mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-04-05 23:05:34 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1
mariadb/mysql-test/suite/rpl/t/rpl_semi_sync.test
Michael Widenius 7af50e4df4 MDEV-32551: "Read semi-sync reply magic number error" warnings on master 
rpl_semi_sync_slave_enabled_consistent.test and the first part of
the commit message comes from Brandon Nesterenko.

A test to show how to induce the "Read semi-sync reply magic number
error" message on a primary. In short, if semi-sync is turned on
during the hand-shake process between a primary and replica, but
later a user negates the rpl_semi_sync_slave_enabled variable while
the replica's IO thread is running; if the io thread exits, the
replica can skip a necessary call to kill_connection() in
repl_semisync_slave.slave_stop() due to its reliance on a global
variable. Then, the replica will send a COM_QUIT packet to the
primary on an active semi-sync connection, causing the magic number
error.

The test in this patch exits the IO thread by forcing an error;
though note a call to STOP SLAVE could also do this, but it ends up
needing more synchronization. That is, the STOP SLAVE command also
tries to kill the VIO of the replica, which makes a race with the IO
thread to try and send the COM_QUIT before this happens (which would
need more debug_sync to get around). See THD::awake_no_mutex for
details as to the killing of the replica’s vio.

Notes:
- The MariaDB documentation does not make it clear that when one
  enables semi-sync replication it does not matter if one enables
  it first in the master or slave. Any order works.

Changes done:
- The rpl_semi_sync_slave_enabled variable is now a default value for
  when semisync is started. The variable does not anymore affect
  semisync if it is already running. This fixes the original reported
  bug.  Internally we now use repl_semisync_slave.get_slave_enabled()
  instead of rpl_semi_sync_slave_enabled. To check if semisync is
  active on should check the @@rpl_semi_sync_slave_status variable (as
  before).
- The semisync protocol conflicts in the way that the original
  MySQL/MariaDB client-server protocol was designed (client-server
  send and reply packets are strictly ordered and includes a packet
  number to allow one to check if a packet is lost). When using
  semi-sync the master and slave can send packets at 'any time', so
  packet numbering does not work. The 'solution' has been that each
  communication starts with packet number 1, but in some cases there
  is still a chance that the packet number check can fail.  Fixed by
  adding a flag (pkt_nr_can_be_reset) in the NET struct that one can
  use to signal that packet number checking should not be done. This
  is flag is set when semi-sync is used.
- Added Master_info::semi_sync_reply_enabled to allow one to configure
  some slaves with semisync and other other slaves without semisync.
  Removed global variable semi_sync_need_reply that would not work
  with multi-master.
- Repl_semi_sync_master::report_reply_packet() can now recognize
  the COM_QUIT packet from semisync slave and not give a
  "Read semi-sync reply magic number error" error for this case.
  The slave will be removed from the Ack listener.
- On Windows, don't stop semisync Ack listener just because one
  slave connection is using socket_id > FD_SETSIZE.
- Removed busy loop in Ack_receiver::run() by using
 "Self-pipe trick" to signal new slave and stop Ack_receiver.
- Changed some Repl_semi_sync_slave functions that always returns 0
  from int to void.
- Added Repl_semi_sync_slave::slave_reconnect().
- Removed dummy_function Repl_semi_sync_slave::reset_slave().
- Removed some duplicate semisync notes from the error log.
- Add test of "if (get_slave_enabled() && semi_sync_need_reply)"
  before calling Repl_semi_sync_slave::slave_reply().
  (Speeds up the code as we can skip all initializations).
- If epl_semisync_slave.slave_reply() fails, we disable semisync
  for that connection.
- We do not call semisync.switch_off() if there are no active slaves.
  Instead we check in Repl_semi_sync_master::commit_trx() if there are
  no active threads. This simplices the code.
- Changed assert() to DBUG_ASSERT() to ensure that the DBUG log is
  flushed in case of asserts.
- Removed the internal rpl_semi_sync_slave_status as it is not needed
  anymore. The @@rpl_semi_sync_slave_status status variable is now
  mapped to rpl_semi_sync_enabled.
- Removed rpl_semi_sync_slave_enabled  as it is not needed anymore.
  Repl_semi_sync_slave::get_slave_enabled() contains the active status.
- Added checking that we do not add a slave twice with
  Ack_receiver::add_slave(). This could happen with old code.
- Removed Repl_semi_sync_master::check_and_switch() as it is not
  needed anymore.
- Ensure that when we call Ack_receiver::remove_slave() that the slave
  is removed from the listener before function returns.
- Call listener.listen_on_sockets() outside of mutex for better
  performance and less contested mutex.
- Ensure that listening is ignoring newly added slaves when checking for
  responses.
- Fixed the master ack_receiver listener is not killed if there are no
  connected slaves (and thus stop semisync handling of future
  connections). This could happen if all slaves sockets where would be
  marked as unreliable.
- Added unlink() to base_ilist_iterator and remove() to
  I_List_iterator. This enables us to remove 'dead' slaves in
  Ack_recever::run().
- kill_zombie_dump_threads() now does killing of dump threads properly.
  - It can now kill several threads (should be impossible but could
    happen if IO slaves reconnects very fast).
  - We now wait until the dump thread is done before starting the
    dump.
- Added an error if kill_zombie_dump_threads() fails.
- Set thd->variables.server_id before calling
  kill_zombie_dump_threads(). This simplies the code.
- Added a lot of comments both in code and tests.
- Removed DBUG_EVALUATE_IF "failed_slave_start" as it is not used.

Test changes:
- rpl.rpl_session_var2 added which runs rpl.rpl_session_var test with
  semisync enabled.
- Some timings changed slight with startup of slave which caused
  rpl_binlog_dump_slave_gtid_state_info.text to fail as it checked the
  error log file before the slave had started properly. Fixed by
  adding wait_for_pattern_in_file.inc that allows waiting for the
  pattern to appear in the log file.
- Tests have been updated so that we first set
  rpl_semi_sync_master_enabled on the master and then set
  rpl_semi_sync_slave_enabled on the slaves (this is according to how
  the MariaDB documentation document how to setup semi-sync).
- Error text "Master server does not have semi-sync enabled" has been
  replaced with "Master server does not support semi-sync" for the
  case when the master supports semi-sync but semi-sync is not
  enabled.

Other things:
- Some trivial cleanups in Repl_semi_sync_master::update_sync_header().
- We should in 11.3 changed the default value for
  rpl-semi-sync-master-wait-no-slave from TRUE to FALSE as the TRUE
  does not make much sense as default. The main difference with using
  FALSE is that we do not wait for semisync Ack if there are no slave
  threads.  In the case of TRUE we wait once, which did not bring any
  notable benefits except slower startup of master configured for
  using semisync.

Co-author: Brandon Nesterenko <brandon.nesterenko@mariadb.com>

This solves the problem reported in MDEV-32960 where a new
slave may not be registered in time and the master disables
semi sync because of that.
2024-01-23 13:03:11 +02:00

535 lines
16 KiB
Text
Raw Permalink Blame History

#
# This include file is used by more than one test suite
# (currently rpl and binlog_encryption).
# Please check all dependent tests after modifying it
#
source include/not_embedded.inc;
source include/have_innodb.inc;
source include/master-slave.inc;
let $engine_type= InnoDB;
# Suppress warnings that might be generated during the test
connection master;
call mtr.add_suppression("Timeout waiting for reply of binlog");
call mtr.add_suppression("Read semi-sync reply");
call mtr.add_suppression("Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT.");
call mtr.add_suppression("mysqld: Got an error reading communication packets");
connection slave;
call mtr.add_suppression("Semi-sync slave .* reply");
call mtr.add_suppression("Slave SQL.*Request to stop slave SQL Thread received while applying a group that has non-transactional changes; waiting for completion of the group");
connection master;
# wait for dying connections (if any) to disappear
let $wait_condition= select count(*) = 0 from information_schema.processlist where command='killed';
--source include/wait_condition.inc
# After fix of BUG#45848, semi-sync slave should not create any extra
# connections on master, save the count of connections before start
# semi-sync slave for comparison below.
let $_connections_normal_slave= query_get_value(SHOW STATUS LIKE 'Threads_connected', Value, 1);
--echo #
--echo # Uninstall semi-sync plugins on master and slave
--echo #
connection slave;
source include/stop_slave.inc;
reset slave;
set global rpl_semi_sync_master_enabled= 0;
set global rpl_semi_sync_slave_enabled= 0;
connection master;
reset master;
set global rpl_semi_sync_master_enabled= 0;
set global rpl_semi_sync_slave_enabled= 0;
--echo #
--echo # Main test of semi-sync replication start here
--echo #
connection master;
set global rpl_semi_sync_master_timeout= 2000; # 2s
echo [ default state of semi-sync on master should be OFF ];
show variables like 'rpl_semi_sync_master_enabled';
echo [ enable semi-sync on master ];
set global rpl_semi_sync_master_enabled = 1;
show variables like 'rpl_semi_sync_master_enabled';
echo [ status of semi-sync on master should be ON even without any semi-sync slaves ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_yes_tx';
--echo #
--echo # BUG#45672 Semisync repl: ActiveTranx:insert_tranx_node: transaction node allocation failed
--echo # BUG#45673 Semisynch reports correct operation even if no slave is connected
--echo #
# BUG#45672 When semi-sync is enabled on master, it would allocate
# transaction node even without semi-sync slave connected, and would
# finally result in transaction node allocation error.
#
# Semi-sync master will pre-allocate 'max_connections' transaction
# nodes, so here we do more than that much transactions to check if it
# will fail or not.
# select @@global.max_connections + 1;
let $i= `select @@global.max_connections + 1`;
disable_query_log;
eval create table t1 (a int) engine=$engine_type;
while ($i)
{
eval insert into t1 values ($i);
dec $i;
}
drop table t1;
enable_query_log;
# BUG#45673
echo [ status of semi-sync on master should be OFF ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_yes_tx';
# reset master to make sure the following test will start with a clean environment
reset master;
connection slave;
echo [ default state of semi-sync on slave should be OFF ];
show variables like 'rpl_semi_sync_slave_enabled';
echo [ enable semi-sync on slave ];
set global rpl_semi_sync_slave_enabled = 1;
show variables like 'rpl_semi_sync_slave_enabled';
source include/start_slave.inc;
connection master;
# NOTE: Rpl_semi_sync_master_client will only be updated when
# semi-sync slave has started binlog dump request
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 1;
source include/wait_for_status_var.inc;
echo [ initial master state after the semi-sync slave connected ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
replace_result $engine_type ENGINE_TYPE;
eval create table t1(a int) engine = $engine_type;
echo [ master state after CREATE TABLE statement ];
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
# After fix of BUG#45848, semi-sync slave should not create any extra
# connections on master.
let $_connections_semisync_slave= query_get_value(SHOW STATUS LIKE 'Threads_connected', Value, 1);
replace_result $_connections_normal_slave CONNECTIONS_NORMAL_SLAVE $_connections_semisync_slave CONNECTIONS_SEMISYNC_SLAVE;
eval select $_connections_semisync_slave - $_connections_normal_slave as 'Should be 0';
echo [ insert records to table ];
insert t1 values (10);
insert t1 values (9);
insert t1 values (8);
insert t1 values (7);
insert t1 values (6);
insert t1 values (5);
insert t1 values (4);
insert t1 values (3);
insert t1 values (2);
insert t1 values (1);
echo [ master status after inserts ];
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
sync_slave_with_master;
echo [ slave status after replicated inserts ];
show status like 'Rpl_semi_sync_slave_status';
select count(distinct a) from t1;
select min(a) from t1;
select max(a) from t1;
--echo
--echo # BUG#50157
--echo # semi-sync replication crashes when replicating a transaction which
--echo # include 'CREATE TEMPORARY TABLE `MyISAM_t` SELECT * FROM `Innodb_t` ;
connection master;
SET SESSION AUTOCOMMIT= 0;
CREATE TABLE t2(c1 INT) ENGINE=innodb;
sync_slave_with_master;
connection master;
BEGIN;
--echo
--echo # Even though it is in a transaction, this statement is binlogged into binlog
--echo # file immediately.
--disable_warnings
CREATE TEMPORARY TABLE t3 SELECT c1 FROM t2 where 1=1;
--enable_warnings
--echo
--echo # These statements will not be binlogged until the transaction is committed
INSERT INTO t2 VALUES(11);
INSERT INTO t2 VALUES(22);
COMMIT;
DROP TABLE t2, t3;
SET SESSION AUTOCOMMIT= 1;
sync_slave_with_master;
--echo #
--echo # Test semi-sync master will switch OFF after one transaction
--echo # timeout waiting for slave reply.
--echo #
connection master;
show status like "Rpl_semi_sync_master_status";
connection slave;
source include/stop_slave.inc;
connection master;
--source include/kill_binlog_dump_threads.inc
set global rpl_semi_sync_master_timeout= 2000;
# The first semi-sync check should be on because after slave stop,
# there are no transactions on the master.
echo [ master status should be ON ];
let $status_var= Rpl_semi_sync_master_status;
let $status_var_value= ON;
source include/wait_for_status_var.inc;
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 0;
source include/wait_for_status_var.inc;
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
echo [ semi-sync replication of these transactions will fail ];
insert into t1 values (500);
# Wait for the semi-sync replication of this transaction to timeout
let $status_var= Rpl_semi_sync_master_status;
let $status_var_value= OFF;
source include/wait_for_status_var.inc;
# The second semi-sync check should be off because one transaction
# times out during waiting.
echo [ master status should be OFF ];
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
# Semi-sync status on master is now ON, but there are no slaves attached,
# so all these transactions will be replicated asynchronously.
delete from t1 where a=10;
delete from t1 where a=9;
delete from t1 where a=8;
delete from t1 where a=7;
delete from t1 where a=6;
delete from t1 where a=5;
delete from t1 where a=4;
delete from t1 where a=3;
delete from t1 where a=2;
delete from t1 where a=1;
insert into t1 values (100);
echo [ master status should be OFF ];
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
--echo #
--echo # Test semi-sync status on master will be ON again when slave catches up
--echo #
# Save the master position for later use.
save_master_pos;
connection slave;
echo [ slave status should be OFF ];
show status like 'Rpl_semi_sync_slave_status';
source include/start_slave.inc;
sync_with_master;
echo [ slave status should be ON ];
show status like 'Rpl_semi_sync_slave_status';
select count(distinct a) from t1;
select min(a) from t1;
select max(a) from t1;
connection master;
# The master semi-sync status should be on again after slave catches up.
echo [ master status should be ON again after slave catches up ];
let $status_var= Rpl_semi_sync_master_status;
let $status_var_value= ON;
source include/wait_for_status_var.inc;
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
show status like 'Rpl_semi_sync_master_clients';
--echo #
--echo # Test disable/enable master semi-sync on the fly.
--echo #
drop table t1;
sync_slave_with_master;
source include/stop_slave.inc;
--echo #
--echo # Flush status
--echo #
connection master;
echo [ Semi-sync master status variables before FLUSH STATUS ];
SHOW STATUS LIKE 'Rpl_semi_sync_master_no_tx';
SHOW STATUS LIKE 'Rpl_semi_sync_master_yes_tx';
# Do not write the FLUSH STATUS to binlog, to make sure we'll get a
# clean status after this.
FLUSH NO_WRITE_TO_BINLOG STATUS;
echo [ Semi-sync master status variables after FLUSH STATUS ];
SHOW STATUS LIKE 'Rpl_semi_sync_master_no_tx';
SHOW STATUS LIKE 'Rpl_semi_sync_master_yes_tx';
connection master;
source include/show_master_logs.inc;
show variables like 'rpl_semi_sync_master_enabled';
echo [ disable semi-sync on the fly ];
set global rpl_semi_sync_master_enabled=0;
show variables like 'rpl_semi_sync_master_enabled';
show status like 'Rpl_semi_sync_master_status';
echo [ enable semi-sync on the fly ];
set global rpl_semi_sync_master_enabled=1;
show variables like 'rpl_semi_sync_master_enabled';
show status like 'Rpl_semi_sync_master_status';
--echo #
--echo # Test RESET MASTER/SLAVE
--echo #
connection slave;
source include/start_slave.inc;
connection master;
replace_result $engine_type ENGINE_TYPE;
eval create table t1 (a int) engine = $engine_type;
drop table t1;
sync_slave_with_master;
echo [ test reset master ];
connection master;
reset master;
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
connection slave;
source include/stop_slave.inc;
reset slave;
# Kill the dump thread on master for previous slave connection and
--source include/kill_binlog_dump_threads.inc
connection slave;
source include/start_slave.inc;
connection master;
# Wait for dump thread to start, Rpl_semi_sync_master_clients will be
# 1 after dump thread started.
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 1;
source include/wait_for_status_var.inc;
sync_slave_with_master;
connection master;
replace_result $engine_type ENGINE_TYPE;
eval create table t1 (a int) engine = $engine_type;
insert into t1 values (1);
insert into t1 values (2), (3);
sync_slave_with_master;
select * from t1;
connection master;
echo [ master semi-sync status should be ON ];
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
--echo #
--echo # Start semi-sync replication without SUPER privilege
--echo #
connection slave;
source include/stop_slave.inc;
reset slave;
connection master;
reset master;
# Kill the dump thread on master for previous slave connection and wait for it to exit
--source include/kill_binlog_dump_threads.inc
# Do not binlog the following statement because it will generate
# different events for ROW and STATEMENT format
set sql_log_bin=0;
grant replication slave on *.* to rpl@127.0.0.1 identified by 'rpl_password';
flush privileges;
set sql_log_bin=1;
connection slave;
grant replication slave on *.* to rpl@127.0.0.1 identified by 'rpl_password';
flush privileges;
change master to master_user='rpl',master_password='rpl_password';
source include/start_slave.inc;
show status like 'Rpl_semi_sync_slave_status';
connection master;
# Wait for the semi-sync binlog dump thread to start
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 1;
source include/wait_for_status_var.inc;
sync_slave_with_master;
connection master;
echo [ master semi-sync should be ON ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
insert into t1 values (4);
insert into t1 values (5);
echo [ master semi-sync should be ON ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
show status like 'Rpl_semi_sync_master_no_tx';
show status like 'Rpl_semi_sync_master_yes_tx';
--echo #
--echo # Test semi-sync slave connect to non-semi-sync master
--echo #
# Disable semi-sync on master
connection slave;
source include/stop_slave.inc;
SHOW STATUS LIKE 'Rpl_semi_sync_slave_status';
connection master;
# Kill the dump thread on master for previous slave connection and wait for it to exit
--source include/kill_binlog_dump_threads.inc
echo [ Semi-sync status on master should be ON ];
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 0;
source include/wait_for_status_var.inc;
show status like 'Rpl_semi_sync_master_status';
let $status_var= Rpl_semi_sync_master_status;
let $status_var_value= ON;
source include/wait_for_status_var.inc;
set global rpl_semi_sync_master_enabled= 0;
connection slave;
SHOW VARIABLES LIKE 'rpl_semi_sync_slave_enabled';
source include/start_slave.inc;
connection master;
insert into t1 values (8);
let $status_var= Rpl_semi_sync_master_clients;
let $status_var_value= 1;
source include/wait_for_status_var.inc;
echo [ master semi-sync clients should be 1, status should be OFF ];
show status like 'Rpl_semi_sync_master_clients';
show status like 'Rpl_semi_sync_master_status';
sync_slave_with_master;
show status like 'Rpl_semi_sync_slave_status';
# Uninstall semi-sync plugin on master
connection slave;
source include/stop_slave.inc;
connection master;
set global rpl_semi_sync_master_enabled= 0;
connection slave;
SHOW VARIABLES LIKE 'rpl_semi_sync_slave_enabled';
source include/start_slave.inc;
connection master;
insert into t1 values (10);
sync_slave_with_master;
--echo #
--echo # Test non-semi-sync slave connect to semi-sync master
--echo #
connection master;
set global rpl_semi_sync_master_timeout= 5000; # 5s
set global rpl_semi_sync_master_enabled= 1;
connection slave;
source include/stop_slave.inc;
SHOW STATUS LIKE 'Rpl_semi_sync_slave_status';
echo [ uninstall semi-sync slave plugin ];
set global rpl_semi_sync_slave_enabled= 0;
echo [ reinstall semi-sync slave plugin and disable semi-sync ];
SHOW VARIABLES LIKE 'rpl_semi_sync_slave_enabled';
SHOW STATUS LIKE 'Rpl_semi_sync_slave_status';
source include/start_slave.inc;
SHOW STATUS LIKE 'Rpl_semi_sync_slave_status';
--echo #
--echo # Clean up
--echo #
connection slave;
source include/stop_slave.inc;
set global rpl_semi_sync_slave_enabled= 0;
connection master;
set global rpl_semi_sync_master_enabled= 0;
connection slave;
change master to master_user='root',master_password='';
source include/start_slave.inc;
connection master;
drop table t1;
sync_slave_with_master;
connection master;
drop user rpl@127.0.0.1;
flush privileges;
set global rpl_semi_sync_master_timeout= default;
--source include/rpl_end.inc
Reference in a new issue View git blame Copy permalink