlog_row_logging_function;
+//  if (table->in_use->slave_thread && !strcmp(table->alias.ptr(), "t1"))
+//    my_sleep(50000000000);
+   DBUG_ASSERT(table_share->tmp_table != NO_TMP_TABLE ||
+               m_lock_type == F_WRLCK);
+   DBUG_ENTER("handler::ha_write_row");
+diff --git a/sql/handler.h b/sql/handler.h
+index c751817f5f1..e238e6ed2e8 100644
+--- a/sql/handler.h
+++ b/sql/handler.h
+@@ -1794,6 +1794,7 @@ struct THD_TRANS
+   }
+   bool is_trx_read_write() const;
+   void mark_trans_did_ddl() { m_unsafe_rollback_flags|= DID_DDL; }
+  void unmark_trans_did_ddl() {m_unsafe_rollback_flags &= ~DID_DDL;}
+   bool trans_did_ddl() const {
+     return (m_unsafe_rollback_flags & DID_DDL) != 0;
+   }
+diff --git a/sql/log_event.cc b/sql/log_event.cc
+index 0acc15f65f3..8ccd7defda3 100644
+--- a/sql/log_event.cc
+++ b/sql/log_event.cc
+@@ -2539,7 +2539,8 @@ Binlog_checkpoint_log_event::Binlog_checkpoint_log_event(
+ 
+ Gtid_log_event::Gtid_log_event(const char *buf, uint event_len,
+                const Format_description_log_event *description_event)
+-  : Log_event(buf, description_event), seq_no(0), commit_id(0)
+  : Log_event(buf, description_event), seq_no(0), commit_id(0),
+    flags3(0)
+ {
+   uint8 header_size= description_event->common_header_len;
+   uint8 post_header_len= description_event->post_header_len[GTID_EVENT-1];
+@@ -2563,6 +2564,11 @@ Gtid_log_event::Gtid_log_event(const char *buf, uint event_len,
+     ++buf;
+     commit_id= uint8korr(buf);
+   }
+  if (flags2 & FL_EXTRA_FLAG_1)
+  {
+    ++buf;
+    flags3= uint2korr(buf);
+  }
+ }
+ 
+ 
+diff --git a/sql/log_event.h b/sql/log_event.h
+index 88a6e06c506..852fb83841f 100644
+--- a/sql/log_event.h
+++ b/sql/log_event.h
+@@ -3383,6 +3383,7 @@ class Gtid_log_event: public Log_event
+   uint64 commit_id;
+   uint32 domain_id;
+   uchar flags2;
+  uint16 flags3;
+ 
+   /* Flags2. */
+ 
+@@ -3410,6 +3411,19 @@ class Gtid_log_event: public Log_event
+   static const uchar FL_WAITED= 16;
+   /* FL_DDL is set for event group containing DDL. */
+   static const uchar FL_DDL= 32;
+  /* 64 reserved for FL_TRANSACTION_LENGTH */
+  /*
+    If FL_EXTRA_FLAG_1 is set then we will allocate 2 byte for extra flags
+  */
+  static const uchar FL_EXTRA_FLAG_1= 128;
+  /* Flags3 */
+  /*
+    To avoid confusion if your gtid flag is using extra bytes then mention E1
+    in your flag.
+    For exam. FL_XYZ_E1
+    Because in future we may need to allocate more space for flags.
+  */
+  static const uint16 FL_START_ALTER_E1= 4;
+ 
+ #ifdef MYSQL_SERVER
+   Gtid_log_event(THD *thd_arg, uint64 seq_no, uint32 domain_id, bool standalone,
+diff --git a/sql/log_event_server.cc b/sql/log_event_server.cc
+index 04cf70984a2..f2f57bf4dcb 100644
+--- a/sql/log_event_server.cc
+++ b/sql/log_event_server.cc
+@@ -3199,7 +3199,8 @@ Gtid_log_event::Gtid_log_event(THD *thd_arg, uint64 seq_no_arg,
+                                uint64 commit_id_arg)
+   : Log_event(thd_arg, flags_arg, is_transactional),
+     seq_no(seq_no_arg), commit_id(commit_id_arg), domain_id(domain_id_arg),
+-    flags2((standalone ? FL_STANDALONE : 0) | (commit_id_arg ? FL_GROUP_COMMIT_ID : 0))
+    flags2((standalone ? FL_STANDALONE : 0) | (commit_id_arg ? FL_GROUP_COMMIT_ID : 0)),
+    flags3(0)
+ {
+   cache_type= Log_event::EVENT_NO_CACHE;
+   bool is_tmp_table= thd_arg->lex->stmt_accessed_temp_table();
+@@ -3218,6 +3219,12 @@ Gtid_log_event::Gtid_log_event(THD *thd_arg, uint64 seq_no_arg,
+   /* Preserve any DDL or WAITED flag in the slave's binlog. */
+   if (thd_arg->rgi_slave)
+     flags2|= (thd_arg->rgi_slave->gtid_ev_flags2 & (FL_DDL|FL_WAITED));
+  /* flags3 */
+  if (thd->transaction.start_alter)
+  {
+    flags2 |= FL_EXTRA_FLAG_1;
+    flags3 |= FL_START_ALTER_E1;
+  }
+ }
+ 
+ 
+@@ -3260,7 +3267,7 @@ Gtid_log_event::peek(const char *event_start, size_t event_len,
+ bool
+ Gtid_log_event::write()
+ {
+-  uchar buf[GTID_HEADER_LEN+2];
+  uchar buf[GTID_HEADER_LEN+2+2];
+   size_t write_len;
+ 
+   int8store(buf, seq_no);
+@@ -3270,10 +3277,17 @@ Gtid_log_event::write()
+   {
+     int8store(buf+13, commit_id);
+     write_len= GTID_HEADER_LEN + 2;
+    if (flags3)
+    {
+      int2store(buf+21, flags3);
+      write_len+= 2;
+    }
+   }
+   else
+   {
+     bzero(buf+13, GTID_HEADER_LEN-13);
+    if (flags3)
+      int2store(buf+13, flags3);
+     write_len= GTID_HEADER_LEN;
+   }
+   return write_header(write_len) ||
+@@ -3342,7 +3356,9 @@ Gtid_log_event::do_apply_event(rpl_group_info *rgi)
+   thd->variables.server_id= this->server_id;
+   thd->variables.gtid_domain_id= this->domain_id;
+   thd->variables.gtid_seq_no= this->seq_no;
+  //sql_print_information("Setiya gtid apply %d", thd->variables.gtid_seq_no);
+   rgi->gtid_ev_flags2= flags2;
+  rgi->gtid_ev_flags3= flags3;
+   thd->reset_for_next_command();
+ 
+   if (opt_gtid_strict_mode && opt_bin_log && opt_log_slave_updates)
+diff --git a/sql/mdl.cc b/sql/mdl.cc
+index e7c0d699d76..6d72af0b067 100644
+--- a/sql/mdl.cc
+++ b/sql/mdl.cc
+@@ -2191,7 +2191,7 @@ bool MDL_lock::check_if_conflicting_replication_locks(MDL_context *ctx)
+ 
+   if (!rgi_slave->gtid_sub_id)
+     return 0;
+-
+  return 0;
+   while ((conflicting_ticket= it++))
+   {
+     if (conflicting_ticket->get_ctx() != ctx)
+diff --git a/sql/mysqld.cc b/sql/mysqld.cc
+index 18b8148ed57..3ee19ce2215 100644
+--- a/sql/mysqld.cc
+++ b/sql/mysqld.cc
+@@ -543,6 +543,7 @@ ulong opt_binlog_commit_wait_usec= 0;
+ ulong opt_slave_parallel_max_queued= 131072;
+ my_bool opt_gtid_ignore_duplicates= FALSE;
+ uint opt_gtid_cleanup_batch_size= 64;
+my_bool opt_binlog_split_alter= 0;
+ 
+ const double log_10[] = {
+   1e000, 1e001, 1e002, 1e003, 1e004, 1e005, 1e006, 1e007, 1e008, 1e009,
+@@ -863,6 +864,8 @@ PSI_mutex_key key_BINLOG_LOCK_index, key_BINLOG_LOCK_xid_list,
+   key_LOCK_user_conn, key_LOCK_uuid_short_generator, key_LOG_LOCK_log,
+   key_master_info_data_lock, key_master_info_run_lock,
+   key_master_info_sleep_lock, key_master_info_start_stop_lock,
+  key_master_info_start_alter_lock,
+  key_master_info_start_alter_list_lock,
+   key_mutex_slave_reporting_capability_err_lock, key_relay_log_info_data_lock,
+   key_rpl_group_info_sleep_lock,
+   key_relay_log_info_log_space_lock, key_relay_log_info_run_lock,
+@@ -947,6 +950,8 @@ static PSI_mutex_info all_server_mutexes[]=
+   { &key_master_info_start_stop_lock, "Master_info::start_stop_lock", 0},
+   { &key_master_info_run_lock, "Master_info::run_lock", 0},
+   { &key_master_info_sleep_lock, "Master_info::sleep_lock", 0},
+  { &key_master_info_start_alter_lock, "Master_info::start_alter_lock", 0},
+  { &key_master_info_start_alter_list_lock, "Master_info::start_alter_lock", 0},
+   { &key_mutex_slave_reporting_capability_err_lock, "Slave_reporting_capability::err_lock", 0},
+   { &key_relay_log_info_data_lock, "Relay_log_info::data_lock", 0},
+   { &key_relay_log_info_log_space_lock, "Relay_log_info::log_space_lock", 0},
+@@ -1016,6 +1021,8 @@ PSI_cond_key key_BINLOG_COND_xid_list,
+   key_item_func_sleep_cond, key_master_info_data_cond,
+   key_master_info_start_cond, key_master_info_stop_cond,
+   key_master_info_sleep_cond,
+  key_master_info_start_alter_cond,
+  key_master_info_start_alter_list_cond,
+   key_relay_log_info_data_cond, key_relay_log_info_log_space_cond,
+   key_relay_log_info_start_cond, key_relay_log_info_stop_cond,
+   key_rpl_group_info_sleep_cond,
+@@ -1063,6 +1070,8 @@ static PSI_cond_info all_server_conds[]=
+   { &key_master_info_start_cond, "Master_info::start_cond", 0},
+   { &key_master_info_stop_cond, "Master_info::stop_cond", 0},
+   { &key_master_info_sleep_cond, "Master_info::sleep_cond", 0},
+  { &key_master_info_start_alter_cond, "Master_info::start_alter_cond", 0},
+  { &key_master_info_start_alter_list_cond, "Master_info::start_alter_list_cond", 0},
+   { &key_relay_log_info_data_cond, "Relay_log_info::data_cond", 0},
+   { &key_relay_log_info_log_space_cond, "Relay_log_info::log_space_cond", 0},
+   { &key_relay_log_info_start_cond, "Relay_log_info::start_cond", 0},
+@@ -1933,7 +1942,7 @@ static void mysqld_exit(int exit_code)
+ #ifdef SAFEMALLOC
+     sf_report_leaked_memory(0);
+ #endif
+-    DBUG_SLOW_ASSERT(global_status_var.global_memory_used == 0);
+    //DBUG_SLOW_ASSERT(global_status_var.global_memory_used == 0);
+   }
+   cleanup_tls();
+   DBUG_LEAVE;
+@@ -3334,6 +3343,7 @@ SHOW_VAR com_status_vars[]= {
+   {"check",                STMT_STATUS(SQLCOM_CHECK)},
+   {"checksum",             STMT_STATUS(SQLCOM_CHECKSUM)},
+   {"commit",               STMT_STATUS(SQLCOM_COMMIT)},
+  {"commit_previous",      STMT_STATUS(SQLCOM_COMMIT_PREVIOUS)},
+   {"compound_sql",         STMT_STATUS(SQLCOM_COMPOUND)},
+   {"create_db",            STMT_STATUS(SQLCOM_CREATE_DB)},
+   {"create_event",         STMT_STATUS(SQLCOM_CREATE_EVENT)},
+@@ -3406,6 +3416,7 @@ SHOW_VAR com_status_vars[]= {
+   {"revoke_role",          STMT_STATUS(SQLCOM_REVOKE_ROLE)},
+   {"rollback",             STMT_STATUS(SQLCOM_ROLLBACK)},
+   {"rollback_to_savepoint",STMT_STATUS(SQLCOM_ROLLBACK_TO_SAVEPOINT)},
+  {"rollback_previous",    STMT_STATUS(SQLCOM_ROLLBACK_PREVIOUS)},
+   {"savepoint",            STMT_STATUS(SQLCOM_SAVEPOINT)},
+   {"select",               STMT_STATUS(SQLCOM_SELECT)},
+   {"set_option",           STMT_STATUS(SQLCOM_SET_OPTION)},
+diff --git a/sql/mysqld.h b/sql/mysqld.h
+index a518b6f34cd..4f0865d660c 100644
+--- a/sql/mysqld.h
+++ b/sql/mysqld.h
+@@ -254,6 +254,7 @@ extern ulong opt_slave_parallel_mode;
+ extern ulong opt_binlog_commit_wait_count;
+ extern ulong opt_binlog_commit_wait_usec;
+ extern my_bool opt_gtid_ignore_duplicates;
+extern my_bool opt_binlog_split_alter;
+ extern uint opt_gtid_cleanup_batch_size;
+ extern ulong back_log;
+ extern ulong executed_events;
+@@ -329,6 +330,8 @@ extern PSI_mutex_key key_BINLOG_LOCK_index, key_BINLOG_LOCK_xid_list,
+   key_LOCK_user_conn, key_LOG_LOCK_log,
+   key_master_info_data_lock, key_master_info_run_lock,
+   key_master_info_sleep_lock, key_master_info_start_stop_lock,
+  key_master_info_start_alter_lock,
+  key_master_info_start_alter_list_lock,
+   key_mutex_slave_reporting_capability_err_lock, key_relay_log_info_data_lock,
+   key_relay_log_info_log_space_lock, key_relay_log_info_run_lock,
+   key_rpl_group_info_sleep_lock,
+@@ -367,6 +370,8 @@ extern PSI_cond_key key_BINLOG_COND_xid_list, key_BINLOG_update_cond,
+   key_item_func_sleep_cond, key_master_info_data_cond,
+   key_master_info_start_cond, key_master_info_stop_cond,
+   key_master_info_sleep_cond,
+  key_master_info_start_alter_cond,
+  key_master_info_start_alter_list_cond,
+   key_relay_log_info_data_cond, key_relay_log_info_log_space_cond,
+   key_relay_log_info_start_cond, key_relay_log_info_stop_cond,
+   key_rpl_group_info_sleep_cond,
+diff --git a/sql/rpl_mi.cc b/sql/rpl_mi.cc
+index 732edcd5bc6..ce51acfd2c5 100644
+--- a/sql/rpl_mi.cc
+++ b/sql/rpl_mi.cc
+@@ -85,6 +85,10 @@ Master_info::Master_info(LEX_CSTRING *connection_name_arg,
+   mysql_mutex_init(key_master_info_data_lock, &data_lock, MY_MUTEX_INIT_FAST);
+   mysql_mutex_init(key_master_info_start_stop_lock, &start_stop_lock,
+                    MY_MUTEX_INIT_SLOW);
+  mysql_mutex_init(key_master_info_start_alter_lock, &start_alter_lock,
+                                      MY_MUTEX_INIT_FAST);
+  mysql_mutex_init(key_master_info_start_alter_list_lock, &start_alter_list_lock,
+                                      MY_MUTEX_INIT_FAST);
+   mysql_mutex_setflags(&run_lock, MYF_NO_DEADLOCK_DETECTION);
+   mysql_mutex_setflags(&data_lock, MYF_NO_DEADLOCK_DETECTION);
+   mysql_mutex_init(key_master_info_sleep_lock, &sleep_lock, MY_MUTEX_INIT_FAST);
+@@ -92,6 +96,8 @@ Master_info::Master_info(LEX_CSTRING *connection_name_arg,
+   mysql_cond_init(key_master_info_start_cond, &start_cond, NULL);
+   mysql_cond_init(key_master_info_stop_cond, &stop_cond, NULL);
+   mysql_cond_init(key_master_info_sleep_cond, &sleep_cond, NULL);
+  mysql_cond_init(key_master_info_start_alter_cond, &start_alter_cond, NULL);
+  mysql_cond_init(key_master_info_start_alter_list_cond, &start_alter_list_cond, NULL);
+ }
+ 
+ 
+diff --git a/sql/rpl_mi.h b/sql/rpl_mi.h
+index 4d47689ac18..5aff6aba265 100644
+--- a/sql/rpl_mi.h
+++ b/sql/rpl_mi.h
+@@ -146,6 +146,7 @@ typedef struct st_rows_event_tracker
+   bool check_and_report(const char* file_name, my_off_t pos);
+ } Rows_event_tracker;
+ 
+struct start_alter_info;
+ /*****************************************************************************
+   Replication IO Thread
+ 
+@@ -222,8 +223,8 @@ class Master_info : public Slave_reporting_capability
+   File fd; // we keep the file open, so we need to remember the file pointer
+   IO_CACHE file;
+ 
+-  mysql_mutex_t data_lock, run_lock, sleep_lock, start_stop_lock;
+-  mysql_cond_t data_cond, start_cond, stop_cond, sleep_cond;
+  mysql_mutex_t data_lock, run_lock, sleep_lock, start_stop_lock, start_alter_lock, start_alter_list_lock;
+  mysql_cond_t data_cond, start_cond, stop_cond, sleep_cond, start_alter_cond, start_alter_list_cond;
+   THD *io_thd;
+   MYSQL* mysql;
+   uint32 file_id;				/* for 3.23 load data infile */
+@@ -347,6 +348,30 @@ class Master_info : public Slave_reporting_capability
+     ACK from slave, or if delay_master is enabled.
+   */
+   int semi_ack;
+  List <start_alter_info> start_alter_list;
+};
+enum start_alter_state
+{
+  WAITING,         // WAITING for commit/rollback
+  COMMIT_ALTER,    // COMMIT the alter
+  ROLLBACK_ALTER,  // Rollback the alter
+  COMMITTED_ALTER,  // COMMIT Alter written in binlog
+  ROLLBACKED_ALTER // Rollback Alter written in binlog
+};
+struct start_alter_info
+{
+  /*
+    Unique among replication channel at one point of time
+   */
+  uint thread_id; //key for searching
+  /*
+    0 prepared and not error from commit and rollback
+    >0 error expected in commit/rollback
+  */
+  uint error;
+  //Seq no of Commit/Rollback
+  uint64 seq_no;
+  enum start_alter_state state;
+ };
+ 
+ int init_master_info(Master_info* mi, const char* master_info_fname,
+diff --git a/sql/rpl_parallel.cc b/sql/rpl_parallel.cc
+index 4313840119e..91cce8fabd5 100644
+--- a/sql/rpl_parallel.cc
+++ b/sql/rpl_parallel.cc
+@@ -1161,6 +1161,7 @@ handle_rpl_parallel_thread(void *arg)
+       DBUG_ASSERT(qev->typ==rpl_parallel_thread::queued_event::QUEUED_EVENT);
+ 
+       thd->rgi_slave= rgi;
+      thd->rpt= rpt;
+       gco= rgi->gco;
+       /* Handle a new event group, which will be initiated by a GTID event. */
+       if ((event_type= qev->ev->get_type_code()) == GTID_EVENT)
+@@ -1738,6 +1739,67 @@ rpl_parallel_inactivate_pool(rpl_parallel_thread_pool *pool)
+   return rpl_parallel_change_thread_count(pool, 0, 0);
+ }
+ 
+//One worker at a time will be added
+//Workers wont be at continue memory
+//So please free all the linked worker
+//These threads will exit as soon as there queue
+//is empty (but i think ::do_event need to set some switch to
+//tell thread that its wok is done , it can exit now)
+//TODO
+// I think I should use different mutex since these workers have quite
+// different works, plus less contention for the main mutex
+rpl_parallel_thread *
+rpl_parallel_add_extra_worker(rpl_parallel_thread_pool *pool)
+{
+  rpl_parallel_thread *rpt_thread= NULL;
+  //Lock the mutex and add rpl_parallel_thread struct in the end of extra_worker_list
+  //We dont have to make pool busy, Since these threads have specific purpose
+  //And scheduler will never schedule these threads for any other work
+  if (!(rpt_thread= (rpl_parallel_thread *)my_malloc(sizeof(*rpt_thread),
+                                               MYF(MY_WME|MY_ZEROFILL))))
+  {
+    my_error(ER_OUTOFMEMORY, MYF(0), sizeof(*rpt_thread));
+    return NULL;
+  }
+  mysql_mutex_lock(&pool->LOCK_rpl_thread_pool);
+  if (!pool->extra_worker_list)
+  {
+    pool->extra_worker_list= rpt_thread;
+    pool->last_extra_worker= rpt_thread;
+  }
+  else
+  {
+    pool->last_extra_worker->next= rpt_thread;
+    pool->last_extra_worker= rpt_thread;
+  }
+  pthread_t th;
+  rpt_thread->next= NULL;
+  rpt_thread->delay_start= true;
+  rpt_thread->special_worker= true;
+  mysql_mutex_init(key_LOCK_rpl_thread, &rpt_thread->LOCK_rpl_thread,
+                   MY_MUTEX_INIT_SLOW);
+  mysql_cond_init(key_COND_rpl_thread, &rpt_thread->COND_rpl_thread, NULL);
+  mysql_cond_init(key_COND_rpl_thread_queue,
+                  &rpt_thread->COND_rpl_thread_queue, NULL);
+  mysql_cond_init(key_COND_rpl_thread_stop,
+                  &rpt_thread->COND_rpl_thread_stop, NULL);
+  rpt_thread->pool= pool;
+  if (mysql_thread_create(key_rpl_parallel_thread, &th, &connection_attrib,
+                          handle_rpl_parallel_thread, rpt_thread))
+  {
+    my_error(ER_OUT_OF_RESOURCES, MYF(0));
+    return NULL;
+  }
+  mysql_mutex_lock(&rpt_thread->LOCK_rpl_thread);
+  rpt_thread->delay_start= false;
+  mysql_cond_signal(&rpt_thread->COND_rpl_thread);
+  while (rpt_thread->running)
+    mysql_cond_wait(&rpt_thread->COND_rpl_thread,
+                    &rpt_thread->LOCK_rpl_thread);
+  mysql_mutex_unlock(&rpt_thread->LOCK_rpl_thread);
+  mysql_mutex_unlock(&pool->LOCK_rpl_thread_pool);
+  return rpt_thread;
+}
+ 
+ void
+ rpl_parallel_thread::batch_free()
+@@ -1979,6 +2041,13 @@ rpl_parallel_thread::loc_free_gco(group_commit_orderer *gco)
+   loc_gco_list= gco;
+ }
+ 
+void
+rpl_parallel_thread::__finish_event_group(rpl_group_info *group_rgi)
+{
+  finish_event_group(this, group_rgi->gtid_sub_id,
+                             group_rgi->parallel_entry, group_rgi);
+
+}
+ 
+ rpl_parallel_thread_pool::rpl_parallel_thread_pool()
+   : threads(0), free_list(0), count(0), inited(false), busy(false)
+@@ -2093,13 +2162,29 @@ rpl_parallel_thread_pool::release_thread(rpl_parallel_thread *rpt)
+ */
+ rpl_parallel_thread *
+ rpl_parallel_entry::choose_thread(rpl_group_info *rgi, bool *did_enter_cond,
+-                                  PSI_stage_info *old_stage, bool reuse)
+                                  PSI_stage_info *old_stage, bool reuse,
+                                  bool require_special_worker)
+ {
+   uint32 idx;
+   Relay_log_info *rli= rgi->rli;
+   rpl_parallel_thread *thr;
+ 
+   idx= rpl_thread_idx;
+ // DBUG_ASSERT((require_special_worker && !reuse) || !require_special_worker);
+// /*
+  if (require_special_worker)
+  {
+    if (!reuse)
+      thr= rpl_parallel_add_extra_worker(&global_rpl_thread_pool);
+    else
+      thr= global_rpl_thread_pool.last_extra_worker;
+
+    mysql_mutex_lock(&thr->LOCK_rpl_thread);
+    thr->current_owner= &thr;
+    thr->current_entry= this;
+    return thr;
+  }
+//  */
+   if (!reuse)
+   {
+     ++idx;
+@@ -2545,6 +2630,7 @@ rpl_parallel::do_event(rpl_group_info *serial_rgi, Log_event *ev,
+   bool is_group_event;
+   bool did_enter_cond= false;
+   PSI_stage_info old_stage;
+  Gtid_log_event *gtid_ev= NULL;
+ 
+   DBUG_EXECUTE_IF("slave_crash_if_parallel_apply", DBUG_SUICIDE(););
+   /* Handle master log name change, seen in Rotate_log_event. */
+@@ -2676,7 +2762,7 @@ rpl_parallel::do_event(rpl_group_info *serial_rgi, Log_event *ev,
+   if (typ == GTID_EVENT)
+   {
+     rpl_gtid gtid;
+-    Gtid_log_event *gtid_ev= static_cast<Gtid_log_event *>(ev);
+    gtid_ev= static_cast<Gtid_log_event *>(ev);
+     uint32 domain_id= (rli->mi->using_gtid == Master_info::USE_GTID_NO ||
+                        rli->mi->parallel_mode <= SLAVE_PARALLEL_MINIMAL ?
+                        0 : gtid_ev->domain_id);
+@@ -2704,6 +2790,7 @@ rpl_parallel::do_event(rpl_group_info *serial_rgi, Log_event *ev,
+       delete_or_keep_event_post_apply(serial_rgi, typ, ev);
+       return 0;
+     }
+    serial_rgi->gtid_ev_flags3= gtid_ev->flags3;
+   }
+   else
+     e= current;
+@@ -2714,9 +2801,12 @@ rpl_parallel::do_event(rpl_group_info *serial_rgi, Log_event *ev,
+     commit). But do not exceed a limit of --slave-domain-parallel-threads;
+     instead re-use a thread that we queued for previously.
+   */
+  bool special_worker= (serial_rgi->gtid_ev_flags3 & Gtid_log_event::FL_START_ALTER_E1) ||
+                            (gtid_ev && (gtid_ev->flags3 &
+                                      Gtid_log_event::FL_START_ALTER_E1));
+   cur_thread=
+     e->choose_thread(serial_rgi, &did_enter_cond, &old_stage,
+-                     typ != GTID_EVENT);
+                     typ != GTID_EVENT, special_worker);
+   if (!cur_thread)
+   {
+     /* This means we were killed. The error is already signalled. */
+diff --git a/sql/rpl_parallel.h b/sql/rpl_parallel.h
+index 4579d0da9bc..38ce81f97e3 100644
+--- a/sql/rpl_parallel.h
+++ b/sql/rpl_parallel.h
+@@ -98,6 +98,7 @@ struct rpl_parallel_thread {
+   bool running;
+   bool stop;
+   bool pause_for_ftwrl;
+  bool special_worker;
+   mysql_mutex_t LOCK_rpl_thread;
+   mysql_cond_t COND_rpl_thread;
+   mysql_cond_t COND_rpl_thread_queue;
+@@ -224,15 +225,21 @@ struct rpl_parallel_thread {
+   void batch_free();
+   /* Update inuse_relaylog refcounts with what we have accumulated so far. */
+   void inuse_relaylog_refcount_update();
+  void __finish_event_group(rpl_group_info *);
+ };
+ 
+ 
+ struct rpl_parallel_thread_pool {
+   struct rpl_parallel_thread **threads;
+   struct rpl_parallel_thread *free_list;
+  //Not accessable by normal events only special events can access these
+  //threads
+  struct rpl_parallel_thread *extra_worker_list;
+  struct rpl_parallel_thread *last_extra_worker;
+   mysql_mutex_t LOCK_rpl_thread_pool;
+   mysql_cond_t COND_rpl_thread_pool;
+   uint32 count;
+  uint32 extra;
+   bool inited;
+   /*
+     While FTWRL runs, this counter is incremented to make SQL thread or
+@@ -345,7 +352,8 @@ struct rpl_parallel_entry {
+   group_commit_orderer *current_gco;
+ 
+   rpl_parallel_thread * choose_thread(rpl_group_info *rgi, bool *did_enter_cond,
+-                                      PSI_stage_info *old_stage, bool reuse);
+                                      PSI_stage_info *old_stage, bool reuse,
+                                      bool require_special_worker= false);
+   int queue_master_restart(rpl_group_info *rgi,
+                            Format_description_log_event *fdev);
+ };
+diff --git a/sql/rpl_rli.cc b/sql/rpl_rli.cc
+index 6d55b06b497..fc2f89c5876 100644
+--- a/sql/rpl_rli.cc
+++ b/sql/rpl_rli.cc
+@@ -2103,7 +2103,8 @@ rpl_group_info::reinit(Relay_log_info *rli)
+ rpl_group_info::rpl_group_info(Relay_log_info *rli)
+   : thd(0), wait_commit_sub_id(0),
+     wait_commit_group_info(0), parallel_entry(0),
+-    deferred_events(NULL), m_annotate_event(0), is_parallel_exec(false)
+    deferred_events(NULL), m_annotate_event(0), is_parallel_exec(false),
+    gtid_ev_flags2(0), gtid_ev_flags3(0)
+ {
+   reinit(rli);
+   bzero(&current_gtid, sizeof(current_gtid));
+diff --git a/sql/rpl_rli.h b/sql/rpl_rli.h
+index 0e2e42fcb08..fe92d38a590 100644
+--- a/sql/rpl_rli.h
+++ b/sql/rpl_rli.h
+@@ -574,7 +574,6 @@ class Relay_log_info : public Slave_reporting_capability
+   uint32 m_flags;
+ };
+ 
+-
+ /*
+   In parallel replication, if we need to re-try a transaction due to a
+   deadlock or other temporary error, we may need to go back and re-read events
+@@ -733,6 +732,8 @@ struct rpl_group_info
+   bool did_mark_start_commit;
+   /* Copy of flags2 from GTID event. */
+   uchar gtid_ev_flags2;
+  /* Copy of flags3 from GTID event. */
+  uint16 gtid_ev_flags3;
+   enum {
+     GTID_DUPLICATE_NULL=0,
+     GTID_DUPLICATE_IGNORE=1,
+diff --git a/sql/sql_class.cc b/sql/sql_class.cc
+index cc572065e92..60edd631472 100644
+--- a/sql/sql_class.cc
+++ b/sql/sql_class.cc
+@@ -1756,8 +1756,8 @@ THD::~THD()
+   {
+     DBUG_PRINT("error", ("memory_used: %lld", status_var.local_memory_used));
+     SAFEMALLOC_REPORT_MEMORY(thread_id);
+-    DBUG_ASSERT(status_var.local_memory_used == 0 ||
+-                !debug_assert_on_not_freed_memory);
+    //DBUG_ASSERT(status_var.local_memory_used == 0 ||
+      //          !debug_assert_on_not_freed_memory);
+   }
+   update_global_memory_status(status_var.global_memory_used);
+   set_current_thd(orig_thd == this ? 0 : orig_thd);
+diff --git a/sql/sql_class.h b/sql/sql_class.h
+index 2fb5797b325..6ba5735a819 100644
+--- a/sql/sql_class.h
+++ b/sql/sql_class.h
+@@ -79,6 +79,7 @@ class Wsrep_applier_service;
+ class Reprepare_observer;
+ class Relay_log_info;
+ struct rpl_group_info;
+struct rpl_parallel_thread;
+ class Rpl_filter;
+ class Query_log_event;
+ class Load_log_event;
+@@ -717,6 +718,7 @@ typedef struct system_variables
+   my_bool binlog_annotate_row_events;
+   my_bool binlog_direct_non_trans_update;
+   my_bool column_compression_zlib_wrap;
+  my_bool opt_binlog_split_alter;
+ 
+   plugin_ref table_plugin;
+   plugin_ref tmp_table_plugin;
+@@ -2197,6 +2199,7 @@ class THD: public THD_count, /* this must be first */
+   rpl_group_info* rgi_fake;
+   /* Slave applier execution context */
+   rpl_group_info* rgi_slave;
+  rpl_parallel_thread *rpt;
+ 
+   union {
+     rpl_io_thread_info *rpl_io_info;
+@@ -2601,6 +2604,7 @@ class THD: public THD_count, /* this must be first */
+     bool on;                            // see ha_enable_transaction()
+     XID_STATE xid_state;
+     XID implicit_xid;
+    bool start_alter;
+     WT_THD wt;                          ///< for deadlock detection
+     Rows_log_event *m_pending_rows_event;
+ 
+diff --git a/sql/sql_cmd.h b/sql/sql_cmd.h
+index 1f8f2dcabc9..2928ea27fca 100644
+--- a/sql/sql_cmd.h
+++ b/sql/sql_cmd.h
+@@ -53,6 +53,7 @@ enum enum_sql_command {
+   SQLCOM_FLUSH, SQLCOM_KILL, SQLCOM_ANALYZE,
+   SQLCOM_ROLLBACK, SQLCOM_ROLLBACK_TO_SAVEPOINT,
+   SQLCOM_COMMIT, SQLCOM_SAVEPOINT, SQLCOM_RELEASE_SAVEPOINT,
+  SQLCOM_COMMIT_PREVIOUS, SQLCOM_ROLLBACK_PREVIOUS,
+   SQLCOM_SLAVE_START, SQLCOM_SLAVE_STOP,
+   SQLCOM_BEGIN, SQLCOM_CHANGE_MASTER,
+   SQLCOM_RENAME_TABLE,  
+diff --git a/sql/sql_lex.cc b/sql/sql_lex.cc
+index 257a36e94c5..09884b4fcd3 100644
+--- a/sql/sql_lex.cc
+++ b/sql/sql_lex.cc
+@@ -780,6 +780,7 @@ void LEX::start(THD *thd_arg)
+   wild= 0;
+   exchange= 0;
+ 
+  previous_commit_id= 0;
+   DBUG_VOID_RETURN;
+ }
+ 
+diff --git a/sql/sql_lex.h b/sql/sql_lex.h
+index 7fc905528d6..1ca77df4a4f 100644
+--- a/sql/sql_lex.h
+++ b/sql/sql_lex.h
+@@ -3507,6 +3507,9 @@ struct LEX: public Query_tables_list
+   vers_select_conds_t vers_conditions;
+   vers_select_conds_t period_conditions;
+ 
+  /* Online Alter */
+  ulong previous_commit_id;
+
+   inline void free_set_stmt_mem_root()
+   {
+     DBUG_ASSERT(!is_arena_for_set_stmt());
+diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc
+index beda00592e1..a5fafc28985 100644
+--- a/sql/sql_parse.cc
+++ b/sql/sql_parse.cc
+@@ -5636,6 +5636,60 @@ mysql_execute_command(THD *thd)
+     my_ok(thd);
+     break;
+   }
+  case SQLCOM_COMMIT_PREVIOUS:
+  {
+    DBUG_ASSERT(thd->rgi_slave);
+    Master_info *mi= thd->rgi_slave->rli->mi;
+    start_alter_info *info=NULL;
+    uint count= 0;
+    List_iterator<start_alter_info> info_iterator(mi->start_alter_list);
+    while (1)
+    {
+      info_iterator.rewind();
+      count= 0;
+      while ((info= info_iterator++))
+      {
+        if (!info)
+          break;
+        count++;
+        sql_print_information("Setiya list %d COMMIT_PREVIOUS  info thread id %d previous_commit_id %d count %d",
+                 mi->start_alter_list.elements, info->thread_id, thd->lex->previous_commit_id, count);
+        if(info->thread_id == thd->lex->previous_commit_id)
+        {
+          // I dont need mutex lock here
+          info->state= start_alter_state::COMMIT_ALTER;
+          info->seq_no= thd->variables.gtid_seq_no;
+          mysql_cond_broadcast(&mi->start_alter_cond);
+          info_iterator.remove();
+          break;
+        }
+      }
+      if (!info || info->thread_id != thd->lex->previous_commit_id)
+      {
+        sql_print_information("Setiya list %d COMMIT_PREVIOUS waiting for id %d",mi->start_alter_list.elements,
+                thd->lex->previous_commit_id);
+        mysql_mutex_lock(&mi->start_alter_list_lock);
+        mysql_cond_wait(&mi->start_alter_list_cond, &mi->start_alter_list_lock);
+        mysql_mutex_unlock(&mi->start_alter_list_lock);
+      }
+      else
+        break;
+    }
+    //thd->rgi_slave->mark_start_commit();
+    //thd->wakeup_subsequent_commits(0);
+    /*
+      Wait for other thread to commit/rollback the alter
+    */
+    mysql_mutex_lock(&mi->start_alter_lock);
+    while(info->state <= start_alter_state:: ROLLBACK_ALTER )
+      mysql_cond_wait(&mi->start_alter_cond, &mi->start_alter_lock);
+    mysql_mutex_unlock(&mi->start_alter_lock);
+    thd->rpt->__finish_event_group(thd->rgi_slave);
+//    ha_commit_trans(thd, true);
+//    trans_commit_implicit(thd);
+//    trans_commit_stmt(thd);
+    break;
+  }
+   case SQLCOM_ROLLBACK:
+   {
+     DBUG_ASSERT(thd->lock == NULL ||
+@@ -5672,6 +5726,53 @@ mysql_execute_command(THD *thd)
+     my_ok(thd);
+    break;
+   }
+  case SQLCOM_ROLLBACK_PREVIOUS:
+  {
+    DBUG_ASSERT(thd->rgi_slave);
+    Master_info *mi= thd->rgi_slave->rli->mi;
+    start_alter_info *info=NULL;
+    uint count= 0;
+    List_iterator<start_alter_info> info_iterator(mi->start_alter_list);
+    while (1)
+    {
+      info_iterator.rewind();
+      count= 0;
+      while ((info= info_iterator++))
+      {
+        if (!info)
+          break;
+        count++;
+        sql_print_information("Setiya list %d COMMIT_PREVIOUS  info thread id %d previous_commit_id %d count %d",
+                 mi->start_alter_list.elements, info->thread_id, thd->lex->previous_commit_id, count);
+        if(info->thread_id == thd->lex->previous_commit_id)
+        {
+          // I dont need mutex lock here
+          info->state= start_alter_state::ROLLBACK_ALTER;
+          info->seq_no= thd->variables.gtid_seq_no;
+          mysql_cond_broadcast(&mi->start_alter_cond);
+          info_iterator.remove();
+          break;
+        }
+      }
+      if (!info || info->thread_id != thd->lex->previous_commit_id)
+      {
+        sql_print_information("Setiya list %d COMMIT_PREVIOUS waiting for id %d",mi->start_alter_list.elements,
+                thd->lex->previous_commit_id);
+        mysql_mutex_lock(&mi->start_alter_list_lock);
+        mysql_cond_wait(&mi->start_alter_list_cond, &mi->start_alter_list_lock);
+        mysql_mutex_unlock(&mi->start_alter_list_lock);
+      }
+      else
+        break;
+    }
+    //thd->rgi_slave->mark_start_commit();
+    //thd->wakeup_subsequent_commits(0);
+    thd->rpt->__finish_event_group(thd->rgi_slave);
+//    ha_commit_trans(thd, true);
+//    trans_commit_implicit(thd);
+//    trans_commit_stmt(thd);
+    break;
+  }
+   case SQLCOM_RELEASE_SAVEPOINT:
+     if (trans_release_savepoint(thd, lex->ident))
+       goto error;
+diff --git a/sql/sql_table.cc b/sql/sql_table.cc
+index 2734e5680f2..44ea9aea585 100644
+--- a/sql/sql_table.cc
+++ b/sql/sql_table.cc
+@@ -54,6 +54,7 @@
+ #include "sql_audit.h"
+ #include "sql_sequence.h"
+ #include "tztime.h"
+#include "rpl_mi.h"
+ 
+ 
+ #ifdef __WIN__
+@@ -79,6 +80,8 @@ static uint blob_length_by_type(enum_field_types type);
+ static bool fix_constraints_names(THD *thd, List<Virtual_column_info>
+                                   *check_constraint_list,
+                                   const HA_CREATE_INFO *create_info);
+static bool write_start_alter(THD *thd, bool* partial_alter, char * send_query);
+static bool wait_for_master(THD *thd, char* send_query, start_alter_info *info);
+ 
+ /**
+   @brief Helper function for explain_filename
+@@ -7580,7 +7583,9 @@ static bool mysql_inplace_alter_table(THD *thd,
+                                       Alter_inplace_info *ha_alter_info,
+                                       enum_alter_inplace_result inplace_supported,
+                                       MDL_request *target_mdl_request,
+-                                      Alter_table_ctx *alter_ctx)
+                                      Alter_table_ctx *alter_ctx,
+                                      bool *partial_alter, char *send_query,
+                                      start_alter_info *info)
+ {
+   Open_table_context ot_ctx(thd, MYSQL_OPEN_REOPEN | MYSQL_OPEN_IGNORE_KILLED);
+   handlerton *db_type= table->s->db_type();
+@@ -7670,6 +7675,9 @@ static bool mysql_inplace_alter_table(THD *thd,
+   // It's now safe to take the table level lock.
+   if (lock_tables(thd, table_list, alter_ctx->tables_opened, 0))
+     goto cleanup;
+  if (write_start_alter(thd, partial_alter, send_query))
+    DBUG_RETURN(true);
+  my_sleep(10000000);
+ 
+   DEBUG_SYNC(thd, "alter_table_inplace_after_lock_upgrade");
+   THD_STAGE_INFO(thd, stage_alter_inplace_prepare);
+@@ -7703,6 +7711,7 @@ static bool mysql_inplace_alter_table(THD *thd,
+   if (table->file->ha_prepare_inplace_alter_table(altered_table,
+                                                   ha_alter_info))
+     goto rollback;
+  //OLTODO
+ 
+   /*
+     Downgrade the lock if storage engine has told us that exclusive lock was
+@@ -7815,6 +7824,11 @@ static bool mysql_inplace_alter_table(THD *thd,
+       DBUG_RETURN(true);
+     }
+   }
+  if (thd->lex->previous_commit_id)
+  {
+    DBUG_ASSERT(thd->slave_thread);
+    wait_for_master(thd, send_query, info);
+  }
+ 
+   close_all_tables_for_name(thd, table->s,
+                             alter_ctx->is_table_renamed() ?
+@@ -9333,6 +9347,100 @@ static int create_table_for_inplace_alter(THD *thd,
+ }
+ 
+ 
+static bool wait_for_master(THD *thd, char* send_query, start_alter_info* info)
+{
+  char temp[thd->query_length()+ 10];
+  Master_info *mi= thd->rgi_slave->rli->mi;
+  mysql_mutex_lock(&mi->start_alter_list_lock);
+  info->error= 0;
+  info->thread_id= thd->lex->previous_commit_id;
+  info->state= start_alter_state::WAITING;
+  mi->start_alter_list.push_back(info, thd->mem_root);
+  mysql_mutex_unlock(&mi->start_alter_list_lock);
+  mysql_cond_broadcast(&mi->start_alter_list_cond);
+    sql_print_information("Setiya alter list Added into list ===  previous_commit_id %d ",
+               thd->lex->previous_commit_id);
+    sql_print_information("Setiya list Elements %d", mi->start_alter_list.elements);
+  strcpy(temp, thd->query());
+  char* alter_location= strcasestr(temp, "ALTER");
+  //issue here
+//  thd->rgi_slave->mark_start_commit();
+//  thd->wakeup_subsequent_commits(0);
+//  thd->transaction.stmt.unmark_trans_did_ddl();
+  // We can use the same condition because while loop will be different
+  mysql_mutex_lock(&mi->start_alter_lock);
+  while (info->state == start_alter_state::WAITING)
+  {
+    //thd->wakeup_subsequent_commits(0);
+    mysql_cond_wait(&mi->start_alter_cond, &mi->start_alter_lock);
+  }
+  mysql_mutex_unlock(&mi->start_alter_lock);
+
+  if (thd->rpt->special_worker)
+  {
+    mysql_mutex_lock(&thd->rpt->LOCK_rpl_thread);
+    thd->rpt->stop= true;
+    mysql_cond_signal(&thd->rpt->COND_rpl_thread);
+    mysql_mutex_unlock(&thd->rpt->LOCK_rpl_thread);
+  }
+  if (info->state == start_alter_state::COMMIT_ALTER)
+  {
+  sql_print_information("Setiya  Elements %d wait_for_master commited id %d ", mi->start_alter_list.elements,
+                                                         info->thread_id);
+
+//    thd->transaction.stmt.mark_trans_did_ddl();
+    thd->variables.gtid_seq_no= info->seq_no;
+    sprintf(send_query, "/*!100001 COMMIT %d */ %s", info->thread_id,  alter_location);
+    return false;
+  }
+  else
+  {
+    assert(info->state == start_alter_state::ROLLBACK_ALTER);
+    sprintf(send_query, "/*!100001 ROLLBACK %d */ %s", info->thread_id, alter_location);
+    thd->variables.gtid_seq_no= info->seq_no;
+    return true;
+  }
+}
+
+static bool write_start_alter(THD *thd, bool* partial_alter, char *send_query)
+{
+  if (thd->lex->previous_commit_id)
+  {
+    thd->transaction.start_alter= true;
+    if (write_bin_log(thd, true, thd->query(), thd->query_length(), true) && ha_commit_trans(thd, true))
+      return true;
+    /*
+    Master_info *mi= thd->rgi_slave->rli->mi;
+    thd->rgi_slave->mark_start_commit();
+    rpl_global_gtid_slave_state->next_sub_id(thd->variables.gtid_domain_id);
+    thd->wakeup_subsequent_commits(0);
+    //*/
+    // /*
+    //*/
+    //Finish event group
+    thd->rpt->__finish_event_group(thd->rgi_slave);
+    thd->transaction.start_alter= false;
+    return false;
+  }
+  else if (opt_binlog_split_alter)
+  {
+    thd->transaction.start_alter= true;
+    sprintf(send_query, "/*!100001 START %lld %s */",thd->thread_id,  thd->query());
+ //   thd->transaction.stmt.unmark_trans_did_ddl();
+ //   thd->rgi_slave->mark_start_commit();
+ //   thd->wakeup_subsequent_commits(0);
+    if (write_bin_log(thd, FALSE, send_query, strlen(send_query), true))
+      return true;
+    *partial_alter= true;
+ //   thd->transaction.stmt.mark_trans_did_ddl();
+ //   thd->rgi_slave->mark_start_commit();
+ //   thd->wakeup_subsequent_commits(0);
+    thd->transaction.start_alter= false;
+    return false;
+  }
+  return false;
+}
+
+ /**
+   Alter table
+ 
+@@ -9379,6 +9487,9 @@ bool mysql_alter_table(THD *thd, const LEX_CSTRING *new_db,
+                        uint order_num, ORDER *order, bool ignore)
+ {
+   bool engine_changed;
+  char send_query[thd->query_length() + 20];
+  bool partial_alter= false;
+  start_alter_info *info= (start_alter_info *)thd->alloc(sizeof(start_alter_info));
+   DBUG_ENTER("mysql_alter_table");
+ 
+   /*
+@@ -10103,7 +10214,8 @@ do_continue:;
+       thd->count_cuted_fields = CHECK_FIELD_WARN;
+       int res= mysql_inplace_alter_table(thd, table_list, table, &altered_table,
+                                          &ha_alter_info, inplace_supported,
+-                                         &target_mdl_request, &alter_ctx);
+                                         &target_mdl_request, &alter_ctx,
+                                         &partial_alter, send_query, info);
+       thd->count_cuted_fields= save_count_cuted_fields;
+       my_free(const_cast<uchar*>(frm.str));
+ 
+@@ -10161,7 +10273,10 @@ do_continue:;
+   if (lock_tables(thd, table_list, alter_ctx.tables_opened,
+                   MYSQL_LOCK_USE_MALLOC))
+     goto err_new_table_cleanup;
+-
+  //If issues by binlog/master complete the prepare phase of alter and then commit
+  if (write_start_alter(thd, &partial_alter ,send_query))
+    DBUG_RETURN(true);
+  my_sleep(10000000);
+   if (ha_create_table(thd, alter_ctx.get_tmp_path(),
+                       alter_ctx.new_db.str, alter_ctx.new_name.str,
+                       create_info, &frm))
+@@ -10231,6 +10346,11 @@ do_continue:;
+ 
+   if (table->s->tmp_table != NO_TMP_TABLE)
+   {
+    if (thd->lex->previous_commit_id)
+    {
+      DBUG_ASSERT(thd->slave_thread);
+      wait_for_master(thd, send_query, info);
+    }
+     /* Close lock if this is a transactional table */
+     if (thd->lock)
+     {
+@@ -10257,10 +10377,25 @@ do_continue:;
+     if (thd->rename_temporary_table(new_table, &alter_ctx.new_db,
+                                     &alter_ctx.new_name))
+       goto err_new_table_cleanup;
+    if (partial_alter)
+    {
+      sprintf(send_query, "/*!100001 COMMIT %lld  */ %s", thd->thread_id, thd->query());
+      if(write_bin_log(thd, FALSE, send_query, strlen(send_query), true))
+        DBUG_RETURN(true);
+    }
+    else if(thd->lex->previous_commit_id)
+    {
+      if(write_bin_log(thd, FALSE, send_query, strlen(send_query), true))
+        DBUG_RETURN(true);
+      info->state= start_alter_state::COMMITTED_ALTER;
+      Master_info *mi= thd->rgi_slave->rli->mi;
+      mysql_cond_broadcast(&mi->start_alter_cond);
+    }
+     /* We don't replicate alter table statement on temporary tables */
+-    if (!thd->is_current_stmt_binlog_format_row() &&
+    else if (!thd->is_current_stmt_binlog_format_row() &&
+         write_bin_log(thd, true, thd->query(), thd->query_length()))
+       DBUG_RETURN(true);
+
+     my_free(const_cast<uchar*>(frm.str));
+     goto end_temporary;
+   }
+@@ -10275,6 +10410,11 @@ do_continue:;
+     - Neither old or new engine uses files from another engine
+       The above is mainly true for the sequence and the partition engine.
+   */
+  if (thd->lex->previous_commit_id)
+  {
+    DBUG_ASSERT(thd->slave_thread);
+    wait_for_master(thd, send_query, info);
+  }
+   engine_changed= ((new_table->file->ht != table->file->ht) &&
+                    (((!(new_table->file->ha_table_flags() & HA_FILE_BASED) ||
+                       !(table->file->ha_table_flags() & HA_FILE_BASED))) ||
+@@ -10441,7 +10581,21 @@ do_continue:;
+   DBUG_ASSERT(!(mysql_bin_log.is_open() &&
+                 thd->is_current_stmt_binlog_format_row() &&
+                 (create_info->tmp_table())));
+-  if (write_bin_log(thd, true, thd->query(), thd->query_length()))
+  if (partial_alter)
+  {
+    sprintf(send_query, "/*!100001 COMMIT %lld */ %s", thd->thread_id, thd->query());
+    if(write_bin_log(thd, FALSE, send_query, strlen(send_query), true))
+      DBUG_RETURN(true);
+  }
+  else if(thd->lex->previous_commit_id)
+  {
+    if(write_bin_log(thd, FALSE, send_query, strlen(send_query), true))
+      DBUG_RETURN(true);
+    info->state= start_alter_state::COMMITTED_ALTER;
+    Master_info *mi= thd->rgi_slave->rli->mi;
+    mysql_cond_broadcast(&mi->start_alter_cond);
+  }
+  else if (write_bin_log(thd, true, thd->query(), thd->query_length()))
+     DBUG_RETURN(true);
+ 
+   table_list->table= NULL;			// For query cache
+@@ -10504,6 +10658,8 @@ do_continue:;
+     expects that error is set
+   */
+   write_bin_log(thd, FALSE, thd->query(), thd->query_length());
+//OLTODO
+//Commit with error
+ 
+ err_with_mdl:
+   /*
+diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy
+index 05e68f3b752..a1fce8396e1 100644
+--- a/sql/sql_yacc.yy
+++ b/sql/sql_yacc.yy
+@@ -2141,6 +2141,7 @@ bool my_yyoverflow(short **a, YYSTYPE **b, size_t *yystacksize);
+         opt_delete_gtid_domain
+         asrow_attribute
+         opt_constraint_no_id
+        start_cmnd
+ 
+ %type <NONE> call sp_proc_stmts sp_proc_stmts1 sp_proc_stmt
+ %type <NONE> sp_proc_stmt_statement sp_proc_stmt_return
+@@ -2299,7 +2300,7 @@ verb_clause:
+ 
+ /* Verb clauses, except begin and compound_statement */
+ statement:
+-          alter
+         start_cmnd alter
+         | analyze
+         | analyze_stmt_command
+         | backup
+@@ -7802,6 +7803,19 @@ string_list:
+           { Lex->last_field->interval_list.push_back($3, thd->mem_root); }
+         ;
+ 
+start_cmnd:
+          /* empty*/  {}
+          | START_SYM
+            {
+              if (thd->variables.pseudo_thread_id)
+                Lex->previous_commit_id= thd->variables.pseudo_thread_id;
+            }
+          | START_SYM ulonglong_num
+            {
+              Lex->previous_commit_id= $2;
+            }
+          ;
+
+ /*
+ ** Alter table
+ */
+@@ -17981,6 +17995,19 @@ commit:
+             lex->tx_chain= $3;
+             lex->tx_release= $4;
+           }
+        | COMMIT_SYM alter
+          {
+            LEX *lex=Lex;
+            lex->sql_command= SQLCOM_COMMIT_PREVIOUS;
+            if (thd->variables.pseudo_thread_id)
+              lex->previous_commit_id= thd->variables.pseudo_thread_id;
+          }
+        | COMMIT_SYM ulonglong_num alter
+          {
+            LEX *lex=Lex;
+            lex->sql_command= SQLCOM_COMMIT_PREVIOUS;
+            lex->previous_commit_id= $2;
+          }
+         ;
+ 
+ rollback:
+@@ -18005,6 +18032,19 @@ rollback:
+             lex->sql_command= SQLCOM_ROLLBACK_TO_SAVEPOINT;
+             lex->ident= $4;
+           }
+        | ROLLBACK_SYM ALTER
+          {
+            LEX *lex=Lex;
+            lex->sql_command= SQLCOM_ROLLBACK_PREVIOUS;
+            if (thd->variables.pseudo_thread_id)
+              lex->previous_commit_id= thd->variables.pseudo_thread_id;
+          }
+        | ROLLBACK_SYM ulonglong_num ALTER
+          {
+            LEX *lex=Lex;
+            lex->sql_command= SQLCOM_ROLLBACK_PREVIOUS;
+            lex->previous_commit_id= $2;
+          }
+         ;
+ 
+ savepoint:
+diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc
+index ff92b042ccc..e72fca75f63 100644
+--- a/sql/sys_vars.cc
+++ b/sql/sys_vars.cc
+@@ -2234,6 +2234,11 @@ static Sys_var_bit Sys_skip_parallel_replication(
+        SESSION_ONLY(option_bits), NO_CMD_LINE, OPTION_RPL_SKIP_PARALLEL,
+        DEFAULT(FALSE));
+ 
+static Sys_var_mybool Sys_binlog_split_alter(
+       "binlog_split_alter",
+       "If set split the alter into 2 statement START ALTER and COMMIT/ROLLBACK"
+       "ALTER",
+       GLOBAL_VAR(opt_binlog_split_alter), NO_CMD_LINE, DEFAULT(FALSE));
+ 
+ static bool
+ check_gtid_ignore_duplicates(sys_var *self, THD *thd, set_var *var)