SECONDARY INDEX IN INNODB
This is a follow-up patch.
This patch moves part of the new test coverage to a test
file that is only run on debug builds since it used debug-
only features and therefore broke the test case on
release builds.
SECONDARY INDEX IN INNODB
The patches for Bug#11751388 and Bug#11784056 enabled concurrent
reads while creating secondary indexes in InnoDB. However, they
introduced a regression. This regression occured if ALTER TABLE
failed after the index had been added, for example during the
lock upgrade needed to update .FRM. If this happened, InnoDB
and the server got out of sync with regards to which indexes
actually existed. Therefore the patch for Bug#11815600 again
disabled concurrent reads.
This patch re-enables concurrent reads. The original regression
is fixed by splitting the ADD INDEX operation into two parts.
First the new index is created but not made active. This is
done while concurrent reads are allowed. The second part of
the operation makes the index active (or reverts the change).
This is done after lock upgrade, which prevents the original
regression.
In order to implement this change, the patch changes the storage
API for in-place index creation. handler::add_index() is split
into two functions, handler_add_index() and
handler::final_add_index(). The former for creating indexes without
making them visible and the latter for commiting (i.e. making
visible) new indexes or reverting the changes.
Large parts of this patch were written by Marko Mäkelä.
Test case added to innodb_mysql_lock.test.
KEY NO 0 FOR TABLE IN ERROR LOG
With the changes made by the patches for Bug#11751388 and
Bug#11784056, concurrent reads are allowed while secondary
indexes are created in InnoDB. This means that the metadata
lock on the affected table is not upgraded to exclusive
until the .FRM is updated at the end of ALTER TABLE processing.
The problem was that if this lock upgrade failed for some
reason (e.g. timeout), the index information in the server
and inside InnoDB would be out of sync. This would happen
since the add index operation already was committed inside
InnoDB but the table metadata inside the server had not been
updated yet.
This patch fixes the problem by (for now) reverting the
effects of the patches for Bug#11751388 and Bug#11784056.
Concurrent reads will now again be blocked during creation
of secondary indexes in InnoDB.
Test case added to innodb_mysql_lock.test.
The problem was that if a query accessing a view was blocked due to
conflicting locks on tables in the view definition, it would be possible
for a different connection to alter the view definition before the view
query completed. When the view query later resumed, it used the old view
definition. This meant that if the view query was later repeated inside
the same transaction, the two executions of the query would give different
results, thus breaking repeatable read. (The first query used the old
view definition, the second used the new view definition).
This bug is no longer repeatable with the recent changes to the metadata
locking subsystem (revno: 3040). The view query will no longer back-off
and release the lock on the view definiton. Instead it will wait for
the conflicting lock(s) to go away while keeping the view definition lock.
This means that it is no longer possible for a concurrent connection to
alter the view definition. Instead, any such attempt will be blocked.
In the case from the bug report where the same view query was executed
twice inside the same transaction, any ALTER VIEW from other connections
will now be blocked until the transaction has completed (or aborted).
The view queries will therefore use the same view definition and we will
have repeatable read.
Test case added to innodb_mysql_lock.test.
This patch contains no code changes.
The problem was that OPTMIZE TABLE was allowed to run on a table
in use by a transaction in a different connection. This caused
repeatable read to break.
This bug was fixed by the introduction of metadata locking, WL#4284.
OPTIMIZE TABLE will now be blocked until the transaction using the
table, has ended.
This patch contains a regression test added to innodb_mysql_lock.test
and no code changes.
Add a wait-for graph based deadlock detector to the
MDL subsystem.
Fixes bug #46272 "MySQL 5.4.4, new MDL: unnecessary deadlock" and
bug #37346 "innodb does not detect deadlock between update and
alter table".
The first bug manifested itself as an unwarranted abort of a
transaction with ER_LOCK_DEADLOCK error by a concurrent ALTER
statement, when this transaction tried to repeat use of a
table, which it has already used in a similar fashion before
ALTER started.
The second bug showed up as a deadlock between table-level
locks and InnoDB row locks, which was "detected" only after
innodb_lock_wait_timeout timeout.
A transaction would start using the table and modify a few
rows.
Then ALTER TABLE would come in, and start copying rows
into a temporary table. Eventually it would stumble on
the modified records and get blocked on a row lock.
The first transaction would try to do more updates, and get
blocked on thr_lock.c lock.
This situation of circular wait would only get resolved
by a timeout.
Both these bugs stemmed from inadequate solutions to the
problem of deadlocks occurring between different
locking subsystems.
In the first case we tried to avoid deadlocks between metadata
locking and table-level locking subsystems, when upgrading shared
metadata lock to exclusive one.
Transactions holding the shared lock on the table and waiting for
some table-level lock used to be aborted too aggressively.
We also allowed ALTER TABLE to start in presence of transactions
that modify the subject table. ALTER TABLE acquires
TL_WRITE_ALLOW_READ lock at start, and that block all writes
against the table (naturally, we don't want any writes to be lost
when switching the old and the new table). TL_WRITE_ALLOW_READ
lock, in turn, would block the started transaction on thr_lock.c
lock, should they do more updates. This, again, lead to the need
to abort such transactions.
The second bug occurred simply because we didn't have any
mechanism to detect deadlocks between the table-level locks
in thr_lock.c and row-level locks in InnoDB, other than
innodb_lock_wait_timeout.
This patch solves both these problems by moving lock conflicts
which are causing these deadlocks into the metadata locking
subsystem, thus making it possible to avoid or detect such
deadlocks inside MDL.
To do this we introduce new type-of-operation-aware metadata
locks, which allow MDL subsystem to know not only the fact that
transaction has used or is going to use some object but also what
kind of operation it has carried out or going to carry out on the
object.
This, along with the addition of a special kind of upgradable
metadata lock, allows ALTER TABLE to wait until all
transactions which has updated the table to go away.
This solves the second issue.
Another special type of upgradable metadata lock is acquired
by LOCK TABLE WRITE. This second lock type allows to solve the
first issue, since abortion of table-level locks in event of
DDL under LOCK TABLES becomes also unnecessary.
Below follows the list of incompatible changes introduced by
this patch:
- From now on, ALTER TABLE and CREATE/DROP TRIGGER SQL (i.e. those
statements that acquire TL_WRITE_ALLOW_READ lock)
wait for all transactions which has *updated* the table to
complete.
- From now on, LOCK TABLES ... WRITE, REPAIR/OPTIMIZE TABLE
(i.e. all statements which acquire TL_WRITE table-level lock) wait
for all transaction which *updated or read* from the table
to complete.
As a consequence, innodb_table_locks=0 option no longer applies
to LOCK TABLES ... WRITE.
- DROP DATABASE, DROP TABLE, RENAME TABLE no longer abort
statements or transactions which use tables being dropped or
renamed, and instead wait for these transactions to complete.
- Since LOCK TABLES WRITE now takes a special metadata lock,
not compatible with with reads or writes against the subject table
and transaction-wide, thr_lock.c deadlock avoidance algorithm
that used to ensure absence of deadlocks between LOCK TABLES
WRITE and other statements is no longer sufficient, even for
MyISAM. The wait-for graph based deadlock detector of MDL
subsystem may sometimes be necessary and is involved. This may
lead to ER_LOCK_DEADLOCK error produced for multi-statement
transactions even if these only use MyISAM:
session 1: session 2:
begin;
update t1 ... lock table t2 write, t1 write;
-- gets a lock on t2, blocks on t1
update t2 ...
(ER_LOCK_DEADLOCK)
- Finally, support of LOW_PRIORITY option for LOCK TABLES ... WRITE
was abandoned.
LOCK TABLE ... LOW_PRIORITY WRITE from now on has the same
priority as the usual LOCK TABLE ... WRITE.
SELECT HIGH PRIORITY no longer trumps LOCK TABLE ... WRITE in
the wait queue.
- We do not take upgradable metadata locks on implicitly
locked tables. So if one has, say, a view v1 that uses
table t1, and issues:
LOCK TABLE v1 WRITE;
FLUSH TABLE t1; -- (or just 'FLUSH TABLES'),
an error is produced.
In order to be able to perform DDL on a table under LOCK TABLES,
the table must be locked explicitly in the LOCK TABLES list.
Bug #42147 Concurrent DML and LOCK TABLE ... READ for InnoDB
table cause warnings in errlog
Concurrent execution of LOCK TABLES ... READ statement and DML statements
affecting the same InnoDB table on debug builds of MySQL server might lead
to "Found lock of type 6 that is write and read locked" warnings appearing
in error log.
The problem is that the table-level locking code allows a thread to acquire
TL_READ_NO_INSERT lock on a table even if there is another thread which holds
TL_WRITE_ALLOW_WRITE lock on the same table. At the same time, the locking
code assumes that that such locks are incompatible (for example, see check_locks()).
This doesn't lead to any problems other than warnings in error log for
debug builds of server since for InnoDB tables TL_READ_NO_INSERT type of
lock is only used for LOCK TABLES and for this statement InnoDB also
performs its own table-level locking.
Unfortunately, the table lock compatibility matrix cannot be updated to disallow
TL_READ_NO_INSERT when another thread holds TL_WRITE_ALLOW_WRITE without
causing starvation of LOCK TABLE READ in InnoDB under high write load.
This patch therefore contains no code changes.
The issue will be fixed later when LOCK TABLE READ has been updated
to not use table locks. This bug will therefore be marked as
"To be fixed later".
Code comment in thr_lock.c expanded to clarify the issue and a
test case based on the bug description added to innodb_mysql_lock.test.
Note that a global suppression rule has been added to both MTR v1 and v2
for the "Found lock of type 6 that is write and read locked" warning.
These suppression rules must be removed once this bug is properly fixed.
Bug #22876 Four-way deadlock
This bug was fixed as a part of Bug#989
"If DROP TABLE while there's an active transaction, wrong binlog order"
A statement which would have caused circular wait will now
be aborted with ER_LOCK_DEADLOCK.
Test case based on bug description added to innodb_mysql_lock.test.
Note that innodb_lock_wait_timeout is set to 5 mins to
prevent race conditions in the test.
