The problem is that when a worker thread is (user) killed in
wait_for_prior_commit, the event group may complete out-of-order since the
wait for prior commit was aborted by the kill.
This fix ensures that event groups will always complete in-order, even
in the error case. This is done in finish_event_group() by doing an
extra wait_for_prior_commit(), if necessary, that ignores kills.
This fix supersedes the fix for MDEV-30780, so the earlier fix for
that is reverted in this patch.
Also fix that an error from wait_for_prior_commit() inside
finish_event_group() would not signal the error to
wakeup_subsequent_commits().
Based on earlier work by Brandon Nesterenko and Andrei Elkin, with
some changes to simplify the semantics of wait_for_prior_commit() and
make the code more robust to future changes.
Reviewed-by: Andrei Elkin <andrei.elkin@mariadb.com>
Signed-off-by: Kristian Nielsen <knielsen@knielsen-hq.org>
Where a read-only server permits writes through replication, it
should not permit user connections to commit/rollback XA
transactions prepared via replication. The bug reported in
MDEV-30978 shows that this can happen. This is because there is no
read only check in the XA transaction logic, the most relevant one
occurs in ha_commit_trans() for normal statements/transactions.
This patch extends the XA transaction logic to check the read only
status of the server before performing an XA COMMIT or ROLLBACK.
Reviewed By:
Andrei Elkin <andrei.elkin@mariadb.com>
The new statistics is enabled by adding the "engine", "innodb" or "full"
option to --log-slow-verbosity
Example output:
# Pages_accessed: 184 Pages_read: 95 Pages_updated: 0 Old_rows_read: 1
# Pages_read_time: 17.0204 Engine_time: 248.1297
Page_read_time is time doing physical reads inside a storage engine.
(Writes cannot be tracked as these are usually done in the background).
Engine_time is the time spent inside the storage engine for the full
duration of the read/write/update calls. It uses the same code as
'analyze statement' for calculating the time spent.
The engine statistics is done with a generic interface that should be
easy for any engine to use. It can also easily be extended to provide
even more statistics.
Currently only InnoDB has counters for Pages_% and Undo_% status.
Engine_time works for all engines.
Implementation details:
class ha_handler_stats holds all engine stats. This class is included
in handler and THD classes.
While a query is running, all statistics is updated in the handler. In
close_thread_tables() the statistics is added to the THD.
handler::handler_stats is a pointer to where statistics should be
collected. This is set to point to handler::active_handler_stats if
stats are requested. If not, it is set to 0.
handler_stats has also an element, 'active' that is 1 if stats are
requested. This is to allow engines to avoid doing any 'if's while
updating the statistics.
Cloned or partition tables have the pointer set to the base table if
status are requested.
There is a small performance impact when using --log-slow-verbosity=engine:
- All engine calls in 'select' will be timed.
- IO calls for InnoDB reads will be timed.
- Incrementation of counters are done on local variables and accesses
are inline, so these should have very little impact.
- Statistics has to be reset for each statement for the THD and each
used handler. This is only 40 bytes, which should be neglectable.
- For partition tables we have to loop over all partitions to update
the handler_status as part of table_init(). Can be optimized in the
future to only do this is log-slow-verbosity changes. For this to work
we have to update handler_status for all opened partitions and
also for all partitions opened in the future.
Other things:
- Added options 'engine' and 'full' to log-slow-verbosity.
- Some of the new files in the test suite comes from Percona server, which
has similar status information.
- buf_page_optimistic_get(): Do not increment any counter, since we are
only validating a pointer, not performing any buf_pool.page_hash lookup.
- Added THD argument to save_explain_data_intern().
- Switched arguments for save_explain_.*_data() to have
always THD first (generates better code as other functions also have THD
first).
The problem is that a parallel replica would not immediately stop
running/queued transactions when issued STOP SLAVE. That is, it
allowed the current group of transactions to run, and sometimes the
transactions which belong to the next group could be started and run
through commit after STOP SLAVE was issued too, if the last group
had started committing. This would lead to long periods to wait for
all waiting transactions to finish.
This patch updates a parallel replica to try and abort immediately
and roll-back any ongoing transactions. The exception to this is any
transactions which are non-transactional (e.g. those modifying
sequences or non-transactional tables), and any prior transactions,
will be run to completion.
The specifics are as follows:
1. A new stage was added to SHOW PROCESSLIST output for the SQL
Thread when it is waiting for a replica thread to either rollback or
finish its transaction before stopping. This stage presents as
“Waiting for worker thread to stop”
2. Worker threads which error or are killed no longer perform GCO
cleanup if there is a concurrently running prior transaction. This
is because a worker thread scheduled to run in a future GCO could be
killed and incorrectly perform cleanup of the active GCO.
3. Refined cases when the FL_TRANSACTIONAL flag is added to GTID
binlog events to disallow adding it to transactions which modify
both transactional and non-transactional engines when the binlogging
configuration allow the modifications to exist in the same event,
i.e. when using binlog_direct_non_trans_update == 0 and
binlog_format == statement.
4. A few existing MTR tests relied on the completion of certain
transactions after issuing STOP SLAVE, and were re-recorded
(potentially with added synchronizations) under the new rollback
behavior.
Reviewed By
===========
Andrei Elkin <andrei.elkin@mariadb.com>
This commit contains a merge from 10.5-MDEV-29293-squash
into 10.6.
Although the bug MDEV-29293 was not reproducible with 10.6,
the fix contains several improvements for wsrep KILL query and
BF abort handling, and addresses the following issues:
* MDEV-30307 KILL command issued inside a transaction is
problematic for galera replication:
This commit will remove KILL TOI replication, so Galera side
transaction context is not lost during KILL.
* MDEV-21075 KILL QUERY maintains nodes data consistency but
breaks GTID sequence: This is fixed as well as KILL does not
use TOI, and thus does not change GTID state.
* MDEV-30372 Assertion in wsrep-lib state: This was caused by
BF abort or KILL when local transaction was in the middle
of group commit. This commit disables THD::killed handling
during commit, so the problem is avoided.
* MDEV-30963 Assertion failure !lock.was_chosen_as_deadlock_victim
in trx0trx.h:1065: The assertion happened when the victim was
BF aborted via MDL while it was committing. This commit changes
MDL BF aborts so that transactions which are committing cannot
be BF aborted via MDL. The RQG grammar attached in the issue
could not reproduce the crash anymore.
Original commit message from 10.5 fix:
MDEV-29293 MariaDB stuck on starting commit state
The problem seems to be a deadlock between KILL command execution
and BF abort issued by an applier, where:
* KILL has locked victim's LOCK_thd_kill and LOCK_thd_data.
* Applier has innodb side global lock mutex and victim trx mutex.
* KILL is calling innobase_kill_query, and is blocked by innodb
global lock mutex.
* Applier is in wsrep_innobase_kill_one_trx and is blocked by
victim's LOCK_thd_kill.
The fix in this commit removes the TOI replication of KILL command
and makes KILL execution less intrusive operation. Aborting the
victim happens now by using awake_no_mutex() and ha_abort_transaction().
If the KILL happens when the transaction is committing, the
KILL operation is postponed to happen after the statement
has completed in order to avoid KILL to interrupt commit
processing.
Notable changes in this commit:
* wsrep client connections's error state may remain sticky after
client connection is closed. This error message will then pop
up for the next client session issuing first SQL statement.
This problem raised with test galera.galera_bf_kill.
The fix is to reset wsrep client error state, before a THD is
reused for next connetion.
* Release THD locks in wsrep_abort_transaction when locking
innodb mutexes. This guarantees same locking order as with applier
BF aborting.
* BF abort from MDL was changed to do BF abort on server/wsrep-lib
side first, and only then do the BF abort on InnoDB side. This
removes the need to call back from InnoDB for BF aborts which originate
from MDL and simplifies the locking.
* Removed wsrep_thd_set_wsrep_aborter() from service_wsrep.h.
The manipulation of the wsrep_aborter can be done solely on
server side. Moreover, it is now debug only variable and
could be excluded from optimized builds.
* Remove LOCK_thd_kill from wsrep_thd_LOCK/UNLOCK to allow more
fine grained locking for SR BF abort which may require locking
of victim LOCK_thd_kill. Added explicit call for
wsrep_thd_kill_LOCK/UNLOCK where appropriate.
* Wsrep-lib was updated to version which allows external
locking for BF abort calls.
Changes to MTR tests:
* Disable galera_bf_abort_group_commit. This test is going to
be removed (MDEV-30855).
* Make galera_var_retry_autocommit result more readable by echoing
cases and expectations into result. Only one expected result for
reap to verify that server returns expected status for query.
* Record galera_gcache_recover_manytrx as result file was incomplete.
Trivial change.
* Make galera_create_table_as_select more deterministic:
Wait until CTAS execution has reached MDL wait for multi-master
conflict case. Expected error from multi-master conflict is
ER_QUERY_INTERRUPTED. This is because CTAS does not yet have open
wsrep transaction when it is waiting for MDL, query gets interrupted
instead of BF aborted. This should be addressed in separate task.
* A new test galera_bf_abort_registering to check that registering trx gets
BF aborted through MDL.
* A new test galera_kill_group_commit to verify correct behavior
when KILL is executed while the transaction is committing.
Co-authored-by: Seppo Jaakola <seppo.jaakola@iki.fi>
Co-authored-by: Jan Lindström <jan.lindstrom@galeracluster.com>
Signed-off-by: Julius Goryavsky <julius.goryavsky@mariadb.com>
The problem seems to be a deadlock between KILL command execution
and BF abort issued by an applier, where:
* KILL has locked victim's LOCK_thd_kill and LOCK_thd_data.
* Applier has innodb side global lock mutex and victim trx mutex.
* KILL is calling innobase_kill_query, and is blocked by innodb
global lock mutex.
* Applier is in wsrep_innobase_kill_one_trx and is blocked by
victim's LOCK_thd_kill.
The fix in this commit removes the TOI replication of KILL command
and makes KILL execution less intrusive operation. Aborting the
victim happens now by using awake_no_mutex() and ha_abort_transaction().
If the KILL happens when the transaction is committing, the
KILL operation is postponed to happen after the statement
has completed in order to avoid KILL to interrupt commit
processing.
Notable changes in this commit:
* wsrep client connections's error state may remain sticky after
client connection is closed. This error message will then pop
up for the next client session issuing first SQL statement.
This problem raised with test galera.galera_bf_kill.
The fix is to reset wsrep client error state, before a THD is
reused for next connetion.
* Release THD locks in wsrep_abort_transaction when locking
innodb mutexes. This guarantees same locking order as with applier
BF aborting.
* BF abort from MDL was changed to do BF abort on server/wsrep-lib
side first, and only then do the BF abort on InnoDB side. This
removes the need to call back from InnoDB for BF aborts which originate
from MDL and simplifies the locking.
* Removed wsrep_thd_set_wsrep_aborter() from service_wsrep.h.
The manipulation of the wsrep_aborter can be done solely on
server side. Moreover, it is now debug only variable and
could be excluded from optimized builds.
* Remove LOCK_thd_kill from wsrep_thd_LOCK/UNLOCK to allow more
fine grained locking for SR BF abort which may require locking
of victim LOCK_thd_kill. Added explicit call for
wsrep_thd_kill_LOCK/UNLOCK where appropriate.
* Wsrep-lib was updated to version which allows external
locking for BF abort calls.
Changes to MTR tests:
* Disable galera_bf_abort_group_commit. This test is going to
be removed (MDEV-30855).
* Record galera_gcache_recover_manytrx as result file was incomplete.
Trivial change.
* Make galera_create_table_as_select more deterministic:
Wait until CTAS execution has reached MDL wait for multi-master
conflict case. Expected error from multi-master conflict is
ER_QUERY_INTERRUPTED. This is because CTAS does not yet have open
wsrep transaction when it is waiting for MDL, query gets interrupted
instead of BF aborted. This should be addressed in separate task.
* A new test galera_kill_group_commit to verify correct behavior
when KILL is executed while the transaction is committing.
Co-authored-by: Seppo Jaakola <seppo.jaakola@iki.fi>
Co-authored-by: Jan Lindström <jan.lindstrom@galeracluster.com>
Signed-off-by: Julius Goryavsky <julius.goryavsky@mariadb.com>
This is a backport from 10.5.
The problem seems to be a deadlock between KILL command execution
and BF abort issued by an applier, where:
* KILL has locked victim's LOCK_thd_kill and LOCK_thd_data.
* Applier has innodb side global lock mutex and victim trx mutex.
* KILL is calling innobase_kill_query, and is blocked by innodb
global lock mutex.
* Applier is in wsrep_innobase_kill_one_trx and is blocked by
victim's LOCK_thd_kill.
The fix in this commit removes the TOI replication of KILL command
and makes KILL execution less intrusive operation. Aborting the
victim happens now by using awake_no_mutex() and ha_abort_transaction().
If the KILL happens when the transaction is committing, the
KILL operation is postponed to happen after the statement
has completed in order to avoid KILL to interrupt commit
processing.
Notable changes in this commit:
* wsrep client connections's error state may remain sticky after
client connection is closed. This error message will then pop
up for the next client session issuing first SQL statement.
This problem raised with test galera.galera_bf_kill.
The fix is to reset wsrep client error state, before a THD is
reused for next connetion.
* Release THD locks in wsrep_abort_transaction when locking
innodb mutexes. This guarantees same locking order as with applier
BF aborting.
* BF abort from MDL was changed to do BF abort on server/wsrep-lib
side first, and only then do the BF abort on InnoDB side. This
removes the need to call back from InnoDB for BF aborts which originate
from MDL and simplifies the locking.
* Removed wsrep_thd_set_wsrep_aborter() from service_wsrep.h.
The manipulation of the wsrep_aborter can be done solely on
server side. Moreover, it is now debug only variable and
could be excluded from optimized builds.
* Remove LOCK_thd_kill from wsrep_thd_LOCK/UNLOCK to allow more
fine grained locking for SR BF abort which may require locking
of victim LOCK_thd_kill. Added explicit call for
wsrep_thd_kill_LOCK/UNLOCK where appropriate.
* Wsrep-lib was updated to version which allows external
locking for BF abort calls.
Changes to MTR tests:
* Disable galera_bf_abort_group_commit. This test is going to
be removed (MDEV-30855).
* Record galera_gcache_recover_manytrx as result file was incomplete.
Trivial change.
* Make galera_create_table_as_select more deterministic:
Wait until CTAS execution has reached MDL wait for multi-master
conflict case. Expected error from multi-master conflict is
ER_QUERY_INTERRUPTED. This is because CTAS does not yet have open
wsrep transaction when it is waiting for MDL, query gets interrupted
instead of BF aborted. This should be addressed in separate task.
* A new test galera_kill_group_commit to verify correct behavior
when KILL is executed while the transaction is committing.
Co-authored-by: Seppo Jaakola <seppo.jaakola@iki.fi>
Co-authored-by: Jan Lindström <jan.lindstrom@galeracluster.com>
Signed-off-by: Julius Goryavsky <julius.goryavsky@mariadb.com>
In MariaDB, we have a confusing problem where:
* The transaction_isolation option can be set in a configuration file, but it cannot be set dynamically.
* The tx_isolation system variable can be set dynamically, but it cannot be set in a configuration file.
Therefore, we have two different names for the same thing in different contexts. This is needlessly confusing, and it complicates the documentation. The same thing applys for transaction_read_only.
MySQL 5.7 solved this problem by making them into system variables. https://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-20.html
This commit takes a similar approach by adding new system variables and marking the original ones as deprecated. This commit also resolves some legacy problems related to SET STATEMENT and transaction_isolation.
This patch introduces a new way of handling UPDATE and DELETE commands at
the top level after the parsing phase. This new way of processing update
and delete statements can be seen in the implementation of the prepare()
and execute() methods from the new Sql_cmd_dml class. This class derived
from the Sql_cmd class can be considered as an interface class for processing
such commands as SELECT, INSERT, UPDATE, DELETE and other comands
manipulating data in tables.
With this patch processing of update and delete statements after parsing
proceeds by the following schema:
- precheck of the access rights is performed for the used tables
- the used tables are opened
- context analysis phase is performed for the statement
- the used tables are locked
- the statement is optimized and executed
- clean-up is performed for the statement
The implementation of the method Sql_cmd_dml::execute() adheres this schema.
The virtual functions of the class Sql_cmd_dml used for precheck of the
access rights, context analysis, optimization and execution allow to adjust
this schema for processing data manipulation statements of any types.
This schema of processing data manipulation statements is taken from the
current MySQL code. Moreover the definition the class Sql_cmd_dml introduced
in this patch is almost a full replica of such class in the existing MySQL.
However the implementation of the derived classes for update and delete
statements is quite different. This implementation employs the JOIN class
for all kinds of update and delete statements. It allows to perform main
bulk of context analysis actions by the function JOIN::prepare(). This
guarantees that characteristics and properties of the statement tree
discovered for optimization phase when doing context analysis are the same
for single-table and multi-table updates and deletes.
With this patch the following functions are gone:
mysql_prepare_update(), mysql_multi_update_prepare(),
mysql_update(), mysql_multi_update(),
mysql_prepare_delete(), mysql_multi_delete_prepare(), mysql_delete().
The code within these functions have been used as much as possible though.
The functions mysql_test_update() and mysql_test_delete() are also not
needed anymore. The method Sql_cmd_dml::prepare() serves processing
- update/delete statement
- PREPARE stmt FROM "<update/delete statement>"
- EXECUTE stmt when stmt is prepared from update/delete statement.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
This patch also fixes some bugs detected by valgrind after this
patch:
- Not enough copy_func elements was allocated by Create_tmp_table() which
causes an memory overwrite in Create_tmp_table::add_fields()
I added an ASSERT() to be able to detect this also without valgrind.
The bug was that TMP_TABLE_PARAM::copy_fields was not correctly set
when calling create_tmp_table().
- Aria::empty_bits is not allocated if there is no varchar/char/blob
fields in the table. Fixed code to take this into account.
This cannot cause any issues as this is just a memory access
into other Aria memory and the content of the memory would not be used.
- Aria::last_key_buff was not allocated big enough. This may have caused
issues with rtrees and ma_extra(HA_EXTRA_REMEMBER_POS) as they
would use the same memory area.
- Aria and MyISAM didn't take extended key parts into account, which
caused problems when copying rec_per_key from engine to sql level.
- Mark asan builds with 'asan' in version strihng to detect these in
not_valgrind_build.inc.
This is needed to not have main.sp-no-valgrind fail with asan.
Extended keys works by first checking if the engine supports extended
keys.
If yes, it extends secondary key with primary key components and mark the
secondary keys as HA_EXT_NOSAME (unique).
If we later notice that there where no primary key, the extended key
information for secondary keys in share->key_info is reset. However the
key_info->flag HA_EXT_NOSAME was not reset!
This causes some strange things to happen:
- Tables that have no primary key or secondary index that contained the
primary key would be wrongly optimized as the secondary key could be
thought to be unique when it was not and not unique when it was.
- The problem was not shown in EXPLAIN because of a bug in
create_ref_for_key() that caused EQ_REF to be displayed by EXPLAIN as REF
when extended keys where used and the secondary key contained the primary
key.
This is fixed with:
- Removed wrong test in make_join_select() which did not detect that key
where unique when a secondary key contains the primary.
- Moved initialization of extended keys from create_key_infos() to
init_from_binary_frm_image() after we know if there is a usable primary
key or not. One disadvantage with this approach is that
key_info->key_parts may have not used slots (for keys we thought could
be extended but could not). Fixed by adding a check for unused key_parts
to copy_keys_from_share().
Other things:
- Simplified copying of first key part in create_key_infos().
- Added a lot of code comments in code that I had to check as part of
finding the issue.
- Fixed some indentation.
- Replaced a couple of looks using references to pointers in C
context where the reference does not give any benefit.
- Updated Aria and Maria to not assume the all key_info->rec_per_key
are in one memory block (this could happen when using dervived
tables with many keys).
- Fixed a bug where key_info->rec_per_key where not allocated
- Optimized TABLE::add_tmp_key() to only call alloc() once.
(No logic changes)
Test case changes:
- innodb_mysql.test changed index as an index the optimizer thought
was unique, was not. (Table had no primary key)
TODO:
- Move code that checks for partial or too long keys to the primary loop
earlier that initally decides if we should add extended key fields.
This is needed to ensure that HA_EXT_NOSAME is not set for partial or
too long keys. It will also shorten the current code notable.
