This is an extension of MDEV-30423 "Deadlock on Replica during BACKUP
STAGE BLOCK_COMMIT on XA transactions"
The original commit in MDEV-30423 was not complete as some usage in XA of
MDL_BACKUP_COMMIT locks did not set thd->backup_commit_lock.
This is required to be set when using parallel replication.
Fixed by ensuring that all usage of BACKUP_COMMIT lock i XA is uniform and
all sets thd->backup_commit_lock. I also changed all locks to be
MDL_EXPLICIT to keep also that part uniform.
A regression test is added.
If semi-sync is switched off then on while a transaction is
in-between binlogging and waiting for an ACK, the semi-sync state of
the transaction is removed, leading to a debug assertion that
indicates the transaction tried to wait, but cannot receive an ACK
signal. More specifically, when semi-sync is switched off, the
Active_tranx list is cleared (where a transaction adds an entry to
this list during binlogging), and each entry in this list saves the
thread which will wait for an ACK, and the thread has the COND
variable to signal to wake itself. So if the entry is lost, the
Ack_receiver thread won’t be able to find the thread to wake up when
an ACK comes in
The fix is to ensure that the entry exists before awaiting the ACK,
and if there is no entry, skip the wait. In debug builds, an
informative message is written explaining that the transaction is
skipping its wait. Additional debug-build only logic is added to
ensure that the cause of the missing entry is due to semi-sync being
turned off and on
Reviewed By:
============
Kristian Nielsen <knielsen@knielsen-hq.org>
Implement variable legacy_xa_rollback_at_disconnect to support
backwards compatibility for applications that rely on the pre-10.5
behavior for connection disconnect, which is to rollback the
transaction (in violation of the XA specification).
Signed-off-by: Kristian Nielsen <knielsen@knielsen-hq.org>
The failing test case validates Seconds_Behind_Master for a delayed
slave, while STOP SLAVE is executed during a delay. The test fixes
initially added to the test (commit b04c857596) added a table lock
to ensure a transaction could not finish before validating the
Seconds_Behind_Master field after SLAVE START, but did not address a
possibility that the transaction could finish before running the
STOP SLAVE command, which invalidates the validations for the rest
of the test case. Specifically, this would result in 1) a timeout in
“Waiting for table metadata lock” on the replica, which expects the
transaction to retry after slave restart and hit a lock conflict on
the locked tables (added in b04c857596), and 2) that
Seconds_Behind_Master should have increased, but did not.
The failure can be reproduced by synchronizing the slave to the master
before the MDEV-32265 echo statement (i.e. before the SLAVE STOP).
This patch fixes the test by adding a mechanism to use DEBUG_SYNC to
synchronize a MASTER_DELAY, rather than continually increase the
duration of the delay each time the test fails on buildbot. This is
to ensure that on slow machines, a delay does not pass before the
test gets a chance to validate results. Additionally, it decreases
overall test time because the test can continue immediately after
validation, thereby bypassing the remainder of a full delay for each
transaction.
Remove the test for MDEV-14528. This is supposed to test that parallel
replication from pre-10.0 master will update Seconds_Behind_Master. But
after MDEV-12179 the SQL thread is blocked from even beginning to fetch
events from the relay log due to FLUSH TABLES WITH READ LOCK, so the test
case is no longer testing what is was intended to. And pre-10.0 versions are
long since out of support, so does not seem worthwhile to try to rewrite the
test to work another way.
The root cause of the test failure is MDEV-34778. Briefly, depending on
exact timing during slave stop, the rli->sql_thread_caught_up flag may end
up with different value. If it ends up as "true", this causes
Seconds_Behind_Master to be 0 during next slave start; and this caused test
case timeout as the test was waiting for Seconds_Behind_Master to become
non-zero.
Signed-off-by: Kristian Nielsen <knielsen@knielsen-hq.org>
In a merge, an mtr.call_suppression was erroneously removed
for "Got an error writing communication packets". So when
the error would expectedly occur, the test would fail.
This patch adds this suppression back.
Note that the test will still fail due MDEV-34799, which
is to be fixed in 10.6.
When mysqldump is run to dump the `mysql` system database, it generates
INSERT statements into the table `mysql.gtid_slave_pos`.
After running the backup script
those inserts did not produce the expected gtid state on slave. In
particular the maximum of mysql.gtid_slave_pos.sub_id did not make
into
rpl_global_gtid_slave_state.last_sub_id
an in-memory object that is supposed to match the current state of the
table. And that was regardless of whether --gtid option was specified
or not. Later when the backup recipient server starts as slave
in *non-gtid* mode this desychronization may lead to a duplicate key
error.
This effect is corrected for --gtid mode mysqldump/mariadb-dump only
as the following. The fixes ensure the insert block of the dump
script is followed with a "summing-up" SET @global.gtid_slave_pos
assignment.
For the implemenation part, note a deferred print-out of
SET-gtid_slave_pos and associated comments is prefered over relocating
of the entire blocks if (opt_master,slave_data &&
do_show_master,slave_status) ... because of compatiblity
concern. Namely an error inside do_show_*() is handled in the new code
the same way, as early as, as before.
A regression test can be run in how-to-reproduce mode as well.
One affected mtr test observed.
rpl_mysqldump_slave.result "mismatch" shows now the new deferring print
of SET-gtid_slave_pos policy in action.
The test was missing a save_master_gtid.inc on the master,
leading to the slave thinking it was in sync after executing
sync_with_master_gtid.inc, despite not having executed the
latest transaction. This skipped transaction, XA COMMIT,
was supposed to error-to-be-ignored because its XID could not
be found, but be thrown out because the replication filters
would filter out the target database. However, if the slave
was able to stop before executing the transaction, then
the replication filer is reset (to empty), and when the
slave is later restarted, that transactions error would
no longer be ignored.
Additionally, as the test cases added in MDEV-33921 rely
on GTID synchronization, the test cases now force
master_use_gtid=slave_pos for consistency
MDEV-34274 did not fix the test failure. The test has a START SLAVE
UNTIL condition, where we can't use sync_with_master_gtid.inc,
wait_for_slave_to_start.inc, or wait_for_slave_to_stop.inc because
our MTR connection thread races with the start/stop of the SQL/IO
threads. So instead, for slave start, we prove the threads started
by waiting for the connection count to increase by 2; and for slave
stop, we wait for the processlist count to return to its pre start
slave number.
Note this is a backport of 8c8b3ab784
from 11.1.
The test rpl.rpl_change_master_demote used a `sleep 1` command
to give time for a START SLAVE UNTIL to start the slave threads
and wait for them to automatically die by UNTIL. On machines
with heavy load (especially MSAN bb builders), one second was
not enough, and the test would fail due to the IO thread
still being up.
This patch fixes the test by replacing the sleep with specific
conditions to wait for. The test cannot wait for the IO or SQL
threads to start, as it would be possible that they would be
started and stopped by the time the MTR executor would check
the slave status. So instead, we test for proof that they
existed via the Connections status variable being incremented
by at least 2 (Connections just shows the global thread id).
At this point, we still can't use the wait_for_slave_to_stop
helper, as the SQL/IO_Running fields of SHOW SLAVE STATUS
may not be updated yet. So instead, we use
information_schema.processlist, which would show the presence
of the Slave_SQL/IO threads. So to "wait for the slave to stop",
we wait for the Slave_SQL/IO threads to be gone from the
processlist.
There are two problems.
First, replication fails when XA transactions are used where the
slave has replicate_do_db set and the client has touched a different
database when running DML such as inserts. This is because XA
commands are not treated as keywords, and are thereby not exempt
from the replication filter. The effect of this is that during an XA
transaction, if its logged “use db” from the master is filtered out
by the replication filter, then XA END will be ignored, yet its
corresponding XA PREPARE will be executed in an invalid state,
thereby breaking replication.
Second, if the slave replicates an XA transaction which results in
an empty transaction, the XA START through XA PREPARE first phase of
the transaction won’t be binlogged, yet the XA COMMIT will be
binlogged. This will break replication in chain configurations.
The first problem is fixed by treating XA commands in
Query_log_event as keywords, thus allowing them to bypass the
replication filter. Note that Query_log_event::is_trans_keyword() is
changed to accept a new parameter to define its mode, to either
check for XA commands or regular transaction commands, but not both.
In addition, mysqlbinlog is adapted to use this mode so its
--database filter does not remove XA commands from its output.
The second problem fixed by overwriting the XA state in the XID
cache to be XA_ROLLBACK_ONLY, so at commit time, the server knows to
rollback the transaction and skip its binlogging. If the xid cache
is cleared before an XA transaction receives its completion command
(e.g. on server shutdown), then before reporting ER_XAER_NOTA when
the completion command is executed, the filter is first checked if
the database is ignored, and if so, the error is ignored.
Reviewed By:
============
Kristian Nielsen <knielsen@knielsen-hq.org>
Andrei Elkin <andrei.elkin@mariadb.com>
PURGE BINARY LOGS did not always purge binary logs. This commit fixes
some of the issues and adds notifications if a binary log cannot be
purged.
User visible changes:
- 'PURGE BINARY LOG TO log_name' and 'PURGE BINARY LOGS BEFORE date'
worked differently. 'TO' ignored 'slave_connections_needed_for_purge'
while 'BEFORE' did not. Now both versions ignores the
'slave_connections_needed_for_purge variable'.
- 'PURGE BINARY LOG..' commands now returns 'note' if a binary log cannot
be deleted like
Note 1375 Binary log 'master-bin.000004' is not purged because it is
the current active binlog
- Automatic binary log purges, based on date or size, will write a
note to the error log if a binary log matching the size or date
cannot yet be deleted.
- If 'slave_connections_needed_for_purge' is set from a config or
command line, it is set to 0 if Galera is enabled and 1 otherwise
(old default). This ensures that automatic binary log purge works
with Galera as before the addition of
'slave_connections_needed_for_purge'.
If the variable is changed to 0, a warning will be printed to the error
log.
Code changes:
- Added THD argument to several purge_logs related functions that needed
THD.
- Added 'interactive' options to purge_logs functions. This allowed
me to remove testing of sql_command == SQLCOM_PURGE.
- Changed purge_logs_before_date() to first check if log is applicable
before calling can_purge_logs(). This ensures we do not get a
notification for logs that does not match the remove criteria.
- MYSQL_BIN_LOG::can_purge_log() will write notifications to the user
or error log if a log cannot yet be removed.
- log_in_use() will return reason why a binary log cannot be removed.
Changes to keep code consistent:
- Moved checking of binlog_format for Galera to be after Galera is
initialized (The old check never worked). If Galera is enabled
we now change the binlog_format to ROW, with a warning, instead of
aborting the server. If this change happens a warning will be printed to
the error log.
- Print a warning if Galera or FLASHBACK changes the binlog_format
to ROW. Before it was done silently.
Reviewed by: Sergei Golubchik <serg@mariadb.com>,
Kristian Nielsen <knielsen@knielsen-hq.org>
The IO thread can report error code 2013 into the error log when it
is stopped during the initial connection process to the primary, as
well as when trying to read an event. However, because the IO thread
is being stopped, its connection to the primary is force-killed by
the signaling thread (see THD::awake_no_mutex()), and thereby these
connection errors should be ignored.
Reviewed By:
============
Kristian Nielsen <knielsen@knielsen-hq.org>
The current semi-sync binlog fail-over recovery process uses
rpl_semi_sync_slave_enabled==TRUE as its condition to truncate a
primary server’s binlog, as it is anticipating the server to re-join
a replication topology as a replica. However, for servers configured
with both rpl_semi_sync_master_enabled=1 and
rpl_semi_sync_slave_enabled=1, if a primary is just re-started (i.e.
retaining its role as master), it can truncate its binlog to drop
transactions which its replica(s) has already received and executed.
If this happens, when the replica reconnects, its gtid_slave_pos can
be ahead of the recovered primary’s gtid_binlog_pos, resulting in an
error state where the replica’s state is ahead of the primary’s.
This patch changes the condition for semi-sync recovery to truncate
the binlog to instead use the configuration variable
--init-rpl-role, when set to SLAVE. This allows for both
rpl_semi_sync_master_enabled and rpl_semi_sync_slave_enabled to be
set for a primary that is restarted, and no transactions will be
lost, so long as --init-rpl-role is not set to SLAVE.
Reviewed By:
============
Sergei Golubchik <serg@mariadb.com>
The special logic used by the memory storage engine
to keep slaves in sync with the master on a restart can
break replication. In particular, after a restart, the
master writes DELETE statements in the binlog for
each MEMORY-based table so the slave can empty its
data. If the DELETE is not executable, e.g. due to
invalid triggers, the slave will error and fail, whereas
the master will never see the problem.
Instead of DELETE statements, use TRUNCATE to
keep slaves in-sync with the master, thereby bypassing
triggers.
Reviewed By:
===========
Kristian Nielsen <knielsen@knielsen-hq.org>
Andrei Elkin <andrei.elkin@mariadb.com>
The test rpl.rpl_change_master_demote used a `sleep 1` command
to give time for a START SLAVE UNTIL to start the slave threads
and wait for them to automatically die by UNTIL. On machines
with heavy load (especially MSAN bb builders), one second was
not enough, and the test would fail due to the IO thread
still being up.
This patch fixes the test by replacing the sleep with specific
conditions to wait for. The test cannot wait for the IO or SQL
threads to start, as it would be possible that they would be
started and stopped by the time the MTR executor would check
the slave status. So instead, we test for proof that they
existed via the Connections status variable being incremented
by at least 2 (Connections just shows the global thread id).
At this point, we still can't use the wait_for_slave_to_stop
helper, as the SQL/IO_Running fields of SHOW SLAVE STATUS
may not be updated yet. So instead, we use
information_schema.processlist, which would show the presence
of the Slave_SQL/IO threads. So to "wait for the slave to stop",
we wait for the Slave_SQL/IO threads to be gone from the
processlist.
