I am trying something with transaction levels on mysql.
I have Session S1 and Session S2. S1 is working with default isolation lavel Repeatable Read. For S2 i set Isolation Level Serializable.
Here is the szenario:
S1:
set innodb_lock_wait_timeout = 5;
start transaction;
S2:
set session transaction isolation level serializable;
start transaction;
select count(*) from produkt;
S1:
select count(*) from produkt;
update kategorie set bezeichnung = 'Smartphone' where kategorieid = 1;
S2:
insert into produkt(produktid, bezeichnung, kategorieid_fk) values (201, 'iPhone 8z', 1);
Can someone explain, why the insert into produkt
from S2 is now blocked?
Here is the Tableschema:
-- Exportiere Datenbank Struktur für transaktiondb
CREATE DATABASE IF NOT EXISTS `transaktiondb`;
USE `transaktiondb`;
-- Exportiere Struktur von Tabelle transaktiondb.kategorie
CREATE TABLE IF NOT EXISTS `kategorie` (
`KategorieID` int(11) NOT NULL AUTO_INCREMENT,
`Bezeichnung` varchar(255) NOT NULL,
PRIMARY KEY (`KategorieID`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;
-- Exportiere Struktur von Tabelle transaktiondb.produkt
CREATE TABLE IF NOT EXISTS `produkt` (
`ProduktID` int(11) NOT NULL AUTO_INCREMENT,
`Bezeichnung` varchar(255) NOT NULL,
`KategorieID_FK` int(11) NOT NULL,
PRIMARY KEY (`ProduktID`),
KEY `fk_Produkt_Kategorie_idx` (`KategorieID_FK`),
CONSTRAINT `fk_Produkt_Kategorie` FOREIGN KEY (`KategorieID_FK`) REFERENCES `kategorie` (`KategorieID`) ON DELETE NO ACTION ON UPDATE NO ACTION
) ENGINE=InnoDB DEFAULT CHARSET=utf8;
You have a deadlock which MySQL describes as:
A situation where different transactions are unable to proceed, because each holds a lock that the other needs. Because both transactions are waiting for a resource to become available, neither will ever release the locks it holds.
Looking closer at the How to Minimize and Handle Deadlocks reads:
InnoDB uses automatic row-level locking. You can get deadlocks even in the case of transactions that just insert or delete a single row. That is because these operations are not really “atomic”; they automatically set locks on the (possibly several) index records of the row inserted or deleted.
Regarding the nature of innodb_lock_wait_timeout
the documentation describe that it only applies to scenarios where inno_db_detect
is disabled, which is not the default configuration:
The lock wait timeout value does not apply to deadlocks when innodb_deadlock_detect is enabled (the default) because InnoDB detects deadlocks immediately and rolls back one of the deadlocked transactions. When innodb_deadlock_detect is disabled, InnoDB relies on innodb_lock_wait_timeout for transaction rollback when a deadlock occurs. See Section 14.7.5.2, “Deadlock Detection and Rollback”.
Also, as a general tip, you'll often want to use start transaction
followed as quickly as possible by a commit
or rollback
and not leave the session "hanging" to minimize these issues.