Understanding the business in Hibernate

1. The concept of database transactions

A database transaction is a unit of work consisting of one or more SQL statements that are dependent on each other, and the entire unit of work must be revoked if one of the SQL statements fails to execute.

In a concurrent environment, when multiple transactions access the same data resource at the same time, it can cause various concurrency problems, can be avoided by setting the transaction isolation level of the database, and pessimistic and optimistic locks can be used to resolve the concurrency problem of missing updates.

The database transaction must have acid characteristics:

A:atomic atomicity, the entire transaction is inseparable, either succeeds or is withdrawn.

C:consistency consistency, transactions do not break the integrity of relational data and the consistency of business logic, such as transfers, should ensure that the total deposit of the two accounts after the end of the transaction is unchanged.

I:isolation isolation, when multiple transactions manipulate the same data at the same time, each transaction has its own full data space d:durability persistence, as long as the transaction succeeds, the update to the database must be persisted, even if the system

System crashes and the database is restarted, the database can be restored to the state at the end of the transaction's success.

2. Transaction boundary declarations

As long as a transaction is declared, the database system automatically guarantees the acid characteristics of the transaction.

The Declaration transaction contains:

Start boundary of a transaction

Normal end boundary of transaction (COMMIT): Commit TRANSACTION, persist

Exception end boundary of a transaction (rollback): Undo transaction, database fallback to state before execution of transaction

The database supports two transaction modes:

autocommit mode: Each SQL statement is a separate transaction, and the transaction is automatically committed when the database finishes executing an SQL statement.

Manual commit mode: The start and end boundaries of a transaction must be explicitly specified by the client of the database

The method of transaction control for the JDBC connection class:

Setautocommit (Boolean autocommit) sets whether transactions are committed automatically, by default automatically

Commit () COMMIT Transaction

Rollback () revoking a transaction
Hibernate methods for controlling transactions:

1. Call Sessionfactory the Opensession method without parameters, get the connection from the connection pool, the session automatically sets the connection to commit the transaction mode manually.

Session session = Sessionfactory.opensession ();

If you call opensession with the connection parameter, you will need to set the manual submission yourself:

Connection.setautocommit (FALSE);

Session session = Sessionfactory.opensession (connection);
2. Declaring the start boundary of a transaction

Transaction tx = Session.begintransaction ();

3. Commit a transaction

Tx.commit ();

4. Revoking a transaction:

Tx.rollback ();

A session can correspond to multiple transactions, but priority should be given to a session that only corresponds to one transaction, and when a transaction ends or is revoked, the session is closed.

Whether the transaction succeeds or not, the session's close session should finally be called

At any one time, only one uncommitted transaction is allowed in a session, and two transactions cannot be started at the same time

3. Multi-transaction concurrency issues

There are 5 types of concurrency problems that can occur when you run multiple transactions concurrently to access the same data:

1. First category missing updates: When a transaction is revoked, the updates that have been committed by other transactions are overwritten

2. Dirty reads: One transaction reads uncommitted update data to another transaction

3. Virtual read: One transaction reads the newly inserted data that has been committed by another transaction

4. Non-repeatable reads: One transaction reads the updated data that has been committed by another transaction

5. Category Two missing updates: One transaction overwrites the updated data submitted by another transaction, non-repeatable read exceptions

1. First category missing updates:

T1 Start a transaction

T2 Start a transaction

T3 Enquiry Deposit balance of $1000

T4 Enquiry Deposit balance of $1000

T5 RMB 100, change the balance of deposit to 1100 yuan

T6 COMMIT Transaction

T7 withdraw $100 and change the deposit balance to $900

T8 Cancel the transaction, the balance of the account recovered to 1000 yuan

2. Dirty reads:

T1 Start a transaction

T2 Start a transaction
T3 Enquiry Deposit balance of $1000

T4

T5 withdraw $100 and change the deposit balance to $900

T6 Enquiry account balance of 900 yuan (dirty Read)

T7 Cancel the transaction, the balance of the account recovered to 1000 yuan

T8 RMB 100, deposit balance changed to $1000

T9 COMMIT Transaction

3. Virtual read:

T1 Start a transaction

T2 Start a transaction

T3 statistics website registered for 1000 people

T4 Registering a new user

T5 COMMIT Transaction

T6 Statistics Website Registration number is 1000 people (virtual reading)

T7 which statistic data is valid.

4. Non-repeatable reading:

T1 Start a transaction

T2 Start a transaction
T3 Enquiry account balance of 1000 yuan

T4 Enquiry account balance of 1000 yuan

T5 withdraw $100 and change the balance to $900

T6 COMMIT Transaction

T7 Enquiry account balance of 900 yuan

T8 whether the balance is 1100 yuan or 900+100 yuan.

5. The second category of missing updates:

T1 Start a transaction

T2 Start a transaction

T3 Enquiry account balance of 1000 yuan

T4 Enquiry account balance of 1000 yuan

T5 withdraw $100 and change the balance to $900

T6 COMMIT Transaction

T7 RMB 100, change the balance of deposit to 1100 yuan

T8 COMMIT Transaction

4. Database lock

The database must have the ability to isolate the individual transactions that are running concurrently, and the database uses locks to implement the transaction isolation

The release of sex.

Depending on the resources that the database can lock, it can be divided into the following locks:

Database-Level Locks

Table-Level Locks

Zone-Level locks

Page-level Locks

Key-value Lock: Locks a row of data in a database table with an index

Row-level Locks: Locks a single row of records in a database table

The greater the lock's blocking granularity, the higher the isolation, the lower the concurrency

Lock escalation refers to adjusting the granularity of locks, replacing multiple low-granularity locks with higher-granularity locks to reduce system load.

According to the degree of blockade, locks can be divided into:

Shared lock: For read data, non-exclusive, allows other transactions to read the locked resource at the same time, but does not allow its

His transaction is updated.

Locking: When a SELECT statement is executed, the database assigns a shared lock to the transaction, locking the queried