Menu
I have the complex query that creates a temporary table that can take 5 seconds to run or more. This seems to be causing a deadlock when running another transaction on similar tables at the same time. I cannot reproduce locally, but on production I am able to have it happen 1 time every few days. (I logged the mysql errors)
The query is pretty complicated (shown at bottom of page); but you don't need to understand the logic; just that it selects from a bunch of tables and joins and can take awhile to run.
I also have a transaction that inserts into many of the same tables. I get a mysql error occasionally 1213: Deadlock found when trying to get lock; try restarting transaction.
Here is pseudocode for the transaction
START TRANSACTION
INSERT INTO phppos_sales
INSERT MANY RECORDS INTO phppos_sales_items
INSERT MANY RECORDS INTO phppos_sales_items_taxes
INSERT MANY RECORDS INTO phppos_sales_payments
END TRANSACTION
How do I go about resolving this deadlock? I tried changing isolation level to READ UNCOMMITTED but mysql settings wouldn't allow for this; and I need to make this work in a variety of environments where I don't have control of the server.
ERROR when changing isolation level:
Cannot execute statement: impossible to write to binary log since BINLOG_FORMAT = STATEMENT and at least one table uses a storage engine limited to row-based logging. InnoDB is limited to row-logging when transaction isolation level is READ COMMITTED or READ UNCOMMITTED.
INNODB ENGINE STATUS:
mysql> SHOW ENGINE INNODB STATUS;
| Type | Name | Status |
| InnoDB | |
=====================================
140520 12:00:17 INNODB MONITOR OUTPUT
=====================================
Per second averages calculated from the last 15 seconds
-----------------
BACKGROUND THREAD
-----------------
srv_master_thread loops: 1766819 1_second, 1766816 sleeps, 167043 10_second, 100947 background, 100945 flush
srv_master_thread log flush and writes: 1776023
----------
SEMAPHORES
----------
OS WAIT ARRAY INFO: reservation count 236559, signal count 288374
Mutex spin waits 546890, rounds 1796579, OS waits 33216
RW-shared spins 205374, rounds 5519210, OS waits 176937
RW-excl spins 5661, rounds 841678, OS waits 23933
Spin rounds per wait: 3.29 mutex, 26.87 RW-shared, 148.68 RW-excl
------------------------
LATEST FOREIGN KEY ERROR
------------------------
140520 11:27:44 Transaction:
TRANSACTION 86D125F, ACTIVE 0 sec inserting
mysql tables in use 1, locked 1
15 lock struct(s), heap size 3112, 6 row lock(s), undo log entries 2
MySQL thread id 1910245, OS thread handle 0x7fbf0042e700, query id 56114114 php-pos-web 10.181.16.33 phppoint update
INSERT INTO `phppos_sales_items_taxes` (`sale_id`, `item_id`, `line`, `name`, `percent`, `cumulative`) VALUES (11763, 1115, 3, 'PST', '8.000', '0')
Foreign key constraint fails for table `phppoint_fatpanda`.`phppos_sales_items_taxes`:
,
CONSTRAINT `phppos_sales_items_taxes_ibfk_1` FOREIGN KEY (`sale_id`) REFERENCES `phppos_sales_items` (`sale_id`)
Trying to add in child table, in index `PRIMARY` tuple:
DATA TUPLE: 8 fields;
0: len 4; hex 80002df3; asc - ;;
1: len 4; hex 8000045b; asc [;;
2: len 4; hex 80000003; asc ;;
3: len 3; hex 505354; asc PST;;
4: len 8; hex 8000000000080000; asc ;;
5: len 6; hex 0000086d125f; asc m _;;
6: len 7; hex 00000000000000; asc ;;
7: len 4; hex 80000000; asc ;;
But in parent table `phppoint_fatpanda`.`phppos_sales_items`, in index `PRIMARY`,
the closest match we can find is record:
PHYSICAL RECORD: n_fields 11; compact format; info bits 0
0: len 4; hex 80002df1; asc - ;;
1: len 4; hex 8000049a; asc ;;
2: len 4; hex 80000001; asc ;;
3: len 6; hex 0000086cfd29; asc l );;
4: len 7; hex f400000216012c; asc ,;;
5: len 0; hex ; asc ;;
6: len 0; hex ; asc ;;
7: len 11; hex 8000000000010000000000; asc ;;
8: len 11; hex 8000000000100000000000; asc ;;
9: len 11; hex 80000000002d0000000000; asc - ;;
10: len 4; hex 80000000; asc ;;
------------------------
LATEST DETECTED DEADLOCK
------------------------
140520 11:27:44
*** (1) TRANSACTION:
TRANSACTION 86D11A3, ACTIVE 2 sec fetching rows
mysql tables in use 9, locked 9
LOCK WAIT 364 lock struct(s), heap size 47544, 80177 row lock(s)
MySQL thread id 1910243, OS thread handle 0x7fbeb2090700, query id 56113840 10.181.26.42 phppoint Copying to tmp table
CREATE TEMPORARY TABLE phppos_sales_items_temp
(SELECT phppos_sales.deleted as deleted,phppos_sales.deleted_by as deleted_by, sale_time, date(sale_time) as sale_date, phppos_sales_items.sale_id, comment,payment_type, customer_id, employee_id,
phppos_items.item_id, NULL as item_kit_id, supplier_id, quantity_purchased, item_cost_price, item_unit_price, category,
discount_percent, (item_unit_price*quantity_purchased-item_unit_price*quantity_purchased*discount_percent/100) as subtotal,
phppos_sales_items.line as line, serialnumber, phppos_sales_items.description as description,
(item_unit_price*quantity_purchased-item_unit_price*quantity_purchased*discount_percent/100)+(item_unit_price*quantity_purchased-item_unit_price*quantity_purchased*discount_percent/100)*(SUM(CASE WHEN cumulative != 1 THEN percent ELSE 0 END)/100)
+(((item_unit_price*quantity_purchased-item_unit_price*quanti
*** (1) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 454941 page no 114 n bits 408 index `location_id` of table `phppoint_fatpanda`.`phppos_sales` trx id 86D11A3 lock mode S waiting
Record lock, heap no 335 PHYSICAL RECORD: n_fields 2; compact format; info bits 0
0: len 4; hex 80000001; asc ;;
1: len 4; hex 80002df3; asc - ;;
*** (2) TRANSACTION:
TRANSACTION 86D125D, ACTIVE 0 sec inserting
mysql tables in use 1, locked 1
23 lock struct(s), heap size 3112, 12 row lock(s), undo log entries 10
MySQL thread id 1910245, OS thread handle 0x7fbf0042e700, query id 56114091 php-pos-web 10.181.16.33 phppoint update
INSERT INTO `phppos_sales_items_taxes` (`sale_id`, `item_id`, `line`, `name`, `percent`, `cumulative`) VALUES (11763, 1178, 2, 'GST', '5.000', '0')
*** (2) HOLDS THE LOCK(S):
RECORD LOCKS space id 454941 page no 114 n bits 408 index `location_id` of table `phppoint_fatpanda`.`phppos_sales` trx id 86D125D lock_mode X locks rec but not gap
Record lock, heap no 335 PHYSICAL RECORD: n_fields 2; compact format; info bits 0
0: len 4; hex 80000001; asc ;;
1: len 4; hex 80002df3; asc - ;;
*** (2) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 454945 page no 386 n bits 288 index `PRIMARY` of table `phppoint_fatpanda`.`phppos_sales_items_taxes` trx id 86D125D lock_mode X insert intention waiting
Record lock, heap no 1 PHYSICAL RECORD: n_fields 1; compact format; info bits 0
0: len 8; hex 73757072656d756d; asc supremum;;
*** WE ROLL BACK TRANSACTION (2)
------------
TRANSACTIONS
------------
Trx id counter 86E47F7
Purge done for trx's n:o < 86E45C0 undo n:o < 0
History list length 1418
LIST OF TRANSACTIONS FOR EACH SESSION:
---TRANSACTION 86E47F6, not started
MySQL thread id 1913171, OS thread handle 0x7fbeb2090700, query id 56205829 10.181.26.42 phppoint
---TRANSACTION 0, not started
MySQL thread id 1913095, OS thread handle 0x7fbf005b4700, query id 56205830 localhost root
SHOW ENGINE INNODB STATUS
--------
FILE I/O
--------
I/O thread 0 state: waiting for completed aio requests (insert buffer thread)
I/O thread 1 state: waiting for completed aio requests (log thread)
I/O thread 2 state: waiting for completed aio requests (read thread)
I/O thread 3 state: waiting for completed aio requests (read thread)
I/O thread 4 state: waiting for completed aio requests (read thread)
I/O thread 5 state: waiting for completed aio requests (read thread)
I/O thread 6 state: waiting for completed aio requests (write thread)
I/O thread 7 state: waiting for completed aio requests (write thread)
I/O thread 8 state: waiting for completed aio requests (write thread)
I/O thread 9 state: waiting for completed aio requests (write thread)
Pending normal aio reads: 0 [0, 0, 0, 0] , aio writes: 0 [0, 0, 0, 0] ,
ibuf aio reads: 0, log i/o's: 0, sync i/o's: 0
Pending flushes (fsync) log: 0; buffer pool: 0
3599456 OS file reads, 9300371 OS file writes, 3988632 OS fsyncs
0.27 reads/s, 16384 avg bytes/read, 13.07 writes/s, 7.27 fsyncs/s
-------------------------------------
INSERT BUFFER AND ADAPTIVE HASH INDEX
-------------------------------------
Ibuf: size 1, free list len 84, seg size 86, 55915 merges
merged operations:
insert 68506, delete mark 4761, delete 38
discarded operations:
insert 0, delete mark 0, delete 0
Hash table size 2212699, node heap has 751 buffer(s)
5050.86 hash searches/s, 624.09 non-hash searches/s
---
LOG
---
Log sequence number 184365806376
Log flushed up to 184365806376
Last checkpoint at 184365791715
0 pending log writes, 0 pending chkp writes
2149282 log i/o's done, 3.47 log i/o's/second
----------------------
BUFFER POOL AND MEMORY
----------------------
Total memory allocated 1098907648; in additional pool allocated 0
Dictionary memory allocated 62951505
Buffer pool size 65536
Free buffers 1
Database pages 64784
Old database pages 23894
Modified db pages 88
Pending reads 0
Pending writes: LRU 0, flush list 0, single page 0
Pages made young 4073451, not young 0
0.07 youngs/s, 0.00 non-youngs/s
Pages read 3592230, created 1542046, written 6130789
0.27 reads/s, 18.00 creates/s, 6.40 writes/s
Buffer pool hit rate 1000 / 1000, young-making rate 0 / 1000 not 0 / 1000
Pages read ahead 0.00/s, evicted without access 0.00/s, Random read ahead 0.00/s
LRU len: 64784, unzip_LRU len: 0
I/O sum[382]:cur[0], unzip sum[0]:cur[0]
--------------
ROW OPERATIONS
--------------
0 queries inside InnoDB, 0 queries in queue
1 read views open inside InnoDB
Main thread process no. 15866, id 140457065543424, state: sleeping
Number of rows inserted 77431960, updated 1673031, deleted 160450, read 4825684197
1103.93 inserts/s, 2.53 updates/s, 0.00 deletes/s, 7772.15 reads/s
----------------------------
END OF INNODB MONITOR OUTPUT
============================
|
+-------
QUERY THIS IS PART OF DEADLOCK: (Along with transaction above)
CREATE temporary TABLE phppos_sales_items_temp
(SELECT
phppos_sales.deleted AS deleted,
phppos_sales.deleted_by AS deleted_by,
sale_time,
Date(sale_time) AS sale_date,
phppos_sales_items.sale_id,
comment,
payment_type,
customer_id,
employee_id,
phppos_items.item_id,
NULL AS item_kit_id,
supplier_id,
quantity_purchased,
item_cost_price,
item_unit_price,
category,
discount_percent,
( item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent / 100 ) AS subtotal,
phppos_sales_items.line AS line,
serialnumber,
phppos_sales_items.description AS description,
( item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent / 100 ) + (
item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent
/ 100 ) * ( Sum(CASE
WHEN cumulative != 1 THEN percent
ELSE 0
end) / 100 ) + ( ( (
item_unit_price * quantity_purchased
-
item_unit_price * quantity_purchased
*
discount_percent / 100 ) * (
Sum(CASE
WHEN cumulative != 1 THEN
percent
ELSE 0
end) / 100 ) + (
item_unit_price * quantity_purchased
-
item_unit_price *
quantity_purchased
*
discount_percent /
100
)
)
* ( Sum(CASE
WHEN
cumulative = 1 THEN percent
ELSE
0
end)
)
/ 100 )
AS total,
( item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent / 100 ) * (
Sum(CASE
WHEN cumulative != 1 THEN percent
ELSE 0
end) / 100 ) + ( ( ( item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent / 100 ) * ( Sum(
CASE
WHEN cumulative != 1
THEN
percent
ELSE 0
end) / 100 ) + (
item_unit_price * quantity_purchased
-
item_unit_price * quantity_purchased
*
discount_percent / 100 ) ) * ( Sum(
CASE
WHEN cumulative = 1 THEN percent
ELSE 0
end) ) / 100 ) AS tax,
( item_unit_price * quantity_purchased -
item_unit_price * quantity_purchased *
discount_percent / 100 ) - (
item_cost_price * quantity_purchased ) AS profit
FROM phppos_sales_items
INNER JOIN phppos_sales
ON phppos_sales_items.sale_id = phppos_sales.sale_id
INNER JOIN phppos_items
ON phppos_sales_items.item_id = phppos_items.item_id
LEFT OUTER JOIN phppos_suppliers
ON phppos_items.supplier_id = phppos_suppliers.person_id
LEFT OUTER JOIN phppos_sales_items_taxes
ON phppos_sales_items.sale_id =
phppos_sales_items_taxes.sale_id
AND phppos_sales_items.item_id =
phppos_sales_items_taxes.item_id
AND phppos_sales_items.line =
phppos_sales_items_taxes.line
WHERE sale_time BETWEEN "2014-04-01 00:00:00" AND "2014-04-30 23:59:59"
AND phppos_sales.location_id = '1'
AND phppos_sales.store_account_payment = 0
GROUP BY sale_id,
item_id,
line)
UNION ALL
(SELECT phppos_sales.deleted
AS
deleted
,
phppos_sales.deleted_by
AS deleted_by,
sale_time,
Date(sale_time)
AS
sale_date,
phppos_sales_item_kits.sale_id,
comment,
payment_type,
customer_id,
employee_id,
NULL
AS
item_id,
phppos_item_kits.item_kit_id,
''
AS
supplier_id,
quantity_purchased,
item_kit_cost_price,
item_kit_unit_price,
category,
discount_percent,
( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 )
AS
subtotal,
phppos_sales_item_kits.line
AS
line,
''
AS
serialnumber,
phppos_sales_item_kits.description
AS
description,
( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 )
+
( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 ) *
( Sum( CASE WHEN cumulative != 1 THEN percent ELSE 0 end) / 100 )
+ ( ( ( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 ) *
( Sum(CASE
WHEN cumulative != 1 THEN percent
ELSE 0
end)
/
100 ) + ( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased *
discount_percent / 100 ) ) * (
Sum(
CASE
WHEN cumulative = 1 THEN percent
ELSE 0
end) ) / 100 ) AS total,
( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 ) * ( Sum(
CASE
WHEN cumulative != 1 THEN percent
ELSE 0
end) / 100 ) + ( ( ( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased
*
discount_percent
/ 100 ) * ( Sum(
CASE
WHEN cumulative != 1 THEN percent
ELSE 0
end) / 100 ) + (
item_kit_unit_price * quantity_purchased
-
item_kit_unit_price *
quantity_purchased
* discount_percent / 100 ) ) *
(
Sum(CASE
WHEN cumulative = 1 THEN percent
ELSE 0
end) ) / 100 ) AS tax,
( item_kit_unit_price * quantity_purchased -
item_kit_unit_price * quantity_purchased * discount_percent / 100 ) - (
item_kit_cost_price * quantity_purchased ) AS profit
FROM phppos_sales_item_kits
INNER JOIN phppos_sales
ON phppos_sales_item_kits.sale_id = phppos_sales.sale_id
INNER JOIN phppos_item_kits
ON phppos_sales_item_kits.item_kit_id =
phppos_item_kits.item_kit_id
LEFT OUTER JOIN phppos_sales_item_kits_taxes
ON phppos_sales_item_kits.sale_id =
phppos_sales_item_kits_taxes.sale_id
AND phppos_sales_item_kits.item_kit_id =
phppos_sales_item_kits_taxes.item_kit_id
AND phppos_sales_item_kits.line =
phppos_sales_item_kits_taxes.line
WHERE sale_time BETWEEN "2014-04-01 00:00:00" AND "2014-04-30 23:59:59"
AND phppos_sales.location_id = '1'
AND phppos_sales.store_account_payment = 0
GROUP BY sale_id,
item_kit_id,
line)
ORDER BY sale_id,
line;
486
A deadlock can occur when transactions lock rows in multiple tables (through statements such as UPDATE or SELECT ... FOR UPDATE ), but in the opposite order.
MySQL Locks: Write LocksIt is the session that holds the lock of a table and can read and write data both from the table. It is the only session that accesses the table by holding a lock. And all other sessions cannot access the data of the table until the WRITE lock is released.
These memory tables never have values with data type like “BLOB” or “TEXT”. They use indexes which make them faster. Temporary table : The temporary tables could be very useful in some cases to keep temporary data. Temporary table is that they will be deleted when the current client session terminates.
Occasionally, you could greatly reduce the frequency of deadlocks by splitting a long transaction into smaller ones, so that locks are released sooner. Follow the rule: the transaction must not be longer than the business logic it requires to be.
When you combine a SELECT with a write statement such as INSERT INTO... or CREATE TABLE AS..., then MySQL has to establish a shared lock on the tables involved in the SELECT.
You have another concurrent transaction (2) that holds an exclusive lock on the table phppos_sales, so transaction (1) can't get its S-lock, and transaction (1) waits.
Then transaction (2) requests an X-lock on tale phppos_sales_items_taxes. But transaction (1) is already in queue to get its S-lock on that table, and transaction (2) must wait behind it in the queue.
Therefore transaction (2) is waiting on transaction (1), while transaction (1) is waiting on transaction (2). This is a classic deadlock.
This only happens once every few days because it depends on transaction (2) acquiring its first lock on phppos_sales before transaction (1) starts its SELECT. Then transaction (2) tries to acquire its second lock on phppos_sales_items_taxes after transaction (1) has its S-lock requests queued.
In other words, it's a race condition, and those are hard to reproduce.
If transaction (2) were to request locks on all the tables it needs as an atomic action, then there would be no way for transaction (1) to sneak in between the lock requests.
You can achieve this by explicitly using LOCK TABLES:
START TRANSACTION
LOCK TABLES phppos_sales WRITE, phppos_sales_items WRITE,
phppos_sales_items_taxes WRITE, ...other table(s)...
INSERT INTO phppos_sales
INSERT MANY RECORDS INTO phppos_sales_items
INSERT MANY RECORDS INTO phppos_sales_items_taxes
INSERT MANY RECORDS INTO phppos_sales_payments
UNLOCK TABLES;
COMMIT;
This does mean that transaction (1) that does the long-running SELECT has to wait for transaction (2) to finish its INSERTs and unlock its tables.
Or else if the SELECT is in progress first, then it means that transaction (2) has to wait for that to finish.
You could fill your temp table with no lock contention if you avoid using CREATE TABLE... SELECT or INSERT INTO... SELECT. That is, fetch the result-set of the SELECT back into your application, and then INSERT those rows into the temp table. That way the SELECT won't require any S-locks.
You could do the same thing with a cursor in a stored procedure.
As @BrendanF comments, you can also change your transaction isolation level to READ-COMMITTED instead of the default REPEATABLE-READ. You can either change the default transaction isolation level globally, or you can change isolation level on a session-by-session basis. This changes the semantics of transactions a bit, so you should read about the differences.
But it does eliminate the need for SELECT to do S-locks when reading from tables during those insert/select operations.
186
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With
